Laboratory University of Michigan

THE REPTILIAN VESTIBULAR AND CEREBELLAR
GRAY WITH FI BER CONNECTIONS

J EAN K . WESTON
Laboratory of Comparative Neurology. Department of Anatomy.
University of Michigan
TWENTY-SEVEN FIGURES
(Accepted for publication September 21. 1935)
CONTENTS
Introduction ........................................................
Materials and methods ..............................................
General survey of the literature ......................................
Cytoarchitectonic structure of the vestibular region .....................
General introduction .............................................
General survey and descriptions of vestibular subdivisions as based on
the material ............................................
Nucleus vestibularis ventrolateralis ............................
Nucleus vestibularis tangentialis ..............................
Nucleus vestibularis ventromedialis ............................
Nucleus vestibularis descendens ...............................
Nucleus vestibularis dorsolateralis .............................
Nucleus vestibularis superior .................................
Nucleus laminaris ...........................................
Cytoarchitectonic structure of the cerebellum ..........................
General introduction .............................................
Nucleus lateralis cerebelli ....................................
Nucleus medialis cerebelli ....................................
Description of cerebellar nuclei ..................................
Lateral reticular gray .......................................
Description of cerebellar cortex ..................................
Direct root fibers of the acoustico-facial complex
Fiber connections of the vestibular region
Sensory root of the VI I th nerve
Motor root of the VI I th nerve ...............................
.............................
...................
.............................
94
95
95
96
96
97
98
103
106
109
111
115
117
120
120
121
121
125
128
129
131
131
131
133
A dissertation submitted i n partial fulfillment of the requirements for the
degree of doctor of philosophy .
93
94 J EAN K . WESTON
Acoustic nerve .............................................. 134
Dorsal (cochlear) root of the VI I I th nerve ................ 134
Ventral (vestibular) root of the VI I I th nerve ............. 138
Trigeminal root fibers to the cerebellum ........................... 142
Secondary afferent connections of the vestibular region ............. 144
Tractus cerebello-vestibularis ................... ............. 144
Certain other secondary afferent connections ................... 145
Secondary efferent connections of the vestibular region ............. 145
Tractus spino-vestibularis .................................... 144
Tractus vestibulo-spinalis .................................... 145
Commissural fibers .......................................... 148
Tractus vestibnlo-cerebellaris ................................. 151
Fiber connections of the rerebellum ................................... 153
Afferent cerebellar connections ................................... 153
Tractus spino-cerebellaris .................................... 153
Nucleo-cerebellar system ..................................... 156
Ventral superficial arcuates .............................. 157
Dlivo- and reticulo-cerebellar fibers ....................... 157
Certain other nucleo-cerebellar connections ................. 158
Tractus tecto-cerebellaris ..................................... 160
Efferent cerebellar paths ......................................... 163
Cerebello-tegmental system ................................... 163
Tractus eerebello-motorius et tegmentalis mesencephali ...... 164
Tractus cerebello-motorius et tegmentalis bulbaris .......... 167
Tractus cerebello-spinalis (et cerebello-vestibularis) ............. 169
Connections of the cerebellar plate ............................ 171
Tractus trigemino-cerebellaris ................................. 159
Summary and discussion ............................................. 171
INTRODUCTION
The aim of this contribution is to present a detailed. com-
parative study of the vestibular and cerebellar nuclei and
fiber paths in various reptilian forms . These two regions are
considered together since they are so intimately associated
structurally. functionally. and in their phylogenetic and onto-
genetic development that such joint consideration seemed to
be warranted . Thus far no investigator has presented. as a
unit. a detailed study of these regions in reptiles. and it was
felt that such a study. in a vertebrate class of such funda-
mental importance to comparative neurology and one which
offers so wide a degree of variation among its members. could
be pursued with profit. and would help bridge a rather obvi-
ous gap in our knowledge of the detailed structure of the
brain stem in these forms .
VESTIBULAR AND CEREBELLAR SYSTEMS 95
This study, pursued by anatomic methods, is subject to the
limitations of all comparative neurological studies carried on
by such means, but, while not all the problems encountered
have been brought to final solution, it is felt that sufficient
new data have been uncovered to warrant the presentation of
this study and to incite the interest of other workers in this
field.
This study carries a step caudalward the work of Huber
and Crosby ('26 and '33) on the reptilian thalamus, mid-
brain, and tectum.
I am indebted to Dr. G. Carl Huber and Dr. Elizabeth C.
Crosby for the opportunity to undertake this problem, for
continued help and supervision while pursuing it, and for
constructive criticism freely given; to them I gratefully
acknowledge my indebtedness.
MATERIALS AND METHODS
The material available for study embraces brain series cut
in cross, sagittal, and frontal planes, and respectively pre-
pared by the modified pyridine-silver (Huber and Guild, '13)
and tigroid methods (Huber, '27) used at Michigan, and by
a modified Weigert's medullary sheath stain. The series
studied are the same as those listed by Huber and Crosby
'33).
GENERAL SURVEY OF THE LITERATURE
Since neither the reptilian cerebellum nor vestibular region
has been studied exhaustively by very many investigators,
so that most of the references quoted will be from papers
which have considered these regions more or less incidentally,
it is scarcely deemed desirable or profitable to present a com-
plete review of the literature. It is rather the intention to
refer to the literature at those points in this study where
such references are directly pertinent to the structures under
consideration.
96 JEAN X. WESTON
I n addition to the literature directly pertaining to the
problem, numerous other researches by various investigators
and based on other vertebrates have been consulted. While
in many cases they have had but an indirect bearing on this
investigation, they have been found both suggestive and
stimulating in the interpretation of the material herein con-
sidered. All of these will obviously not be referred to in the
text of this paper despite the fact that they have undoubtedly
had no small effect upon the evaluation of the material.
Chief among these were the contributions of Larsell and
Herrick on the amphibian cerebellum and associated regions.
CYTOARCHITECTONIC STRUCTURE OF THE VESTIBULAR REGION
General introduction
The discussion of the region in which the vestibular nuclei
are located will be initiated by presenting briefly certain gen-
eral observations based on the most generalized order availa-
ble, the Chelonia or turtles.
Most of the lateral limb of the medulla oblongata in rep-
tiles, which forms a large part of the lateral wall of the I Vth
ventricle, has been generally recognized as being primarily
concerned with the reception, elaboration, and projection of
vestibular and acoustic impulses. Ventrally and anteriorly
the vestibular area extends to the chief sensory nucleus and
entering roots of the Vth nerve. More caudally it is de-
limited ventrally by the descending root and nucleus of the
Vth nerve, its ventral boundary throughout being approxi-
mately at the level of the inferolateral angle of the IVth
ventricle. Caudally it grades over into the sphere of the
vagus, and cephalically and dorsally, though more imper-
ceptibly, into the cerebellum. Dorsally it reaches the cochlear
region. Through this area pass the fibers of the VI I th and
VI I I th nerves, together with other fibers of passage, most of
which are concerned with the cerebellum. The above rela-
tions are illustrated in figures 2 to 7 and 17 to 26.
I n Lacertilia, accompanying the eversion (figs. 8 and 26)
of the cerebellum (possibly as a result of it), with the in-
creased development of the trigeminal complex (fig. 26) and
with the advent of a large nucleus tangentialis, the vestibular
region becomes relatively more broadened mediolaterally and
shortened dorsoventrally (figs. 9, 10, 25 and 26), with a con-
comitant narrowing of the I Vth ventricle ; this together with
a variable degree of development of the cochlear region
(Holmes, '03 ; figs. 9, 10, 25, see p. 118). This latter region is
relatively smaller than in the turtle (at least in the forms
studied), though in position and differentiation (particularly
in the larger lizards) it evidently foreshadows the conditions
found in Crocodilia.
I n Ophidia the shape of the vestibular region resembles
that in lizards, although the ventricle is relatively narrower
(figs. 11, 12 and 25). The cochlear region (Holmes, '03) is
very poorly developed (figs. 12 and 25). I n this connection
it is interesting to note that Ditmars ('10, p. 199) observed
that snakes have no external ear and that their tongue is
sensitive to vibrations, i.e., they hear with their tongue.
It is among Crocodilia that the cochlear region becomes
best developed (Holmes, '03). I n the alligator it assumes
almost equal size with the vestibular region (figs. 14, 15 and
25)-provided one considers the nucleus laminaris to be
cochlear-and the IVth ventricle is further encroached upon
through this development (see especially fig. 15).
General survey and descriptions of vestibular subdivisions as
based ow t he material
The terminology of Sanders ('29) for the sparrow is
largely used in this account, chiefly because that terminology
is more directly applicable to the conditions observed in
the forms studied, both by reason of their morphologic ap-
pearance and their comparative anatomic relationships.
Where her terminology does not fully conform with the locali-
zation in reptiles, this is duly mentioned. It should be stated
that the plans of subdivision of the region advanced earlier
TEE JOURNAL OF COMPARATIV% NEUROLOGY, VOL. 65
98 JEAN K. WESTON
by Stieda (1875), C. L. Herrick (1890), Koppen (1892, 1894),
Holmes ('03), Beccari ('12), van Hoevell ('16), and others
deserve recognition and may be quite as correct and permissi-
ble as that which follows. One must recognize that all plans
of subdivision present individual interpretations of the same
structures, and are biased in part by the material at the dis-
posal of the respective observer.
According to the Sanders' ( '29) nomenclature, the veshibu-
lar region is here subdivided as follows: 1) nucleus ves-
tibularis ventrolateralis (nucleus of Deiters) ; 2) nucleus
vestibularis tangentialis ; 3) nucleus vestibularis ventromedi-
alis ; 4) nucleus vestibularis descendens (inferior of Beccari) ;
5) nucleus vestibularis dorsolateralis (superior of Beccari ;
anterior of van Hoevell; see p. 113) ; 6) nucleus vestibularis
superior ; 7) nucleus laminaris.
This nucleus, very
evident and relatively constant in position throughout this
vertebrate class, forms an excellent approach to the more
detailed consideration of this region. The nucleus vestibu-
l ark ventrolateralis is particularly large and obvious in
Chrysemys, being the most conspicuous cell mass in the ves-
tibular region (figs. 3,4,6,7,19,22 to 24), and it is practically
impossible to confuse its characteristically large, deeply stain-
ing, multipolar cells with those of any other cell mass in this
region.
I n the literature this large celled nucleus is commonly re-
ferred to as Deiters' nucleus or the lateral vestibular nucleus,
either as a whole (Koppen, 1892; Holmes, '03; van Hoevell,
'16 ; Ariens Kappers, '21 and '29 ; Larsell, '26 and '32 ; Shank-
lin, '30; Frederikse, '31; and others) or in part (Beccari, '12,
central group), these observers considering its large celled
part homologous with the mammalian Deiters ' or lateral ves-
tibular nucleus. Hindenach ('31, fig. 5) figured this nucleus
in Sphenodon punctatum, where its relations appear essen-
tially similar to those in the turtle. Deiters' nucleus is but
a part of the nucleus ventralis of the older workers, which
was often referred to as a whole as Deiters' nucleus.
Nucleus vestibularis vefitrolateralis.
The nucleus vestibularis ventrolateralis lies somewhat
medially (figs. 3 and 4), having a rostrocaudal extent a trifle
greater than that of the entering VI I I th roots (figs. 22 and
24). It passes over rostrally into the nucleus vestibularis
dorsolateralis (figs. 6, 22, 23) and continues caudally into the
nucleus vestibularis descendens (figs. 7, 22 to 24; see also
fig, 470 in Ariens Kappers, Huber and Crosby, '36). Medi-
ally and dorsally (figs. 3 and 4) the periventricular gray
intercalated among the dorsal arcuate fibers separates it from
the ependymal lining of the I Vth ventricle; medially and
ventrally the nucleus vestibularis ventromedialis serves the
same purpose. Laterally it lies against the long ascending
tracts to the cerebellum (figs. 19 and 23). Dorsally it borders
upon the cochlear nuclei (figs. 3 and 4) and the nucleus
laminaris, where that is present, and ventrally it is limited
by the nucleus and descending root of the Vth nerve (figs.
4, 5, 19, 23 to 25). Longitudinally coursing fibers of passage
and the root fibers of the VI I th and VI I I th nerves aid in
causing the rather broken up appearance of this nucleus
(figs. 19, 22, 23 to 25).
Beccari ( '12), in Lacerta, and Ariens Kappers ( '21), in the
crocodile, have noted a fairly consistent grouping of these
large cells into a central (which Beccari, '12, p. 678, considers
probably the only one homologous to Deiters' nucleus of
mammals), a dorsal, and a ventrocaudal (Beccari ; posterior
of Ariens Kappers) group (figs. 3 and 4). These three divi-
sions are readily observed in the turtle material, the central
and ventrocaudal being rather intimately connected and to-
gether making up the pars ventralis of this account (figs. 3
and 4) . These are also figured but not labeled by Larsell
( '32, fig. 13, p. 313). I n addition there is found, at some
levels, a fairly well-defined lateral group (see figs. 3 and 4).
The shape of the nucleus is quite inconstant throughout. At
certain levels, where all the groups are more or less inter-
connected, the nucleus might be considered as having the
shape of a rough, broken, shallow C, the thicker closed portion
(the central group of Beccari) lying toward the ventricle,
100 JEAN K. WESTON
and the open part (bounded dorsally by the dorsal group;
ventrally by the ventrocaudal group of Beccari) embracing
most of the entering vestibular portion of the VIIIth roots
and the scattered cells intercalated among these fibers (this
relation shows better in the lizard, fig. 9). Where the lateral
group is present it closes the C laterally. One of the factors
responsible for the quasi-separation between dorsal and ven-
tral groups is the entering sensory root of the VI I th nerve
(fig. 25). The ventral (ventrocaudal and central of Beccari,
'12) group continues somewhat farther caudally than does
the dorsal, and is usually more distinct, better delimited, and
larger than the others. This grouping varies but little among
the turtles, the chief difference being in the relative compact-
ness and size of the cell masses.
The cells characterizing this nucleus are relatively large,
deeply staining, and multipolar in appearance (figs. 3 and 4),
resembling the ventral horn cells of the spinal cord of mam-
mals (Spitzka, 1880). This is the only one of the vestibular
nuclei in turtles presenting a distinctly efferent type of cell
(Malone, ,l o), and hence it is logical to assume that it is
probably principally concerned with vestibular reflex paths
(doubtless largely to the neck and extremities in this form
where the trunk is immovable). These large cells possess
several thick processes (seven having been noted in some
cases) and abundant Nissl substance grouped most heavily
around a large, centrally placed nucleus. Scattered among
these large cells are both medium sized and small multipolar
and spindle-shaped varieties, which stain fairly deeply (figs.
3 and 4), as well as some very small, lightly staining, pear-
shaped and bipolar cells. Certain of these latter cells may
be representative of Beccari's nucleus tangentialis, which is
not distinct in the turtles. The above description agrees
essentially with the accounts of this cell group in the litera-
ture and shows little variation among all the turtles examined.
I n lizards, Beccari ( '12 ; see his figs. 7, 8, 9, 12 and 18) gave
a very complete description of this nucleus, based on silver
series of embryos of Lacerta muralis, and all the lizard ma-
terial examined accords with his description, and is very
similar to that of the turtle (figs. 9 and 10). I n Anolis caro-
linensis the nucleus vestibularis ventrolateralis varies from
the above description chiefly as follows: 1) its large cells
appear actually and relatively smaller in size; 2) the nucleus
as a whole is smaller in size, both actually and relatively;
3) relatively, it does not have so great a rostrocaudal extent;
4) the lateral group noted above is not so constant nor so well
developed ; 5 ) it reaches its greatest development slightly
rostra1 to the most cephalic entrance of the roots of the
VI I I th nerve. It must be remembered that Anolis is a very
small, swift moving, active form, while Chrysemys is much
larger and more clumsy (except possibly in water) ; possibly
the size of this efferent nucleus may be to some extent a func-
tion of the amount of muscle present.
Beccari noted that the ventrocaudal group lies close to the
chief sensory nucleus of the Vth nerve and is difficult to
delimit therefrom. This is not true for the turtle, where the
relatively much greater size of the cells of Deiters nucleus
could permit of no confusion with cells of the chief sensory
nucleus of the Vth nerve. Also, the most anterior level at
which Deiters nucleus is found is caudal, in turtles, to the
most caudal pole of the chief sensory nucleus of V, depending,
of course, on where one stops calling it the chief sensory
nucleus and begins calling it the nucleus of the descending
root of V. Neither could Beccaris statement be verified in
the majority of the lizards observed, though in Varanus there
was a small group of Deiters cells (part of the ventrocaudal
group of Beccari) which seemed to lie rather far ventrally,
quite close to certain cells of the Vth which appeared large
and efferent in type, though quite apparently a part of the
nucleus of the descending root.
Shanklin (30, see figs. 30, 33 and 34) also figured and de-
scribed this nucleus in Chameleon vulgaris as a nucleus
ventralis (nucleus magnocellularis of Deiters) . His brief
description is in perfect accord both with that of Beccari and
with the one presented here for Anolis. Prederikse (31,
figs. 28 and 29) showed this nucleus in photomicrographs of
102 J EAN K. WESTON
the brain of Lacerta vivipara, where its appearance is in
agreement with the above account. Larsell ('26, figs. 4 and
23) figured and labeled this nucleus in Anniella and Gerrho-
notus.
Of all the reptiles studied, snakes seem to have received
least attention from the viewpoint of the comparative ana-
tomy of the vestibular system. Hence the general work of
Rabl Riickhard (1894), who barely touched on the vestibular
region, a figure of Larsell ( '26, fig. 17), and the brief mention
by Papez ('29) of this nucleus, which he labeled in his figure
226, include all the relevant literature.
I n Thamnophis the nucleus vestibularis ventrolateralis is
in most respects very similar to the nucleus of that name as
described for Anolis. This comparison concerns two indi-
viduals in which the brains are of about the same size and
in which the degree of activity is on the same general plane.
The chief differences are that the individual cells of the
nucleus are actually much larger in the snake (fig. 12), in
fact being somewhat larger than the homologous neurons in
Chrysemys, and that the nucleus has relatively fewer cells.
Here, again, there can be little possibility of confusion with
the chief sensory nucleus of V. However, in Natrix, where
the cells are slightly smaller, some large cells are noted in the
chief sensory nucleus and the nucleus of the descending root
of V, which are very close to the nucleus vestibularis ventro-
lateralis pars ventralis (see p. 101) of this account (possibly
being aberrant Deiters' cells), and here Beccari's ( '12) ob-
servation on Lacerta muralis can be confirmed, although his
rather generalized interpretation is open to question, since
this relationship appears more in the nature of an individual
variation. The other snakes studied were very similar to
Thamnophis, except that the cells were slightly smaller and
the distinctness of the grouping, as in turtles, was somewhat
variable.
I n crocodiles the nucleus under consideration has been dis-
cussed, in the majority of cases in a very general way, by
many observers (Rabl Riickhard, 1878 ; Holmes, '03 ; van
Hoevell, '16; de Lange, '16; Ariens Kappers, '21 and '29;
Huber and Crosby, '26 ; and others). For Alligator mississip-
piensis, a large and, on land, a rather lethargic animal, the
brains of young specimens were studied, but even so they
were larger than the brains of adult individuals of the other
reptilian forms available. I n them the nucleus vestibularis
ventrolateralis (fig. 14) is very well developed, resembling
more closely the turtle as to the size of its cells, although
these cells are relatively somewhat larger. While the group-
ing described by Ariens Kapper.s ( '21) for the crocodile
is present in the alligator, it is often masked by the distinct
concentration of these cells in the lateral portion of this
region (fig. 14), apparently a rather distinct development of
the variable lateral group noted in the turtle.
The nucleus presents its most marked development anterior
to the level at which the nucleus tangentialis is well developed
and it does not extend so far caudally in relation to the most
caudal entering roots of the VI I I th nerve as is the case in
the turtle. The concentration of the smaller elements of this
nucleus is much more pronounced than in the other forms
studied (fig. 14).
The relations of the ventrolateral vestibular nucleus (fig.
14) are essentially the same as in the turtle, except that
dorsally it is in contact with the pronounced band of longi-
tudinally coursing fibers which lies ventral to the nucleus
laminaris (fig. 14), and that laterally and caudally a well-
developed nucleus tangentialis is present. Here, also, there
is no question of delimitation from either of the sensory
nuclei of the Vth nerve,
Nucleus vestibularis tangentialis. The literature seems to
contain no reference to the presence of a nucleus vestibularis
tangentialis in the turtle other than Beccari's ('12, p. 691)
general statement that it is probably present throughout the
vertebrate series. Chrysemys certainly presents no well-
developed nucleus tangentialis, its only representative con-
sisting of a few small, spindle-shaped cells lying intercalated
in the course of the entering vestibular fibers (figs. 4 and 19),
104 JEAN I(. WESTON
unless certain of the medium-sized cells of the nucleus ves-
tibularis ventrolateralis can be considered as its functional
equivalent. With the exception of one series of Emys, which
exhibited a more marked concentration of cells in the typical
position of the nucleus tangentialis, the other turtle material
observed resembled Chrysemys in this respect.
The relatively distinct nucleus tangent iali s distinguishes
the vestibular region of lizards from that of turtles, and it
was in Lacerta muralis that Beccari ( '12) first described this
cell group in reptiles. Shanklin's ( '30) paper on Chameleon
vulgaris does not consider this nucleus nor do his figures
show it labeled. The photomicrographs of Frederikse ('31,
figs. 28 and 29), based on Lacerta vivipara, show this nucleus,
although it is not labeled.
Beccari's excellent and most detailed description of this
nucleus, based on silver preparations of embryos, leaves little
to be added; the following brief description is chiefly for the
convenience of the reader, and is based on Anolis carolinensis.
The nucleus tangentialis is the well-marked collection of
small and medium-sized cells found intercalated among the
entering vestibular root fibers (figs. 9 and 10). Anteriorly
it passes over gradually into the lateral part of the nucleus
vestibularis dorsolateralis, although at times this relation
may be obscured by an especially well-developed lateral ex-
tension of the anterior portion of the nucleus vestibularis
ventrolateralis. Caudally it gradually merges, along with
the small celled component of Deiters' nucleus, into the
nucleus vestibularis descendens ; this caudal to the most
caudally entering vestibular roots. Medially it borders on
the nucleus vestibularis ventrolateralis (in many cases being
practically continuous with the small celled portion of this
latter mass, figs. 9 and l o), and more caudally, where that
mass is not well developed, it approaches the dorsal arcuate
fibers and the periventricular gray (figs. 9 and 10). Ven-
trally it comes close to the chief sensory nucleus and the
nucleus of the descending root of V, more laterally, and the
nucleus vestibularis ventromedialis, more medially (fig. 10).
It is often difficult to delimit the nucleus tangentialis from
the chief sensory nucleus and the nucleus of the descending
root of TI, since their cells are of practically the same size
and lie quite close together in certain planes, although the
motor root of the VI I th nerve (fig. l o), the ventralmost enter-
ing vestibular fibers (fig. 9), and a fairly constant blood
vessel separating these areas (as Koppen, 1892, noted in the
turtle and which is frequently encountered among all the
reptiles ; see fig. 25D) are very helpful in orientation. Later-
ally the tangential nucleus is scattered among the entering
vestibular roots and the longitudinally coursing tracts ;
dorsally, it reaches the cochlear gray (figs. 9 and 10).
Beccari (12) has noted four groupings; he designated them
as anterior, posterior, central, and lateral, with the added
note that cross sections show more a dorsal and a ventral
division, due to the manner of entrance of the vestibular
fibers. All these groups can be identified to a variable degree
at various levels, and this distribution can probably be
ascribed to fiber paths ramifying through this region; hence
such grouping is of doubtful functional significance.
The cells of this nucleus are of small and medium size, and
stain rather deeply. They are much smaller than the large
Deiters cells (fig. 9), and, although certain of them resemble
the cells of all the other contiguous gray areas, they usually
differ enough em masse to be readily recognizable. Many of
them, especially the larger, more deeply staining cells, are
multangular, though at first glance they appear round; others
are more nearly oval (racket-shaped, as Beccari termed
them) ; still others are spindle-shaped and appear very simi-
lar to the smaller elements of the nucleus vestibularis ventro-
lateralis. The nucleus appears best developed at the more
caudal levels of the entering vestibular roots (fig. l o), and
at these levels there are usually few of the large Deiters
cells in the plane of section (fig. 10).
Only minor variations were noted among the lizards availa-
ble for study. Varanus and the Gila monster present little
evidence of grouping and the latter appeared to exhibit a
106 JEAN X. WESTON
relative diminution in the size of this cell mass, with its cells
apparently more scattered among the cells of the nucleus
vestibularis ventrolateralis.
No description or labeled figure of this nucleus has been
encountered anywhere in the literature on snakes, it appar-
ently having been included under the general term of the
vestibular or Deiters nucleus. However, in the garter snake
the nucleus tangentialis is very similar to Anolis (fig. 12) in
all respects, although the grouping of its cells is more irregu-
lar and inconstant than in lizards, and its boundaries are not
very distinct. I n the other snake material studied (possibly
because less satisfactorily stained), the nucleus does not ex-
hibit quite so marked a development.
Ariens Kappers (21) appears to have been the only ob-
server who has mentioned a nucleus tangentialis for Croco-
dilia, and he does not consider it very fully. It is quite well
developed in Alligator mississippiensis, although not to
nearly the same relative extent as has been noted in lizards
or snakes. Its relations are very similar to those described
in Anolis, except that its dorsal boundary is formed by the
cross cut fibers ventral to the nucleus laminaris (fig. 15), and
it has no distinct and constant subgrouping of its cells. The
cells of the nucleus vary chiefly in that here the more deeply
staining elements are spindle- or lancehead-shaped and very
numerous, being relatively and actually smaller than the more
oval or pear-shaped, deeply staining cells of the lizard (com-
pare figs. 10 and 15) and snake.
Nucleus vestibularis ventromedialis. This rather ill-defined
mass has evidently not been named in the literature consulted
dealing with reptilian forms, probably being included under
the general term of Deiters nucleus or the nucleus ventralis.
However, in the sparrow Sanders (29) designated a mass
of small cells by the name nucleus vestibularis dorsomedi-
ah, which certainly possesses similar relations to the
nucleus under consideration, although in reptiles the term
ventromedialis is more nearly accurate, topographically con-
sidered, and this mass is referred to here as the nucleus
vestibularis ventromedialis (figs. 3 to 5 ) . Sanders ('29) also
suggested that this mass is homologous with the medial ves-
tibular nucleus of mammals.
I n Chrysemys this nucleus is preeminently a small celled
mass, intermingled with which are a variable number of
medium sized, multangular cells (figs. 3 to 5 ) , which stain
rather lightly and are pear- or spindle-shaped. However,
they usually can be readily differentiated from contiguous
cell masses by reason of their relatively much smaller size,
although the whole nucleus is often almost totally eclipsed
at some levels by the invasion of the large cells of the nucleus
vestibularis ventrolateralis.
The rostrocaudal limits of this nucleus are very indefinite
and purely arbitrary. Here its cephalic limit is considered
as being at the level of the cephalic end of the nucleus ves-
tibularis ventrolateralis, since in this region it presents first
a fairly definite concentration of cells, although it is quite
possible to trace such cells somewhat farther rostralward
than this, more or less as periventricular gray scattered
rather thinly among the dorsal arcuate and cerebello-teg-
mental fibers. Followed caudally it becomes better developed,
tending to assume a rather triangular outline in cross sec-
tion (fig. 5 ) , keeping the same relative position just lateral
and inferior to the inferolateral angle of the I Vth ventricle,
and can be followed through the whole extent of the vestibu-
lar region caudal to the above noted point of origin. It is
split up considerably by the dorsal arcuates, the vestibular
fibers, and the secondary Vth fibers coursing through it, which
factors probably partially explain its variable cell density
from section to section. Dorsolaterally it is bounded by the
ventrolateral vestibular nucleus and more caudally by the
inferior vestibular nucleus, certain of the cells of this latter
nucleus practically becoming a part of the nucleus vestibu-
laris ventromedialis (fig. 5 ) . Ventrolaterally it lies fairly
close to the nucleus of the descending root of the Vth almost
throughout its extent (figs. 3 to 5 ) , and at some levels there
appears to be cell continuity between them. Dorsally and
108 JEAN K. WESTON
rnedially it stretches toward the cochlear gray by means of
scattered cells intercalated in the course of the dorsal arcuate
fibers (periventricular gray, figs. 3 to 5 ) . Ventrally and
medially it is similarly elongated around the inf erolateral
angle of the ventricle toward the medial longitudinal fascicu-
lus by small, scattered cells; medially it is limited by the
ependyma of the IVth ventricle (figs. 3 to 5 ) . I ts caudal
boundary is difficult to fix. However, somewhere between
the entrance of the I Xth and Xth nerves it becomes dominated
by these nerves, especially the latter, and quite definitely can
no longer be thought of as being in any relation to vestibular
fibers. This point coincides essentially with the caudal pole
of the acoustic nuclei and of the nucleus vestibularis de-
scendens, all of which are difficult to delimit caudally with
any degree of certainty. The extent to which this cell mass
is in reality vestibular is questionable (see p. 175). Excepting
in some very young specimens of Pseudoemys, where this
nucleus is much better developed as a compact mass than in
the adult forms, the nucleus vestibularis ventromedialis shows
little variation in the turtle.
The relations of the nucleus vestibularis ventromedialis in
lizards differ but little from those in Chrysemys, except that
in Anolis the nucleus appears to be slightly better developed
(figs. 9 and 10). More anteriorly it comes into relation ven-
trally with the dorsomedial motor nucleus and root of the
. Vth nerve. More caudally and laterally it is in contact with
the nucleus tangentialis, which, as observed above, is here
much more developed than in the turtle (fig. 10). Also, due
to the broadening and flattening of this region, as previously
noted, it lies quite near to the cell masses on the dorsolateral
side of the medial longitudinal fasciculus (figs. 9 and l o),
thus resembling in position the nucleus intercalatus of more
caudal levels. The larger lizards, such as Varanus and Gila
monster, resemble the turtle more closely in the development
of this nucleus, while the smaller lizards resemble Anolis
more closely.
I n snakes the relations of the nucleus vestibularis ventro-
medialis are similar to those in Anolis. However, the nucleus
here is relatively much larger (fig. l a), though whether this
may be due to a considerable invasion of the smaller elements
of surrounding cell masses is uncertain, particularly since
the lateral boundary of the nucleus under consideration is a
very indefinite and hence an arbitrary one. Dorsally and
medially it is well developed, and seemingly becomes directly
continuous with the cells of the cochlear region, which are
very similar to the cells of this nucleus in size and type (fig.
12). Laterally, in addition to the boundaries noted in the
turtle and the lizard, the nucleus is limited at some levels by
a rather conspicuous bundle of what appear to be secondary
acoustic fibers, which swing down into the region of the
nucleus of the descending root of the Vth.
I n Crocodilia is found the best relative development of
the nucleus vestibularis ventromedialis among the reptiles,
although there is not much difference in its relative develop-
ment when compared with that of the garter snake (compare
fig. 12 with figs. 14 and 15). With the exceptions 1) that the
dorsal extension of the nucleus into the cochlear region is not
present here, due to the well-developed nucleus laminaris
(figs. 14 and 15), 2) that the cells of this nucleus appear
slightly larger, and 3) that more of the multangular type of
cells are present, the description for the snake can be applied
to Alligator mississippiensis.
Nucl eus vest i bul ari s descendem. The nucleus vestibularis
descendens has usually been included with Deiters nucleus
or nucleus ventralis. Ariens Kappers (21) stated that in
reptilian forms there is no caudal extension of Deiters nu-
cleus, as in mammals, to the level of entrance of the I Xth
nerve.
I n general, the nucleus vestibularis descendens may be
thought of as the more diffuse, caudal, medium sized and
small celled continuation of the nucleus vestibularis ventro-
lateralis and the nucleus tangentialis (figs. 5, 7, 22 to 24).
It represents a fairly well-delimited nucleus of variable shape
110 JEAN K. WESTON
extending from the caudal limits of Deiters ' nucleus, cephali-
cally, to a rather indistinct, purely arbitrary, caudal limit,
which lies somewhere between the entrance of the I Xth and
of the Xth nerves, its most caudal part apparently running
over into the nuclei funiculi posterioris. It lies in close rela-
tion throughout to the descending vestibular root (figs. 22
to 24). Ventrolaterally it is limited by the descending root
of the Vth and its nucleus; ventromedially, by the nucleus
vestibularis ventromedialis. Medially it is bounded by the
dorsal arcuate fibers; dorsally, by the cochlear gray; and
laterally, by the ascending fibers to the cerebellum. The
delimitation from the cochlear gray and from the nucleus
vestibularis ventromedialis is not very distinct at some levels,
there being cell continuity between all three of these masses.
The constituent cells are essentially of the medium sized
and small varieties previously noted for the surrounding
nuclei, the medium sized, multangular variety predominating,
particularly more rostrally. Throughout its extent the mass
appears most dense around and among the descending ves-
tibular fibers.
Beccari ('12) noted that in Lacerta muralis the nucleus
vestibularis descendens is very indistinct. Frederikse ( '31,
figs. 30 and 5 5 ) labeled this nucleus in Lacerta vivipara,
where it seems to be quite comparable to its homologue in
the turtle.
I n the smaller lizards examined, such as Anolis and Hol-
brookia, the above statement of Beccari seems to apply, and
this may be ascribed, at least in part, to the very great de-
velopment caudally of the nucleus tangentialis which appar-
ently occurs at the expense of the nucleus vestibularis
descendens (and as a consequence of the development of the
nucleus tangentialis, the nucleus under consideration no
longer exhibits the same intimate relationship to the nucleus
vestibularis ventrolateralis as was noted in the turtle).
Also, the fact that there appear to be many fiber bundles
coursing through this region and converging upon the coch-
lear area, which tend to split up the gray of the whole region
into columns, serves to effectually mask the nucleus vestibu-
laris descendens in these forms, the more so since the cells
of the nucleus vestibularis descendens are very similar to
those of the surrounding cell masses. However, in the larger
lizards, notably in Gila monster, this nucleus resembles the
homologous cell mass in Chrysemys, with the exception of
the above noted relation to the nucleus tangentialis.
Papez ( '29) figured the nucleus vestibularis descendens and
mentioned it very briefly in his consideration of the brain of
Natrix sipedon. From his figures it appears that in position,
relations, and extent, it resembles closely the foregoing
description for Chrysemys, although a few of his figures
suggest that he includes with it a portion of what here is
ascribed to the cochlear nuclear complex. However, since
his illustrations are based on fiber material it is diflicult to
be certain of a sharp delimitation of a nuclear mass such as
this, the more so since such delimitation is practically im-
possible even in toluidin blue material. I n the garter snake,
as well as in the other species studied, the relations of this
nucleus are so similar in all respects to the conditions in the
larger lizards, that no further description is deemed neces-
sary.
I n Alligator mississippiensis the relations of the nucleus
vestibularis descendens are very similar to those in the turtle.
However, in the alligator the nuclear mass is a trifle more
condensed and much better delimited from the cochlear re-
gion, dorsally, the nucleus vestibularis ventromedialis, medi-
ally, and the nucleus of the descending root of the Vth
ventrally, and appears to fade out at about the level of
entrance of the Xth roots. It is notable that a greater num-
ber of larger cells are present at the cephalic pole of this
nucleus, while caudally it is almost entirely small celled.
Nucleus vestibularis dorsolateralis. As with the other ves-
tibular nuclei, Deiters ' nucleus (or nucleus ventralis) has
also included the nucleus vestibularis dorsolateralis of this
report. The nucleus under consideration is the same as the
nucleus vestibularis superior of Beccari ( '12, Lacerta), and
112 J EAN K. WESTON
undoubtedly forms a large part of van Hoevell's ('16) nucleus
vestibularis anterior in Caiman sklerops (pp. 113 and 115).
Ariens Kappers ('21) verified this nucleus in the crocodile,
and considered it homologous to Bechterew's nucleus of
higher forms. It is the same cell mass which Larsell ('26,
figs. 21 and 23) and Hindenach ( '31, Sphenodon) have labeled
as the nucleus vestibularis superior.
Between the levels of the nucleus vestibularis ventrolater-
ah, caudally, and the cerebellar nuclei, cephalically, practi-
cally the whole of the lateral limb of the medulla oblongata
is taken up by the nucleus vestibularis dorsolateralis (figs.
3, 6, 22 and 23). These boundaries are purely arbitrary,
since caudally there is little difference between the cells of
the caudal part of this nucleus and the smaller cells of the
nucleus vestibularis ventrolateralis, and rostrally the same
is true of this nucleus in relation to the cerebellar and supe-
rior vestibular nuclei. I n the turtle, the most cephalic limit
of the nucleus vestibularis dorsolateralis (likewise the most
caudal limit of the cerebellar nuclei) is approximately at the
rostralmost level of entrance of the Vth roots (figs. 6 and 23).
The mass is very irregular in shape and rather diffuse
throughout, appearing most dense caudally and medioven-
trally (fig. 3). Medially it lies against the cerebello-tegmental
fibers, anteriorly, and the dorsal arcuates and their interca-
lated periventricular gray, more posteriorly. Dorsally the
cochlear gray (and the nucleus laminaris), wliich is best de-
veloped more caudally, serves as a rather indefinite boundary,
and rather more rostrally the cerebello-oblongata junction,
with its attendant fiber masses sweeping up into the cere-
bellum, answers the same purpose. Laterally the long ascend-
ing tracts to the cerebellum limit the nucleus (fig. 22) , and
ventrally it is usually very poorly delimited from the chief
sensory nucleus of the Vth (figs. 3 and 6).
The cells of this nucleus are predominantly medium-sized,
multangular, and stain fairly darkly. However, many smaller
cells of various shapes are scattered among them, particularly
caudally. As van Hoevell ('16) has noted for the nucleus
vestibularis anterior (see p. 112) of Caiman (incorrectly desig-
nated Alligator sklerops ; private communication from Am-
sterdam), the medium sized multangular type predominates
caudally resembling the vestibular variety, while more
rostrally there are more of the medium sized spindle-shaped
or pear-shaped cells, which resemble those of the cerebellar
nuclei, although the extent to which this tendency is mani-
fested depends upon where one draws the line between the
vestibular and the cerebellar nuclei (see p. 115). I n other
words, here is a rather undifferentiated transition area be-
tween the strictly vestibular and the strictly cerebellar
regions. The diffuse nature of the nucleus is largely due to
the fact that it is broken up by the vestibular fibers passing
through it, and these are also very likely responsible for the
inconstant clumping observed at all levels. With the excep-
tion of Emys, where this mass appears more compact through-
out, there is little evidence of variation in any of the other
turtles studied.
I n the lizards the nucleus vestibularis dorsolateralis is
formed, as in turtles, by the fusion caudally of the cerebellar
nuclei, which occurs slightly anterior to the caudalmost part
of the cerebello-oblongata junction-and hence slightly ante-
rior to the rostralmost entrance of the Vth roots. The nucleus
is very diffuse here and occupies a greater dorsoventral extent
than in the turtle. I ts rostralmost level is usually relatively
farther cephalad than the same in the turtle. However, in
Varanus it is placed much farther caudalward than in the
turtle, appearing at about the middle of the rostrocaudal
extent of the entering Vth roots. To some extent this varia-
tion is undoubtedly due to a difference in the plane of section,
but it is unquestionable that such variation is in part de-
pendent upon the degree of eversion in these forms. Caudally
it is arbitrarily considered as bounded by the entrance into
the picture of Deiters cells, which, being farther forward
than in turtles, make it appear that in all the lizards (except
Gila monster) there has been a relative, if not an actual
shortening of this portion of the vestibular area.
THE JOWNAL OF COMPARATIVE NEUROLOGY, VOL. 65
114 JEAN K. WESTON
There is, rather consistently, an indefinite grouping, especi-
ally more caudally, into a medioventral group and a dorso-
lateral group. The smaller lizards, such as Anolis, presented
a denser cell arrangement, the cells appearing relatively more
numerous than in the turtle, while in the larger lizards, such
as Gila monster, they are more exactly comparable to those
of the turtle. The sizes and types of cells are relatively very
similar to those of the turtle, though actually the cells ap-
peared slightly smaller.
The nucleus vestibularis dorsolateralis represents the nu-
cleus which Beccari ( '12) and Larsell ('26) in lizards, and
Hindenach ( '31), have labeled the nucleus vestibularis supe-
rior. Beccari's brief description of this nucleus, as a mass
of stellate cells smaller than those of Deiters' and lying in
that portion of the brain stem corresponding to the juxta-
restiform body of mammals, along the course of the ascending
VI I I th root, is more true of the lateral than of the medial
portion of this mass. Beccari also questioned whether this
mass corresponded to the superior vestibular nucleus of
Cajal ('00) in Amphibia, and stated that if its neuraxes do
not add to the medial longitudinal fasciculus, by reason of
its position it could be considered to correspond to the nucleus
vestibulo-cerebellaris. Frederikse ( '31), in his figures 26 and
27, has labeled this nucleus more caudally and in its ventral
portion as nucleus vestibularis adscendens, and more ante-
riorly, apparently, as a part of his nucleus vestibulo-cere-
bellosus ; his nucleus principalis vestibularis (fig. 27) is
probably a part of the chief sensory nucleus of V.
Papez ( '29) mentioned the nucleus vestibularis dorsolater-
alis under the term nucleus vestibularis superior, and consid-
ered it as the place of origin for his vestibulo-mesencephalic
tract, and this apparently constitutes the only reference to
this mass in the literature on snakes. The description of this
nucleus in the lizard is applicable, for the most part, to
snakes. The chief differences are that in the gartei snake
the cells of the nucleus seem actually larger than in the lizard
and more densely arranged (fig. 11). There is no differen-
tiable cochlear gray bounding this nucleus dorsally, the
cerebello-oblongata junction occupying that position through-
out the cephalocaudal extent of the nucleus.
The nucleus vestibularis dorsolateralis has been figured
and described by van Hoevell ('16) for Caiman sklerops
(pp. 112 and 125) under the name of the anterior vestibular
nucleus, and his descriptions are applicable to the conditions
observed in Alligator mississippiensis, where, although some-
what larger, this nucleus resembles closely the condition in
Chrysemys. However, according to van Hoevell 's figures,
the name of anterior vestibular nucleus is applied to a large
part of what is here considered the nucleus lateralis cerebelli.
Nucleus vestibularis superior. Among reptiles, only turtles
present a distinct nucleus vestibularis superior which has a
position similar to that mass as described by Sanders ('29)
for the sparrow. Even here, sagittal series suggest that this
nucleus might well be considered the direct rostroventral con-
tinuation of the nucleus vestibularis dorsolateralis (fig. 6).
Where present this mass corresponds only to the most ante-
rior portion of the nucleus vestibularis superior of Beccari
( W), Ariens Kappers ( 'Zl ), Larsell ('26 and '32), and
others, their nucleus vestibularis superior being largely the
nucleus vestibularis dorsolateralis of this account.
I n transverse series of the turtle, the nucleus vestibularis
superior (figs. 2,6,17) is continuous caudally with the nucleus
vestibularis dorsolateralis. It first becomes clearly evident
slightly rostral to the most rostral limits of the chief sensory
nucleus of V, just ventral to the nucleus lateralis cerebelli
and but poorly delimited from this latter mass, being com-
posed of morphologically similar cells. Ventral to it is the
reticular gray, from which it is also poorly delimited. Medi-
ally lie the cerebello-tegmental fibers and their intercalated
periventricular gray. Laterally the afferent cerebellar tracts
are found. Followed forward it becomes larger, becoming
somewhat round in cross section. Its relations remain nearly
the same throughout, except that anteriorly it appears more
directly continuous with the reticular gray, ventrally, and
with the lateral cerebellar nucleus, dorsally (fig. 6), while
116 JEAN K. WESTON
medially it is in relation with the mesencephalic root of the
Vth nerve and the scattered cells accompanying that tract
(which are Iabeled the nucleus loci cerulei by van Hoevell,
'16, in a similar relation to the mesencephalic root of the Vth
nerve in Caiman sklerops). It finally passes over cephali-
cally, caudal to the large celled portion of the nucleus isthmi,
into a smaller-celled reticular gray mass (see p. 128), rather
broken up by fiber tracts running to the cerebellum, which
swings around the lateral angle of the IVth ventricle (fig. l),
and is here in direct continuity with the most anterior part
of the cerebellar nuclei.
The cells of the superior vestibular nucleus resemble in
size both those of the lateral cerebellar nucleus and those of
medium size found in the nucleus vestibularis dorsolateralis,
although the multangular variety predominates, resembling,
in this respect, the vestibular rather than the cerebellar nuclei
(fig. 2). Smaller cells of various shapes and sizes are also
present. The cells are rather compactly arranged and the
whole mass stains relatively deeply. It is not a very large
nucleus (fig. 2).
While in certain of the other forms studied (fig. 8) it is
possible to make out faintly a nucleus similar to the above,
its delimitation is so indefinite that it would force the issue
were the attempt made to locate such a nucleus in all the
different orders, I n lizards (especially the smaller ones)
and snakes, the small size of the area, and, especially in
lizards, the marked eversion or tendency toward it, probably
Constitute a sufficient anatomic basis to explain the absence
of this nucleus as a distinct mass. I n alligators tlicre is also
a tendency toward eversion in that portion of the cerebellum
most closely attached to the medulla oblongata probably
sufficient to explain in part the absence of this nucleus.
Probably a functionally homologous group is present in all
these forms, although morphologically it may constitute a
part of either the nucleus vestibularis dorsolateralis, the
nucleus cerebelli lateralis, or even the dorsalmost portion of
the reticular gray of this region, or it may be distributed
among these three masses, being submerged until the whole
region reaches fuller phylogenetic development in birds.
Nucleus lamilzaris. There still is not unanimity of opinion
as to whether the nucleus laminaris is to be considered purely
vestibular, purely cochlear, or mixed in its function. Holmes
('03) first considered this nuclear mass under the above name
(although it had been known previously by the older workers
such as Edinger and Koppen as part of the tuberculum acusti-
cum) and described it rather completely in the alligator,
noting that it was variously developed in all the other Reptilia
examined by him. This nuclear mass was seen in Caiman
(van Hoevell, 'I S), in Crocodilia (de Lange, '16; Ariens
Kappers, '21 and 'as), in Alligator (Huber and Crosby, '26),
and in Chameleon (Shanklin, '30).
At about the middle of the rostrocaudal extent of the
nucleus vestibularis ventrolateralis (figs. 4, 5, 7, 21, 24),
Chrysemys series show a nucleus dorsal to this latter mass,
which is unquestionably the nucleus laminaris of Holmes
('03; Testudo and Chelone midas) although there is no defi-
nite lamina. Followed frontally it becomes more and more
diffuse, and before the most anterior level of the nucleus
vestibularis ventrolateralis is reached, it can no longer be
delimited from the cochlear gray of this region as a definite
nuclear mass. Caudally, considerably behind the nucleus
vestibularis ventrolateralis, the nucleus laminaris seems
gradually to become invaded by the larger oval cells (figs.
5, 7, 21, 24) of the nucleus dorsalis magnocellularis (pars
posterior of the nucleus laminaris of Holmes, '03). Ventrally
the nucleus is separated from the vestibular nuclei by a fairly
well-marked mass of cross cut and longitudinally cut fibers.
Medially it abuts against the ependyma of the I Vth ventricle,
and laterally and dorsally it lies among the longitudinally
coursing fibers to the cerebellum.
The characteristic cells (figs. 5, 7, 24) are medium sized
and are rather spindle-shaped or multangular in form,
although there is some admixture of smaller cells. For the
most part, these cells appear to be oriented so that one of
118 JEAN K. WESTON
their two main processes courses dorsally and the other
ventrally.
Beccari ( 'l a), in Lacerta muralis, was unable to establish
a nucleus laminaris comparable to Holmes ' description,
though he would not affirm that it was lacking. Shanklin
('30), in Chameleon, identified this nucleus as that portion
of the cochlear mass having the same relations as the nucleus
laminaris of crocodiles, and noted that the cells are medium
sized and small, multipolar or round in form, and further
stated that they are not lined up in a row but are separated
into groups by means of fiber fasciculi. It is difficult to har-
monize Shanklin 's description of this mass with the labeling
of the same in certain of his figures. Frederikse ('31) did
not attempt to subdivide the cochlear region in Lacerta
vivipara.
I n smaller lizards, the present observations are in accord
with the negative results of Beccari as concerns this nucleus.
Certainly there is cochlear gray throughout this region (figs.
9 and l o), but this resembles the nucleus laminaris only in
position-and even that tends to become atypical at many
levels, where this gray moves more laterally and seems to
have a position more analogous to that of the avian nucleus
angularis. It is apparently the more caudal part of this
cochlear gray that Shanklin described as the nucleus lamin-
aris (and so it may be functionally). No very definite nucleus
dorsalis magnocellularis can be clearly recognized in these
forms, the cells throughout appearing smaller and more dif-
fnse than those composing the nucleus laminaris in turtles.
When the larger lizards are examined, the results of
Holmes ('03) can be verified, since a nuclear mass similar to
the nucleus laminaris of Chrysemys can be readily located,
although it is apparently not so well developed. It may be
noted, parenthetically, that in this order, where the nucleus
laminaris shows the greatest variability of development, the
development of the cochlea also shows its greatest variability,
and this would constitute a major argument for considering
the nucleus laminaris to be chiefly cochlear in function.
Holmes ( '03) mentioned having examined Boa, Python,
Tropidonotus, and Zamenis, but made no comment concerning
the presence of the nucleus laminaris in these forms. It is
not possible to delimit a definite nucleus laminaris in the
snakes studied, which is not surprising since the whole
acoustic region is relatively greatly reduced. I f a functional
nucleus laminaris is present, it has either migrated medio-
ventrally or has become a part of the dorsal cochlear mass,
the only cochlear nuclear mass sufficiently developed here to
deserve consideration, which by reason of its topographic
relations is more homologous to the nucleus angularis of
higher forms (fig. 12).
It is in Crocodilia, concomitant with the extraordinarily
great development of the cochlear portion of the ear and the
consequent relative enlargement of the acoustic portion of
the VI I I th nerve, that there is an unmistakable and highly
developed nucleus laminaris, another argument for its being
primarily cochlear in function. I n these forms, the descrip-
tions of Holmes ('03) for Alligator, of van Hoevell ('16) for
Caiman, and of Ariens Kappers ('21) for Crocodilia are
quite complete and the material available is in essential agree-
ment with them. A glance at figures 14, 15 and 25A shows
the relations of this very evident cell mass. However, just
dorsal to the fiber mass bounding the nucleus laminaris
dorsally, there is found a second nuclear mass having the
same linear type of arrangement but being somewhat more
broken (figs. 14, 15, 25A), the cells of which are similar to
those of the nucleus laminaris, but less darkly stained. Later-
ally this mass is directly continuous with the nucleus lamin-
aris by means of a fairly well-defined, semilunar band of
cells, the concavity of this band being directed medially (figs.
14 and 15). This appears not to have been noted by Holmes
or subsequent observers of this region in reptiles. This band
is best shown caudally (fig. 15), and it is also observed to
continue ventrally, apparently in direct continuity with the
smaller cells of the vestibular region. This apparent con-
nection is definitely intercalated among secondary cochlear
120 J EAN I<. WESTON
fibers. This second, more dorsal row of cells probably repre-
sents portions of the nucleus laminaris split off by the longi-
tudinally coursing tracts which bound it dorsally.
The nucleus laminaris passes over caudally, as Holmes
('03) has noted, into a large celled nucleus (the nucleus
dorsalis magnocellularis of van Hoevell, '16, Ariens Kappers,
'21, and Huber and Crosby, '26 ; pars posterior of the nucleus
laminaris of Holmes, '03), which here seems to be composed
of a larger celled mediodorsal portion and a smaller celled
ventrolateral portion, although these respective portions may
possibly be accentuated by the plane of section and the man-
ner in which the fibers pass through this mass, and hence
have relatively little functional significance.
CYTOARCHITECTONIC STRUCTURE OF THE CEREBELLUM
Gemera1 imtroductiom
The reptilian cerebellum may be considered as a plate of
nervous tissue, attached laterally to the medulla oblongata
and rostrally to the anterior medullary velum, which diverges
from this attachment in a variety of ways and which shows
a variable amount of folding, depending upon the species
under consideration. With such attachments it follows that
the cerebellar plate forms a variable portion of the roof of
the IVth ventricle, and serves in all cases as a place of attach-
ment for the tela choroidea of the I Vth ventricle.
Embryologically, Herrick and Liarsell have shown for am-
phibians (although this had been known earlier for man; see
His, 1891, and Streeter, '12; the work of Reese, '08 and '15,
on the alligator is also of interest) that the cerebellum
"appears to be derived from two bilaterally paired sets of
primitive coordinating centers, namely the auricular lobes
and the corpora cerebelli" (Larsell, '32, p. 340) ; and Larsell
( '32) considers it to have a similar origin in reptiles.
So far as the gross morphology is concerned, it will receive
little attention here, since it is fully realized that this phase
of the reptilian cerebellum has received adequate attention
from other investigators (for example, Ingvar, '19, and Lar-
sell, '26 and '32).
VESTIBULAR AND CEREBELLAR SYSTEMS 12s
Description. of cerebellar nuclei
Edinger ('00) and Banchi ( '03) considered briefly a rep-
tilian cerebellar nucleus. Since the work of van Hoevell
( '16), this mass has been generally recognized as consisting
of two groups of cells, usually referred to as the nucleus
lateralis cerebelli and the nucleus medialis cerebelli. How-
ever, it should be added, and with emphasis, that these two
nuclei are very intimately connected morphologically in all
the forms studied, probably being parts of the same slightly
folded gray mass. I n Sphenodon punctatum, Hindenach
('31) has noted no anatomic separation into two nuclear
masses, but found that such a separation may be made on
the basis of the fiber connections.
Nucleus lateralis cerebelli. The nucleus lateralis cerebelli
has received brief consideration in turtles from Larsell ('32)
where it appears to agree with the same mass as described
by van Hoevell ( '16) in Caiman sklerops (see pp. 115 and
125), by Ariens Kappers ( '21 and '29) in Crocodilia, by Lar-
sell ('26) in lizards and snakes, and by Shanklin ('30) in
Chameleon vulgaris (with some exceptions, see p. 123).
I n Chrysemys marginata (and with but slight differences
in the other turtles studied) the nucleus lateralis cerebelli
(figs. 1, 2, 6, 17, 23, 24) is continuous caudally with the nucleus
vestibularis dorsolateralis, and ventrally with the nucleus
vestibularis superior. I Dis difficult to separate the nucleus
lateralis cerebelli from the nucleus vestibularis dorsolateralis
but for convenience the caudal pole of each of the cerebellar
nuclei is considered here to be at the rostralmost level of
entrance of the Vth nerve. At this level (figs. 2 and 6) the
nucleus lateralis cerebelli is a rather dense cell group lying
about in the middle of the lateral limb of the medulla oblon-
gata, just lateral to the IVth ventricle (fig. 2). Ventrally
lies the nucleus vestibularis superior (as yet poorly devel-
oped) and dorsally, the nucleus medialis cerebelli. Laterally
it is limited by fiber bundles to the cerebellum. Followed
rostralward it keeps practically the same relative position
except that it is crowded more dorsalward. At about the
122 JEAN I<. WESTON
middle of its caudorostral extent, it attains a position dorsal
to the lateral angle of the I Vth ventricle (Larsells, 32,
lateral extension of the rostral region of the ventricle),
where ventrally a reticular gray mass takes the place of the
nucleus vestibularis superior and where medially it appears
continuous with the periventricular gray above the I Vth ven-
tricle (van Hoevell s nucleus brachium conjunctivum ; con-
sidered here merely as a well-developed portion of the
periventricular gray), as well as with those swinging around
the lateral angle of the ventricle and running ventromedially
(fig. 1). Here, also, it is in more evident continuity with the
nucleus medialis cerebelli, although this connection is varia-
bly developed throughout its extent, as is its continuity with
the superior vestibular nucleus and the reticular gray ven-
trally. There is doubt as to the relations of this nucleus as
figured by Larsell ( 32, fig. 15). His figure does not present
relations observed in Chrysemys or other turtle material, but
does appear to represent the relations observed in snake
material, leaving the supposition that it was inadvertently
accredited to turtle when it should have been labeled snake
(compare with figs. I, 2, 11and 12 of this account).
Farther forward this nucleus has become more medial,
lying entirely dorsal to the I Vth ventricle, closely intercon-
nected with the nucleus medialis cerebelli (fig. l), and form-
ing practically the most medial portion of the reticular gray
mass which swings up around the lateral angle of the I Vth
ventricle. At this level the nucleus lateralis cerebelli has
reached its most rostral extent. Thus the lateral cerebellar
nucleus is more distinct from the medial one laterally and
caudally, and more intimately connected with it medially and
rostrally (figs. 1, 2, 6).
The cells are chiefly medium sized and stain fairly dark,
being quite densely packed throughout (figs. 1, 2, 6). Many
small cells are also noticeable. Apparently oval and bipolar
types predominate, although certain multangular cells can be
observed, as well as a few scattered, large cells of the type
characteristic of the nucleus medialis cerebelli.
Larsell ( '26) described a nucleus lateralis cerebelli rather
generally in the legless, subterranean, burrowing form, Anni-
ella nigra. He observed that there the lateral nucleus is
considerably larger than the medial, noting that the nuclei
are both small and that the separation between them is not
so evident as in the larger reptilian forms. I n Gerrhonotus
principis (a less active, more generalized lizard), he found
the cerebellar nuclei very similar to those of Anniella but
larger, and his figure of this form (fig. 21) appeared to illus-
trate conditions quite similar to those found in Chrysemys
at more rostra1 levels of this nucleus. This apparent simi-
larity, in part at least, is probably due to the fact that in
Gerrhonotus there is a much less extreme eversion than is
found in Anolis, Holbrookia, and certain other lizards.
Shanklin ( '30) considered this nucleus in Chameleon vul-
garis. From his brief description of the cerebellar nuclei,
it appears that they are completely at variance with those
masses as Larsell ('26) described them in lizards and as they
are considered and identified in this paper, and such of his
figures as illustrate these masses do not seem to materially
clarify them, at least insofar as it is permitted to interpret
such figures. It seems most probable that his nucleus later-
alis is the nucleus medialis of this report, or both the cere-
bellar nuclei in some of the figures, and that his nucleus
medialis may be the periventricular cells corresponding to
the nucleus of the brachium conjunctivum of van Hoevell
( '16).
I n the smaller lizards the available material accords with
Larsell's observations. Here it is very difficult to delimit a
well-defined nucleus lateralis cerebelli (fig. 8) because of the
relatively much smaller size of these lizards and because
the eversion seems to have tended to pull the cerebellar nuclei
dorsalward. As a result the lateral cerebellar nucleus be-
comes very diffusely organized and less clearly delimited
from the medial cerebellar nucleus and the granular layer
of the cerebellum. However, its main relations are very
similar to those in Chrysemys, particularly anteriorly. Ven-
trally, there being here no observable superior vestibular
124 JEAN K. WESTON
nucleus, the lateral cerebellar nucleus is directly continuous
with the reticular gray throughout its course. The cells are
similar to those in the homologous nucleus of the turtle,
although they seem to be somewhat smaller (fig. 8).
I n the larger lizards the nucleus is more nearly comparable
to those found in the turtle and the alligator; Gila monster
being more comparable to the turtle and Varanus to the
alligator.
The nucleus lateralis cerebelli appears relatively smaller
than the medial cerebellar nucleus in the smaller lizards,
which fact does not agree with the condition noted by Larsell
('26) ; in the larger specimens, there is little differencc in the
size of these two nuclei.
Larsell ('26) has noted that the cerebellar nuclei in Tham-
nophis are similar to those in Gerrhonotus, and his figure 18
(of an embryo of Thamnophis) illustrates in general the
conditions encountered among the various snakes studied.
I n these forms there is relatively much less of a separation
of the nucleus lateralis cerebelli from the nucleus medialis
cerebelli, medially, and from the nucleus vestibnlaris clorso-
lateralis, laterally and ventrally, than in any of the reptiles
studied, indicating, as others have suggested, that perhaps
here there is a regressive, rather than a specialized, condition
of the cerebellum. The relations of this nucleus are most
similar to those found in the smaller lizards, although here
it is more truly lateral in relation to the medial nucleus than
in any of the lizards examined (fig. 11) and appears to be
the larger of the two, as Larsell ('26) noted (see fig. 11).
I n Alligator mississippiensis the nucleus lateralis cerebelli
has relations very similar to those found in turtles (fig. 13).
It differs, of course, in its somewhat larger size, and also by
extending farther rostralward than does the medial nucleus.
Van Hoevell ('16) described this nucleus very briefly in
Caiman sklerops, considering it as lying 'lateral and ventro-
lateral' from the cells of the medial cerebellar nucleus. With
this description our observations are in full accord. How-
ever, it is difficult to homologize the figures of van Hoevell
('16) with similar levels of the brains of either the turtle or
the alligator. With his figure 9 the available material is in
agreement, so far as the cerebellar nuclei are concerned.
With his figure 8 this is not the case. The cell mass he has
labeled nucleus vestibularis anterior probably is the caudal
part of the lateral cerebellar nucleus, and the cells he has
labeled the nucleus lateralis cerebelli are merely a part of
the smaller celled component of the medial cerebellar nucleus,
which component at most levels is in direct (though often
masked) continuity with the same sort of cells of the lateral
cerebellar nucleus. The same criticism applies to his figure
7, although here it appears that quite possibly the ventral-
most portion of what is labeled the nucleus vestibularis
anterior is actually that mass due to the tilt of the section;
however, the same criticism as above holds for his nucleus
lateralis cerebelli and the more dorsal part of his nucleus
vestibularis anterior. His figure 6 is in agreement with this
account insofar as the nucleus vestibularis anterior (here
nucleus vestibularis dorsolateralis) is concerned. This fig-
ure appears to be about at the level of the transition from
the cerebellar nuclei to the vestibular nuclei (which occurs
relatively farther rostralward than in the turtle). Since the
nucleus lateralis cerebelli merges caudally more directly with
the nucleus vestibularis dorsolateralis (or van Hoevell 's
nucleus vestibularis anterior), it would appear consistent to
consider that what van Hoevell has labeled in his figure 6 as
the nucleus lateralis cerebelli, is the diffuse, caudal pole of
the nucleus medialis cerebelli, at least in its more dorsal
portion. Ariens Kappers ('21) identified the two cerebellar
nuclei in the crocodile and used van Hoevell's figure 7 to
illustrate them.
Nucleus medialis cerebelli. The literature on the medial
cerebellar nucleus seems to be in quite general agreement as
to its location and relations, and has been contributed by the
same workers as those referred to in the account of the lateral
cerebellar nucleus.
126 J EAN K. WESTON
The nucleus medialis cerebelli as described by Larsell ( 32)
is strictly comparable with the homologous mass in the ma-
terial available for study, with the reservations previously
noted (p. 122) regarding his figure 15. He said (p. 317) :
The more rostra1 part of the medial nucleus in Chrysemys
extends nearly to the midplane, but farther caudad it lies
more laterally, in transverse series. It has, therefore, an
oblique position in the deep part of the cerebellum. He
also noted (p. 316) that: Toluidin-blue series of Chrysemys
show the two nuclei as quite distinct dorsally, but ventrally
they are joined by a bridge of cells. While this is essen-
tially true at the more anterior levels of their extent (and is
well shown in his figure 15), it is scarcely the case more
caudally, at least in the series of Chrysemys available for
study (see p. 122; also figs. 1, 2, 6; and the account of this
mass in the snake, p. 127).
As has been seen (p. 112), the caudalmost level of the
medial cerebellar nucleus in Chrysemys is considered to be
in the same plane as the caudal pole of the lateral nucleus,
both passing over into the nucleus vestibularis dorsolateralis.
However, the medial nucleus lies in a considerably more
dorsal position than the lateral cerebellar nucleus and is not
so compact a mass at these more caudal levels. Throughout
its rostrocaudal extent it is most definitely a cerebellar
nucleus (figs. 1, a) , being at all times, except perhaps most
caudally, an integral part of the cerebellum. Despite the
fact that it lies in the stream of afferent fibers to the cere-
bellum throughout its extent, it appears to be a fairly com-
pact mass of cells. Ventrally it is bounded by the nucleus
lateralis cerebelli ; lateraIly, by the long ascending paths to
the cerebellum; medially, by the IVth ventricle and the granu-
lar layer of the cerebellum; and dorsally, by the granular
layer of the cerebellum, and these relations remain about the
same throughout its extent (figs. 1, 2, 6, 7, 16, 17, 23, 24). I ts
rostralmost extent slightly exceeds that of the lateral cere-
bellar nucleus.
I ts characteristic cells are relatively large and fairly dark
staining, being oval or bipolar in type (figs. 1, 2, 6, 7),
although many of them appear to have at least three proc-
esses. These cells are very nearly the same size as the
Purkinje cells in this form, although the largest of them seem
just a shade larger than most of the Purkinje cells (this
apparent difference in size has been considerably enhanced
in the turtle illustrations so that these would show up well).
It is these larger cells that provide the chief criterion for
dividing a medial from a lateral nucleus in cell preparations.
There is a large component of smaller cells, which appear
very similar to the neurons making up the lateral cerebellar
nucleus and which may be functionally equivalent to these
latter.
Larsell ( '26) has described a nucleus medialis cerebelli in
Anniella and Gerrhonotus, and noted that it is smaller than
the lateral cerebellar nucleus and but poorly separated from
it ; in other respects, it resembles the turtle condition. Shank-
lin ('30) has labeled and described the nucleus medialis
cerebelli of this report as a nucleus lateralis cerebelli.
The medial cerebellar nucleus, in both the smaller and
larger lizards, has essentially the same relations noted for
the turtle (fig. 8). However, the nucleus is not so compact
as in the turtle, its cells being scattered among the granular
cells of the cerebellum laterally (Shanklin, '30). The size
of its characteristic cells relative to the Purkinje cells agrees
with the condition in the turtle, although both types are actu-
ally smaller here than in the turtle. This mass seems to
differ from the homologous group in Gerrhonotus and Anni-
ella (Larsell, '26) in being fully as large as, or even larger
than, the lateral cerebellar nucleus.
I n snakes, the medial cerebellar nucleus agrees perfectly
with Larsell's ('26) description for the garter snake, and
resembles that of the smaller lizards quite closely, as he noted.
I n Thamnophis this nucleus is relatively much smaller than
the lateral cerebellar nucleus and very intimately associated
with it (fig. 11). The large cells, slightly larger than the
128 J EAN K. WESTON
Purkinje cells as a rule, are found intermingled with the
granular cells of the cerebellum, being very few and scattered
in most sections.
I n Alligator mississippiensis this nucleus (fig. 13) agrees
closely with the nucleus medialis cerebelli of Caiman sklerops
(van Hoevell, '16), with the exceptions noted on page 125, as
well as with Ariens Kappers' ( '21) brief description based
on Crocodilia. I t also resembles the homologous nucleus in
Chrysemys, except that it does not extend relatively so far
rostralward, and in the lizards to the extent that the large
cells are rather diffusely scattered among the basal cells of
the granular layer, lying between the fascicles entering the
cerebellum, and the typical cells are much larger than the
Purkinje cells.
Lateral reticular gray. This term embraces the small-
celled, mostly rather diffuse gray masses in close contact
with the cerebellar nuclei laterally and ventrally in all the
forms discussed, as is particularly well shown in sagittal and
in horizontal series (figs. 1, 8, 11, 13). The more caudal
and lateral portion of the reticular gray (just; rostra1 to the
most cephalic extent of the vestibular complex) lies in direct
continuity with the nucleus lateralis cerebclli, and follows it
dorsally and medially as it creeps around the lateral angle
of the I Vth ventricle (fig. 1). This more caudal portion of the
reticular gray seems to be intercalated among the fiber bundles
swinging out of (possibly also into) the cerebellum (figs. 6
and 23, not labeled).
More anteriorly the small-celled portion of the nucleus
isthmi lies in approximately the same relation to the anterior
part of the medial cerebellar nucleus, being more split up by
the afferent fibers to the cerebellum and seemingly continuous
with the smaller cells of the cerebellar nuclei and with the
granular cells of the cerebellum. Farther anteriorly (at the
very anterior limits of the cerebellar plate), the most medial
extension of the nucleus isthmi pars parvocellularis seems
to be continuous forward into the mid-brain as the cell mass
labeled nucleus precerebellaris (fig. 6).
Thus it is possible to think of the cerebellar nuclei as the
apex of a fan-like mass of smaller reticular cells, the rays of
which, concave medially, diverge rostrally into the mid-brain
and caudally into the medulla oblongata, apparently being
threaded by both afferent and efferent cerebellar fibers.
DescriptioN of cerebellar corteix;
It is fully realized that this report has little to contribute
concerning the structure of the cerebellar cortex among rep-
tiles, although the scope of the study demands at least a brief
consideration of it. For a detailed account of its embryo-
logical development, the significance of its various suloi, and
its subdivisions, the reader is referred to the excellent studies
of Ingvar ( '19) and Larsell ( '26 and '32), and, more in gen-
eral, to the accounts of Tilney ('23), Herrick ('24), and
Hausman ('29). For details of the histology of the separate
elements, the work of Ram6n y Cajal ('11) and especially of
Larsell ( '32; in this connection his papers on the amphibian
cerebellum are also interesting) is referred to, the more so
since this latter observer had a very rich collection of Golgi
material at his disposal.
The cerebellum of reptiles is essentially a plate of nervous
tissue. This plate, as has been readily recognized in all rep-
tilian forms studied, and by practically every investigator
who has even glanced at this region, is composed of three
rather definite layers, which in turn have a distinct intrinsic
histologic character.
I n Chrysemys (figs. 1 to 7), where the relations are perhaps
as simple as they are found anywhere among the reptiles,
there is a more superficial, lighter staining area, with diffuse,
faintly stained, round cells, the molecular layer (see all fig-
ures on the turtle). This occupies almost half of the thick-
ness of the plate except laterally, in the auricular lobe
(Larsell 's, '32, floccular lobe), where it is considerably
thinned out relative to the granular layer (fig. 3). Trans-
verse sections, particularly rostrally, seem to make it appear
relatively thicker.
THE JOURNAL OF OOMPARATTVE NEUROLOGY, VOL. 65
130 JEAN K. WESTON
Approximately the deeper half of the plate is taken up by
an exceptionally deeply staining, closely packed mass of
small, round or oval cells, the granular layer (see all figures
of turtle). This also remains about the same thickness
throughout the plate, but laterally, as was the case with the
molecular layer, it thins slightly (fig. 3). From the cerebello-
oblongata junction forward it is not quite so compact ven-
trally, the cells being split apart by the fiber bundles entering
(and leaving) the cerebellum. It also appears to be hollowed
out ventrally and caudally, near the base of the cerebellum,
to make room for the cerebellar nuclei and the cerebellar
commissure. Here, as in the molecular layer, more rostra1
transverse sections appear to demonstrate a relatively enor-
mously thick granular layer, but this is due to the necessarily
oblique plane such sections have followed. I n the midline
there is a rather constant, fairly cell-free area, quite variable
in distinctness and often occupied by a blood vessel.
Lying between these two layers, and taking up most of the
remaining small portion of the cerebellar plate, is the layer
of Purkinje cells (see all figures on the turtle). These are
fairly large, bipolar, deeply staining, and very obvious cells,
seldom more than three or four cells deep and usually not
more than two cells deep (although transverse series appear
to dispute this for the reasons noted above). Between this
layer and the granular cells is a rather thin, relatively cell-
free area, which is readily seen, in silver or Weigert material,
to be occupied by a tangled fiber mass (figs. 17,18, 23, 24, 26).
The layer of Purkinje cells also thins out laterally (in Lar-
sell's, '32, floccular lobe), being reduced there to a single
layer or scattered single cells (fig. 3).
The above relations hold throughout the reptilian forms
examined, provided necessary allowances are made for the
changes in shape and complexity of the simple plate as de-
scribed above in Chrysemys. For instance, in lizards (see
figs. 8 and 26A) the varying degree of eversion causes the
granular layer to appear most superficial and the molecular
layer deeper, thus apparently causing a reversal of these
relations as described in the turtle. The relations in the
snake (figs. 11, 12, 26C) are very similar to those in Chrys-
emys, except that there is but a small auricular lobe and no
layer of Purkinje cells excepting caudally, these cells being
scattered irregularly throughout the molecular layer (figs.
11, 12, 26C). I n Alligator (see figs. 13, 14, 15, 26B) the cere-
bellar plate is folded into three parts: an anterior vertical
part, attached to the anterior medullary velum and showing
a tendency toward eversion; a more horizontal middle part;
and a more caudal vertical part. I n these forms some
troublesome transverse pictures are encountered, enhanced
somewhat by the so-called ventriculus cerebelli, merely a
dorsal extension of the IVth ventricle, which in some trans-
verse planes seems to be a true cerebellar ventricle. I n Alli-
gator apparently the granular layer occupies slightly more
than the half of the plate allotted to it in Chrysemys.
FI BER CONNECTIONS OF THE VESTIBULAR REGION
Direct root fibers of t he acoustico-facial complex;
J ust caudal and dorsal to the entrance of the Vth nerve,
a comparatively large bundle of nerve fibers can be readily
observed entering the medulla oblongata in gross prepara-
tions of Chrysemys marginata. This bundle can be divided
quite easily, microscopically, into four roots, two of which
belong to the VI I th nerve. Here, this whole mass of entering
nerve roots (figs. 19 and 25) is designated as the entering
acoustico-f acial complex, and although this communication is
not concerned with all of this complex, all portions of it will
receive brief consideration.
Sensory root of t he VIIt h fierue. This consists of a bundle
of fine fibers which runs almost directly medialward to near
the ependyma of the I Vth ventricle (fig. 22), having a slight
inclination from dorsal to ventral, and being split into fasci-
cles by the longitudinally coursing tracts through which it
passes nearly at right angles (fig. 25). This necessarily takes
it between the cells of the nucleus vestibularis ventrolateralis,
which it helps to divide into a pars dorsalis and a pars ven-
tralis (fig. 4). Upon approaching the I Vth ventricle, the
132 J EAN K. WESTON
sensory VI I th root turns abruptly caudalward (fig. 22),
occupying the same relative position as does the fasciculus
solitarius (fig. ZO) , with which it becomes directly continuous
caudally (as Ariens Kappers, '21, noted in Crocodilia, form-
ing thus a pre-vagal component of that fasciculus; and as
Tuge, '32, likewise showed in Chrysemys elegans; see also
Barnard, '36). No collaterals or stem fibers could be traced
to any of the nuclei at its level of entrance.
A$ the point where it turns caudalward, a few fine fibers
course rostralward (fig. 22). These are rather difficult to
identify in any but the most favorable preparations. The
greater number appear to be contributed from more caudal
levels, possibly coming from the I Xth and Xth nerves (fig.
22). Those from the sensory VI I th root are even more diffi-
cult to identify; nor could it be determined whether these
represent stem fibers or collaterals. It is most difficult to
trace this bundle rostrally, due to its intermingling with the
cerebello-tegmental system of fibers near the ventricle (fig.
22), which are of the same caliber and staining reaction.
They could be identified no farther rostralward than the point
where the mesencephalic root of the Vth nerve reaches the
dorsolateral angle of the ventricle, even in certain especially
favorable horizontal preparations. (These are doubtless
homologous to Beccari's, '12, VI I th fibers to the cerebellum.
More probably they pass to the much disputed secondary
gustatory nucleus, if such a nucleus be present here.)
No significant differences appear in the several lizards
studied (fig. 25B), which agrees with the results of Reccari
( '12), Ariens Kappers ( '2l ), Shanklin ( '30), and Frederikse
( '31). I n the snake, this root does not lie nearly so close to
the ventricle as in the turtle, where it causes an elevation on
the ventricular wall, but lies lateral to the nucleus vestibularis
ventromedialis (fig. 25C). Furthermore, it enters the medulla
oblongata more dorsally, due to the relatively poor develop-
ment of the cochlear root in the snake. Papez ('29) figured
such a bundle, labeling it the fasciculus solitarius. The alli-
gator resembles the turtle except in the position which this
root occupies with reference t~ the ventricle, wherein it re-
sembles the snake more closely, although here the sensory
V'IIth root is not so easy to separate from the other longi-
tudinally running systems with which it is crowded closely
together, probably due to the extreme development of the
cochlear nuclei in crocodiles. Ariens Kappers ( '21) has noted
its presence in the crocodile and Hindenach ('31) figured and
labeled a comparable root in Sphenodon.
Mot or root of the VIIt h nerve. The coarse, heavily medul-
lated and fairly compact fascicles observed at the ventral and
anterior level of entrance of the ventral VI I I th root make up
the motor root of the VI I th nerve (figs. 4, 19, 23 to 25). From
this point they arch dorsomedially toward (and finally into)
the commissural fibers lying just beneath the floor of the
I Vth ventricle, being split up somewhat by the longitudinally
coursing bundles of the descending Vth root (figs. 19 and
25D). The dorsalmost fibers lie very near to the ventralmost
vestibular fibers, to which they are very similar. The fasci-
cles course rostromedially through the commissural fibers at
a sufficient angle to be fairly readily separable, until the
dorsolateral portion of the fasciculus longitudinalis medialis
is reached, when they turn abruptly caudalward (being appar-
ently a part of the medial longitudinal fasciculus in trans-
verse sections; figs. 19, 20, 25D), and descend in that relation
until about the level of entrance of the I Xth nerve (fig. 20).
Here they leave this position, a few fibers at a time, and
course ventrolaterally to their nucleus of origin. Some of
the fibers of this root cross to the opposite side before turn-
ing caudally, presumably originating from the opposite
nucleus, particularly in lizards.
This root was very similar in the other reptilian forms
studied (see figs. 10, 12, 14, 15, 25), its extent depending
somewhat upon the slightly variable position of the motor
nucleus of the VI I th nerve. Tuge ('32) figured this root in
Chrysemys essentially as noted above; as did Hindenach
( '31) for Sphenodon, Beccari ( '12), Ariens Kappers ( '21,
also in the snake and the crocodile), and Frederikse ( '31)
134 J EAN X. WESTON
for lizards; and Papez ('29) for the water snake. However,
Xhanklin ( ' 30), in Chameleon vulgaris, noted the motor root
of the VI I th nerve as being in close relation with the motor
root of the Vth nerve at the latter's emergence; in fact, lying
dorsal to the motor root of the Vth here (see his fig. 26).
None of the lizard material examined agrees with Shanklin's
figure nor with his description of this root. However, such
a condition may be present in Chameleon, exaggerated per-
haps, by the plane of section.
The VI I I th nerve enters the medulla as
two fairly distinct roots; a dorsal root, which obviously is
related chiefly to the more dorsal (cochlear) group of nuclei,
and a ventral root, which is apparently entirely related to
the vestibular nuclei. I n the turtle both these roots are com-
posed of rather coarse fibers with essentially the same
staining reaction. Practically all investigators who have
examined the VTIIth nerve among Reptilia, since the time of
Stieda (1875), have noted these two roots.
The dorsal,
posterior, or cochlear root, as it has been variously termed,
is of primary interest here only insofar as it carries a vestibu-
lar component (Retzius, 1881-1884 and de Burlet, '29). It
enters the medulla oblongata dorsal and caudal to the sensory
root of the VI I th nerve (figs. 19, 21, 23, 24, 25D).
Various observers (Holmes, '03 ; Beccari, '12 ; Ariens
Kappers, '21; and others) have recognized that a considera-
ble component of this root exhibits vestibular connections.
Beccari ('12) termed this the posterior vestibular root, which
led to some confusion when used in conjuction with the more
inclusive term of posterior root. This vestibular component
will be considered in detail along with the ventral or purely
vestibular root.
The distribution of the purely cochlear portion of the
dorsal root has been well described by Holmes ( ' 03) , Beccari
( '12), Ariens Kappers ( W) , Huber and Crosby ( '26), and
others, most of these observers agreeing that the direct root
fibers are distributed chiefly to the cochlear nuclei (nucleus
Acoustic nevue.
Dorsal (cochlear) root of the VI I I th nerve.
angularis, nucleus laminaris, and nucleus dorsalis magno-
cellularis). Holmes ( '03) found secondary cochlear fibers
joining his acoustico-cerebellar tract and was unwilling to
definitely exclude direct cochlear fibers from this tract,
although he was likewise unable to demonstrate their pres-
ence in his Weigert material. Ariens Kappers ('21) noted,
and later Huber and Crosby ('26) verified, secondary coch-
lear fibers to the cerebellum, which coursed in their cochleo-
vestibulo-cerebellar tract, which have been fully confirmed
(Larsell, '26 and '32 ; Shanklin, '30 ; Frederikse, '31 ; and
others). Beccari ( '12) has apparently stressed more the
direct cochlear fibers to the cerebellum, which he described
and figured quite fully.
Transverse series of the turtle showed the cochlear root
entering the medulla oblongata and passing almost directly
to all the cochlear nuclei present, but chiefly to the nucleus
dorsalis magnocellularis, since the other cochlear nuclei are
poorly developed (figs. 21 and 24). Near these nuclei, the
direct cochlear fibers become so intermingled with secondary
fibers that it is practically impossible to determine definitely
their further course. I n adult series cut in other planes,
particularly horizontal, it appears that certain fibers course
rostrally for a considerable distance, intermingled with the
secondary cochlear and vestibular fibers running in the same
direction, although even here one can do no more than did
Holmes ( '03), when he recorded that he was unwilling to
state definitely that no direct cochlear fibers joined the com-
plex which he labeled the tractus acustico-cerebellaris. How-
ever, in series of very young Chelhydra serpentina ( 3 days
after hatching) cut, with intent, in an oblique frontal plane,
and very well and differentially impregnated after the
pyridine-silver method (Huber and Guild, '13), the whole
sweep of the entering cochlear fibers was very evident (fig.
21). Here unquestionable cochlear root fibers could be traced
so far into the base of the cerebellum at the eerebello-oblon-
gata junction that it was practically certain that they ended
there. This confirms, at least in some measure, the observa-
tion of Beccari ('12) in this respect. Such fibers were not
136 JEAN K. WESTON
very numerous, being far outnumbered by the secondary
cochlear fibers accompanying them. All the other material
examined, particularly in sagittal and horizontal series, ex-
hibited some evidence of such root fibers to the cerebellum,
but nowhere else could they be so conclusively demonstrated
as in the young Chelhydra material. They appeared least
obvious in the alligator, and were difficult to find in the snake
material, probably due to the difference in form relations in
the region. It should be emphasized that it is far from
proven that all the fibers entering with the cochlear portion
of the dorsal rooh in these forms are in reality cochlear in
function. This seems questionable as well from the rela-
tively small development of the cochlear portion of the middle
ear (except among the Crocodilia) as from the debatable part
the sense of hearing actually does play in the life of these
animals. It is apparently true that practically all reptiles
are capable of uttering some sort of sound, at least during
the mating season (see Gadow, '01; Ditmars, '10; Reese,
'15). Consequently, they must possess a certain degree of
hearing. However, it seems quite as likely that a considera-
ble number of the so-called cochlear fibers may be neither
truly cochlear nor truly vestibular at this stage of phylogeny
(or ontogeny, in young forms), but mediate, rather, more of
an undifferentiated vibratory type of sensibility which may
be useful in orienting the animal in space, perhaps making
up in some measure for the loss of the lateral line organs.
An objection that could well be raised here is the fact that
no direct cochlear fibers have been reported among the Am-
phibia, either embryo or adult. However, this does not
entirely deny their existence.
It is likewise entirely possible that some of them may be
purely vestibular in function. While the presence of such
fibers in the cochlear root in the numbers observed passing
directly to the cerebellum does not appear very probable,
still, if true, such fibers could be readily explained as direct
vestibular fibers to the cerebellum running with the cochlear
root.
It would seem significant that this system comprises so few
fibers as to be most apparent only in very young forms where
the nervous elements appear not yet to have reached their
full development. It is further of interest to note that such
fibers become less obvious in the Crocodilia, in which, among
reptiles, are found the best developed cochlea and the largest
relative number of cochlear fibers, and where, consequently,
one should expect to find such a tract best developed were it
an important functional component of the cochlear system.
Probably these facts and the differences of opinion cited
regarding the distribution of the direct cochlear fibers to the
cerebellum can best be coordinated by considering that primi-
tively (as evidenced by their presence in very young material)
such direct fibers are present and rather numerous (particu-
larly in forms where the cochlear nuclei show relatively little
differentiation), probably originating from undifferentiated
endings having to do with orientation in space; from such
endings one would expect direct fibers to the cerebellum. As
these endings differentiate either into receptors for purely
cochlear or purely vestibular stimuli, their fibers entering
the central nervous system would seek centrally either a coch-
lear or a vestibular center (another application of Ariens
Rappers ', '08, principle of neurobiotaxis) .
Thus as the cochlea (and consequently the cochlear nuclei)
becomes progressively more differentiated, and the cochlear
root larger, such direct cochlear fibers to the cerebellum be-
come more and more interrupted by the increasingly large
development of the cochlear nuclei, so that such fibers become
fewer in number, both ontogenetically and phylogenetically,
looking toward the condition in birds and even in mammals,
where they have not yet been reported.
Among the smaller lizards (as well as the snake) the coch-
lear fibers are finer in caliber than the vestibular fibers, as
Beccari ('12) noted, although in Varanus and Heloderma
they appear to be of the same size. I n lizards this root has
been figured and briefly described by Shanklin ( '30), Fred-
erikse ('31), and others. Papez ('29) noted it in the water
138 J EAN K. WESTON
snake, where it resembles the turtle, provided due allowance
is made for the much reduced condition of the cochlear nuclei
and their differences in position (figs. 12 and 25C). This
reduction in size also serves to obscure the direct cochlear
fibers to the cerebellum, since they become mixed with the
vestibular and secondary cochlear fibers almost at once after
their entrance into the medulla oblongata. Holmes ( '03),
Ariens Kappers ('al ), and Huber and Crosby ( '26) have
considered this root in the crocodiles, and are in essential
agreement with the foregoing description. I n these forms,
also, the cochlear fibers are slightly h e r in caliber than the
vestibular fibers (see fig. 25A).
This, the
remaining component of the entering acoustico-facial com-
plex, has been variously referred to as the anterior, ventral,
or vestibular root of the VI I I th nerve. It carries fibers ex-
clusively from that part of the auditory apparatus concerned
primarily with equilibrium (Retzius, 1881-1884 ; de Burlet,
'29). The vestibular component of the cochlear root (the
posterior vestibular root of Beccari, 12) will also be con-
sidered here.
The neuraxes of the vestibular ganglion cells form a rather
closely packed root which enters the medulla oblongata be-
tween the two root of the VI I th nerve (figs. 23 and 25D),
and slightly ventral to the cochlear root. After their en-
trance, these fibers run almost directly medialward, with a
slight ventral tilt. The vestibular component of the cochlear
root pursues a similar course, though this lies dorsal to the
level of the sensory VI I th root (figs. 23, 25D). Transverse
silver series of Chrysemys readily demonstrate the following
possibilities of distribution :
1. Many of the fibers come into relation with both the
large and the small cells of the nucleus vestibularis ventro-
lateralis (fig. 25D).
2. A considerable number pass into the commissural band
of fibers lying ventral to the IVth ventricle (fig. 25D). Here
it is next to impossible to determine accurately what pro-
portion are secondary and what proportion are primary
Ventral (vestibular) root of the VI I I th nerve.
vestibular fibers, though probably the former predominate
(p. 149). However, such direct vestibular fibers may : a) cross
the midline and pass to the opposite vestibular nuclei (these
appear to have lost their medullary sheaths before they reach
the inferolateral angle of the ventricle); b) turn into the
fasciculus longitudinalis medialis, either contra- or homo-
laterally and pass either caudally or cephalically within that
bundle; c) after reaching the inferolateral angle of the ven-
tricle, turn ventrally, in considerable numbers, into the homo-
lateral reticular gray.
3. Sagittal and horizontal series demonstrate conclusively
that many of the entering vestibular fibers run medially to
near the I Vth ventricle and then arch rostrolaterally and
slightly dorsally, a relation noted many years ago by Stieda
(1875). These constitute the direct vestibular fibers to the
cerebellum (fig. 22).
4. Many fibers curve caudolaterally and slightly ventrally,
making up the descending vestibular root, which forms part
of the vestibulo-spinal tract (fig. 22).
5. A considerable number course directly to a position near
the I Vth ventricle, where they bifurcate, one branch ascend-
ing and the other descending in company with the ascending
and descending vestibular roots (fig. 22). This is not always
the case, since either one or both branches may terminate on
the cells of the vestibular nuclei, or one branch may be added
to the commissural fibers noted above. The impression was
received that the ascending and descending vestibular roots
account for a relatively very large number of the direct ves-
tibular fibers, at least in the turtle (see fig. 22), where the
nucleus tangentialis is very small, although some of its cells
showed the cup or spoon type of ending noted first by Ram6n
y Cajal ( '08) in fishes (see also Pearson, '36), and later by
Beccari ('12) in Lacerta muralis.
6. It was also possible to demonstrate a few spoon or cup
type endings about the smaller cells of the nucleus vestibu-
lark ventrolateralis, which supports the idea that (although
very small) a functionally homologous nucleus tangentialis
140 J EAN K. WESTON
may be present here in the turtle. I n the other reptiles, it
was very evident that the large fibers of the vestibular root,
especially, formed spoon- or cup-shaped terminations about
the cells of the1 nucleus tangentialis, particularly in lizards.
Some of these fibers end completely on these cells; others
continue in part toward the commissural fibers, which they
help to form; still others pass rostrally to add to the ascend-
ing vestibular root, or caudally to add to the descending root.
7. A few of the vestibular fibers, or their collaterals, quite
certainly break up around the cells of the nucleus vestibularis
ventromedialis, although these are very difficult to be certain
of due to the many diverse fibers coursing through this
nucleus.
8. Fibers from the vestibular component of the cochlear
root appear to break up in relation to the cochlear nuclei,
particularly the nucleus laminaris, although these may be
aberrant cochlear fibers.
The direct ascending vestibular fibers arch rostrolaterally
and with somewhat of a dorsal inclination (fig. 22). This
course brings them out around (and under) the lateral recess
of the IVth ventricle, then upward and apparently into rela-
tion with the granular cells of the auricular lobe of the
cerebellum (Larsell's floccular lobe, '32). This would seem
to correspond to the brief description of the vestibulo-
floccular tract of Larsell ( '26 and particularly '32). This
tract appears to carry the major portion of the direct ascend-
ing vestibular fibers (fig. 22). Not all of the direct vestibular
fibers follow this course, as Larsell ('32) has noted (see
p. 152). Some secondary fibers accompany this tract. These
direct fibers to the cerebellum are in relation laterally with
the long ascending cerebellar tracts ; in relation dorsally
(mixed to some extent) with the cochleo-vestibulo-cerebellar
tract (Ariens Kappers, '21; Huber and Crosby, '26) ; and in
relation medially with the tractus cerebello-spinalis and the
periventricular gray. Thus they pass dorsolateral to the
nucleus vestibularis superior, though quite possibly they are
in relation with that nucleus via collaterals. These fibers
are in continuous relation with the cells of the nucleus ves-
tibularis dorsolateralis, excepting farther rostrally, where
that nucleus passes over into the cerebellar nuclei. Here
there is a strong possibility of synaptic relation with the
cerebellar nuclei, particularly the medial nucleus, although
not demonstrable in the material available. Some of these
fibers appear to pass into the cerebellar commissure (p. 151).
The descending vestibular root curves in a caudolateral
direction with a slight inclination ventrally, eventually com-
ing to lie rather far laterally, and in relation ventrally with
the descending Vth root (fig. 22). It is closely applied to the
nucleus vestibularis descendens throughout the extent of the
latter (fig. 22), and is relatively a fairly large group of fibers
among the turtles. These fibers could not be followed very
far caudally since it appears that most of them terminate in
relation to the caudal portion of the nucleus vestibularis
ventrolateralis and the nucleus vestibularis descendens. Any
fibers which course beyond the caudal limits of the Iatter
nucleus would do so in company with the tractus vestibulo-
spinalis, and will be considered with that tract. It should be
emphasized here that from the material available the exist-
ence of very many direct vestibular fibers farther caudal-
ward than the caudalmost limits of the nucleus vestibularis
descendens is questionable, although perhaps a few of the
decussating fibers which pass down in the medial longitudinal
fasciculus are of this nature, a point impossible of determina-
tion from the material. Certainly no direct vestibular fibers
can be traced into a position analogous to that which Beccari
( '12) figured as a direct vestibulo-spinal tract (p. 148).
For the most part, the above description obtains in all
other forms studied. However, where a well-developed
nucleus tangentialis is present many more of the vestibular
fibers (variable in size) are seen ending in relation to it.
The direct vestibular fibers, particularly the ascending, are
not relatively so numerous.
The foregoing account of the vestibular root fibers is simi-
lar, in general, to the previous accounts of Holmes ( '03),
142 JEAN K. WESTON
de Lange ('17)' Ariens Kappers ('21)' Huber and Crosby
( '26), Larsell ( '26 and '32)' Papez ( '29)' and others. For a
detailed account and analysis of the vestibular system, refer-
ence is made to Beccari's ('12) excellent study, with which
the above account is in essential accord, although much of the
finer detailed histology of endings, bifurcations, etc., as
Beccari described them, could not be verified completely.
This was due partially to the lack of embryonic material and
also because all the material available was prepared more
for the study of fiber tracts than for that of the finer histology
of the synapses.
Trigerninat root fibe,rs t o t he cerebellum
The presence of direct trigeminal fibers to the cerebellum,
in reptiles, is a point concerning which considerable differ-
ence of opinion exists. Banchi ('03) described such fibers
in turtles, snakes and lizards. Beccari ('12) also observed
their presence (see his figure 4, p. 663). De Lange ( '16)
doubted their existence in Crocodilia, although he stated that
such fibers appeared to be present in that form, as did Ariens
Kappers ('21, fig. 369), who noted further that they accom-
panied the dorso-spino-cerebellar tract (with which are closely
associated nucleo-cerebellar and secondary trigemino-cere-
bellar fibers). Huber and Crosby ( '26) also found evidence
of such fibers, noting particularly their close relation to the
nucleo-cerebellar system. Larsell ( '26) found some indica-
tion of their presence in lizards and snakes. Although most
of the above statements of opinion were in the affirmative,
most of these observers likewise noted that the actual evi-
dence left much to be desired in the way of conclusiveness.
However, Larsell ('32) described and figured a tract in an
embryo of Chrysemys which he considered as distinct evi-
dence for the presence of such direct fibers, and noted sup-
porting evidence in his adult turtle and alligator material
(see also p. 143). Such fibers have also been reported by
Woodburne ( '36).
The first evidence of direct trigeminal fibers to the cere-
bellum appeared in a transverse Weigert series of Chrysemys.
Here well-myelinated fibers, which unquestionably left the
Vth nerve, could be followed dorsalward and slightly rostral-
ward in company with the dorso-spino-cerebellar and nucleo-
cerebellar fibers, up to the base of the cerebellum. These
were never very numerous in any one section, and often could
not be followed so far toward the cerebellum, but their total
number seemed relatively quite large (fig. 26D). Silver
material verified the presence of such fibers, although they
could not be traced so far nor so easily here.
Weigert preparations of the horned toad and the snake
seemed to demonstrate conclusively the presence of direct
trigeminal fibers to the cerebellum (fig. 26A) in rather large
numbers. The figures of these forms (figs. 26A and 26C)
demonstrate such fibers completely enough so as to require
no further description. I n Alligator mississippiensis (fig.
26B) they are apparently much less numerous, considering
the relatively large size of the trigeminal nerve, and here
they could not be demonstrated so conclusively. However,
as shown in figure 26B, fibers can be observed definitely
leaving the Vth nerve and turning dorsofrontally in company
with the fiber band lying along the dorsolateral side of the
medulla oblongata, as Ariens Kappers ('21, fig. 369) ob-
served.
The exact distribution of these fibers in the cerebellum
would require degeneration preparations, but they appear to
distribute chiefly homolaterally and to end particularly in
relation with the granular cells of the lateral portion of the
cerebellum, probably chiefly in the auricular lobe (floccular
lobe of Larsell, '32), though this interpretation may not
account for all of them.
Reference to Larsell's ('32) figure 31 indicates that the
direct Vth fibers of the above description are considerably
more laterally located than those so designated in his figure
of a Chrysemys embryo. It is interesting in this connection
to note that in Beccari's ('12) figure 4, page 663, of an em-
bryo of Lacerta muralis, there are direct trigemino-cerebellar
144 JEAN K. WESTON
fibers labeled which lie considerably more laterally than the
same in Larsells figure, apparently agreeing more with the
position of these fibers as they are here described (see fig. 26).
However, in view of the lack of adequate embryological ma-
terial, it is not possible to venture an opinion as to whether
the tract which Larsell (32) labels the ascending Vth root
is homologous to that tract as described in the foregoing
account or not, since very possibly such a tract could become
forced more laterally with the further development and elabo-
ration of the gray of the region, and hence come to corre-
spond, in the adult, to its position as noted in this account.
Secolzndary af erent connection-s of the vest i bdar regiolz
I n Chrysemys, in addition to receiving the direct vestibular
root fibers, the vestibular region receives impulses originat-
ing in the cerebellum and in lower bulbar and cord centers as
well. It receives these not only through the stem fibers of
tracts which primarily interrelate such centers with the ves-
tibular gray, but also by way of the very rich and widespread
dendritic ramifications it maintains in close relation with
many other tracts coursing through and near the region, as
well as with contiguous nuclear masses.
This tract (figs. 18, 23 to
26) courses in part with the tractus vestibulo-cerebellaris and
in part with the tractus cerebello-spinalis, and will receive
detailed consideration along with those two tracts (pp. 151
and 169).
This tract will receive more
detailed consideration along with the tractus vestibulo-spi-
nalis, together with which it runs (figs. 20, 22, 23; p. 145).
Very likely this tract makes up the largest component of the
fiber system usually designated as the tractus vestibulo-
spinalis in reptiles. I ts origin could not be definitely deter-
mined in the available material. Apparently such origin,
for the most part, is in the spinal cord, but the possibility of
bulbo-vestibular fibers must be kept in mind, though such
could not be definitely demonstrated and are probably not
very numerous.
Tractus cerebello-vestibularis.
Tractus spino-vestibularis.
VESTIBULAR AXD CEREBELLAR SYSTEMS 145
Certain other secondary afferent connnections. It appears
in pyridine-silver and Golgi preparations that the cells of
nearly all the vestibular nuclei are similarly oriented. Hence,
practically throughout this region their dendrites can be ob-
served passing laterally into relation with the more laterally
coursing tracts to.the cerebellum. I n fact, Golgi preparations
show a rather dense network of these dendrites among such
tracts in places where the impregnation has been good.
Other dendrites pass dorsally into the cochlear gray and its
associated fiber tracts (though not to so great an extent in
Alligator, where the cochlear region is highly developed),
and still others pass ventrally into the reticular gray and the
fiber tracts of that region. I n addition, dendrites pass both
rostrally and caudally (and quite possibly across the midline),
thus subserving internuclear functions, as do certain den-
drites observed passing into the periventricular gray.
The richness and extensiveness of this dendritic ramifica-
tion is best developed among the cells of the nucleus vestibu-
laris ventrolateralis (as Beccari, '12, noted), where it is not
at a11unusual to encounter six and seven processes to a cell
of that nuclear mass, each accompanied by numerous fine,
unmedullated fibers which seem to be following along these
processes to enter into the pericellular synaptic net notice-
able about each of these cells. The above is not true for the
tangential nucleus, the dendrites of which are inevident or
very small, nor is it the case to so marked an extent in the
nucleus laminaris or the nucleus vestibularis ventromedialis.
Secondary ef erent cowneetiom of the vestibular regiow
Tractus vestibdo-spinalis. The account of this tract in
Chrysemys includes also a description of the spino-vestibular
and of the direct vestibulo-spinal path (p. 141), these three
being so intermingled that they are impossible of accurate
delimitation in the absence of degeneration material.
I n transverse series the tract is very obvious (fig. 20) and
is closely intermingled with the cells of the nucleus vestibu-
laris descendens (fig. 22). Laterally it is in relation with
THE JOURNAb OP $JOMPAL&ATIVE NEUBOLOGIY, VOL. 65
146 JEAN X. WESTON
the ascending cerebellar tracts and with secondary cochlear
fibers; dorsally, with the cochlear gray and its associated
fiber system ; medially, with the tractus cerebello-spinalis
(which may carry a small vestibulo-spinal component, par-
ticularly from the pars dorsalis of the nucleus vestibularis
ventrolateralis, figs. 20 and 23), and with the nucleus vestibu-
laris ventromedialis and the fasciculus solitarius ; ventrally,
with the descending root of the Vth nerve and its nucleus.
Its fibers appear to intermingle somewhat with those of the
adjacent tracts. Followed caudally, it seems to decrease con-
siderably in size, probably due to the direct component
terminating in relation with the nucleus vestibularis de-
scendens, and to intermingling with contiguous fiber tracts,
particularly the mediodorsal portion of the descending Vth
root. This intermingling becomes more evident and the tract
more difficult to delimit from the descending Vth root as it
is followed caudally, where it also appears to amalgamate
with the cerebello-spinal tract. Caudal to the entrance of
the Xth roots, little fascicles of the tractus vestibulo-spinalis
can be noted detaching themselves from the dorsomedial
border of the descending Vth root and running ventrally and
slightly medially (through and medial to the nucleus of the
descending Vth root) to become lost in the ventral part of
the reticular formation. These little fascicles are no longer
evident at the junction of the spinal cord and the medulla
oblongata, though it is easy to determine that the tract con-
tinues over into the cord. However, such fascicles do not
seem to account for all the fibers of this tract, and one is thus
forced to the conclusion that a considerable component of the
tract is still intermingled laterally with the descending Vth
root and adjacent tracts, or else has ended, little by little, in
relation to some of the nuclei of the brain stem.
The nucleus vestibularis descendens appears to be most
intimately associated with this tract, and while absolute proof
was not attainable, it is probable that the larger part (if not
all) of the descending vestibular root ends in relation to its
cells, and that this relation holds for the spino-vestibular
VESTIBULAR, AND CEREBELLAR SYSTEMS 147
component of the tract also. Furthermore, it would seem
that the vestibulo-spinal component proper (which does not
appear to be very large or obvious, if present at all) is con-
tributed principally by the nucleus vestibularis ventrolater-
alis, pars ventralis ; the nucleus vestibularis descendens
adding but little, if any, to this tract, but discharging by way
of the medial longitudinal fasciculus. It appears doubtful
whether any of the vestibular nuclei rostral to the nucleus
vestibularis ventrolateralis contribute to the vestibulo-spinal
tract, and conversely, whether the spino-vestibular component
proper contributes to the vestibular nuclei farther rostral-
ward than the entering vestibular roots, though this observa-
tion could not be verified satisfactorily. Very likely this tract
also carries a vestibulo-bulbar component, which would dis-
tribute to the motor nuclei of the brain stem. Regarding
such a connection, Tuge ( '32) has noted collaterals from this
tract to the small neurons of the XI I th nucleus, as well as
dendrites of the XI I th nucleus ramifying among the fibers
of the vestibulo-spinal tract (p. 148). This latter observation
has been verified in the material available, and this fact is
equally true for the dendrites of the Xth nucleus, as well as
for those of the nucleus of the descending root of the Vth.
The above description is applicable also to the alligator and
the snake.
I n lizards it is possible to find a vestibulo-spinal tract fully
comparable to the above. I n addition, there is a very notice-
able vestibulo-spinal tract somewhat more medially, lying
ventral to the level of the entering motor VIIth root and
slightly ventrolateral to the inferolateral angle of the rVth
ventricle (fig. 25B). This tract is almost entirely made up
of the relatively large, dark staining neuraxes of the large
cells of the nucleus vestibularis ventrolateralis. A few
neuraxes appear to be contributed by cells of the nucleus
tangentialis, though these are very unusual and it is uncer-
tain how far they accompany this tract. The tract can be
readily followed caudally into the spinal cord, in its course
moving ventromedially and lying finally just lateral to the
148 JEAN K. WESTON
medial longitudinal fasciculus. It has been impossible to
verify such a tract in any of the other forms studied. How-
ever, in the other forms, particularly in the turtle and the
alligator, there is a relatively large vestibulo-reticular con-
nection in the same relation (p. 139). This tract in lizards
might be justifiably termed a Deiterso-spinal path and be
considered homologous to the ventrolateral vestibulo-spinal
path of higher forms. This is the tract which Beccari figured
f or embryos of Lacerta muralis ('12; figs. 7 and 8, fascio
vestibolo spinale-diretto) as a direct vestibulo-spinal path,
though he briefly described a vestibulo-spinal path as origi-
nating from the cells of Deiters' nucleus. However, he
showed (in his fig. 8) a descending vestibular root conform-
ing in its relations to the description here given of the
vestibulo-spinal path (of which it is here considered a com-
ponent).
The vestibulo-spinal tract has received brief mention by
some of the earlier workers (Edinger, '08, and earlier;
Holmes, '03; and others). De Lange ('16) and Ariens
Kappers ( '21) noted it in Alligator ; Papez ( '29) labeled and
briefly described it in the snake, essentially as given in the
above account. Shanklin ( '30) described the vestibulo-spinal
tract as consisting of two parts, a direct vestibulo-spinal path
(corresponding in its relations to the one described above
for the turtle), and a Deiterso-spinal tract, originating in
Deiters ' nucleus (nucleus vestibularis ventrolateralis). This
latter appears to be homologous to the tractus cerebello-
spinalis of this report, which probably carries a small
vestibulo-spinal component, and becomes amalgamated with
the vestibulo-spinal tract caudally. Frederikse ( '31) figured
a descending vestibular root in a position analogous to that
of the present account. Tuge ('32) labeled a tract in Chrys-
emys in the general region of the vestibulo-spinal tract which
is homologous in position with Beccari's direct vestibulo-
spinal path in Lacerta.
Cornrnissural fibers. From all the. vestibular nuclei fibers
pass toward the inferolateral angle of the I Vth ventricle,
where they enter into the band of commissural fibers (figs.
19 and 25) which passes below the ventricle and toward the
midline. This is also true of vestibular root fibers (p. 138).
The nucleus vestibularis ventrolateralis is the largest con-
tributor, although the nucleus vestibularis descendens also
gives rise to a considerable number. Fibers from the coch-
lear nuclei and from the chief sensory nucleus and the nucleus
of the descending root of the Vth help make up this com-
missural system; these are usually of smaller caliber. I n
addition, decussating motor fibers of the Vth nerve and the
motor root of the VI I th nerve are also present, their con-
centration depending somewhat on the level of the section.
From the above diverse list, it can readily be understood that
a considerable degree of error might easily creep into any
attempt at an exact evaluation of their further course. How-
ever, the following statements may be made with some degree
of confidence.
Certain of the secondary vestibular fibers which enter into
the commissural group can be observed to cross the raphe
and pass to the vestibular gray of the other side, hence serv-
ing for coordination of the two sides. A questionable number
of these fibers are doubtless direct vestibular root fibers.
These last appear relatively more numerous in lizards and
snakes, being overshadowed in the turtle and the alligator
by the marked development of the secondary cochlear fibers
crossing here. Most of the commissural fibers quite evidently
run into the medial longitudinal fasciculus, where the larger
number appear to pass caudally, in the contralateral fascicu-
lus, down into the cord regions, as many observers have noted,
probably corresponding to the medial vestibulo-spinal tract
of higher forms. Certain of these secondary fibers pass
rostralward or caudalward in the homolateral medial longi-
tudinal fasciculus.
The relative proportion of fibers which follow these various
courses and their exact nucleus of origin are difficult of
accurate determination. Certainly the nucleus tangentialis
is a large contributor, as is the nucleus vestibularis ventro-
150 JEAN K. WESTON
lateralis, although all the vestibular nuclei here seem to con-
tribute some fibers to the medial longitudinal fasciculus.
Very probably those from the nucleus vestibularis dorso-
lateralis and the nucleus vestibularis superior are more
concerned with the motor portion of the I I I rd and with the
IVth nuclei, and would then correspond to what has been
designated the vestibulo-mesencephalic system (Papez, '29,
for snake; Tuge, '32, for Chrysemys; and others).
Certain of the commissural fibers, after crossing the mid-
line, pass lateroventrally into the reticular gray. The nucleus
of the VIth lies right in the midst of these commissural fibers,
and some of them appear to end in relation to its cells. Cer-
tainly the observations of Tuge ( '32), that some of the den-
drites of the VIth nucleus pass around the inferolateral angle
of the ventricle and get into the vestibular gray, can be readily
verified.
Another component is the rather numerous group of
secondary vestibular fibers that courses toward the infero-
lateral angle of the ventricle as though to add to the com-
missural fibers, but just before reaching that position turns
ventrally into the homolateral reticular gray, where the fibers
appear to terminate (figs. 19 and 25). These vestibulo-
reticular fibers are quite numerous, excepting among the
lizards. Most of these fibers are contributed by the nucleus
vestibularis ventrolateralis. Some of them apparently pass
into relation with motor nuclei at appropriate levels, notably
the motor Vth and motor VI I th nuclei, and it is possible that
some may pass caudally, without synapse, in the reticular
substance to lower motor centers in the cord, but these could
not be demonstrated in forms other than lizards (p. 147).
As observed by Beccari ( '12), Tuge ( '32), and many others,
the dendrites of the reticular cells pass up and into relation
with these commissural fibers, as well as into relation with
the vestibular gray. All the above statements have been
quite well substantiated and carefully checked through a
study of Golgi material.
It should be noted that in large part the above account
verifies the more general observations of many of the earlier
workers (Edinger, '08 and earlier ; Holmes, '03 ; Beccari, '12 ;
and others). Various parts of the above account have been
either briefly noted or expressed in figures by Papea ('29),
Ariens Kappers ( '21 and '29), Shanklin ( '30), Frederikse
( '31), Tuge ( '32) and others.
Tractus vestibulo-cerebellaris. The direct vestibular com-
ponent of this system has already been considered briefly
(p. 140; also fig. 22) and will be more fully reviewed here.
Intermingled with this tract are a considerable number of
cerebello-vestibular fibers (p. 169).
The secondary vestibular component of this tract is com-
posed chiefly of fibers which originate, apparently, from the
cells of the nucleus vestibularis ventrolateralis and the
nucleus vestibularis dorsolateralis. Some of these fibers may
be carrying impulses which originally reached the medulla
oblongata via the vestibular nerve of the opposite side (and
perhaps it is these which cross in the cerebellar commissure).
It is possible that some of the crossed vestibular fibers con-
tributed by the nucleus tangentialis and the nucleus vestibu-
laris descendens add to this tract, but if present, such a
component is very small. I n addition, the nucleus laminaris
contributes a noticeable component to this tract, which is
included in the tractus cochleo-vestibulo-cerebellaris of Ariens
Kappers ( '21) and Huber and Crosby ( '26), and the tractus
lamino-cerebellaris of Shanklin ( '30).
From the level of the entering vestibular roots, caudally,
to the base of the cerebellum, rostrally, the vestibulo-cere-
bellar tract occupies a large part of the middle and inferior
portion of the lateral limb of the medulla (figs. 18, 19, 22, 23,
25 and 26). It is limited dorsally by the cochlear gray
and cochlear fiber paths; medially, by the dorsal arcuate
system (more caudally, and the cerebello-tegmental system,
more rostrally) ; ventrally, it borders on the chief sensory
nucleus and the nucleus of the descending root of the Vth,
part of the descending root of the Vth, more caudally, and
152 J EAN K. WESTON
the reticular gray, more rostrally; laterally, it runs in com-
pany with the ventral superficial arcuate, spino-cerebellar and
direct trigemino-cerebellar systems, and is intermingled with
these to some extent, particularly rostrally. Some nucleo-
cerebellar fibers, notably from the chief sensory nucleus of
the Vth, seem to accompany the fibers of the vestibulo-cere-
bellar tract (fig. 25). Followed rostrally, the tract curves
laterally (always closely applied to the scattered cells of the
nucleus vestibularis dorsolateralis, which are concentrated in
part on its medial side), at the same time sweeping dorsally.
This course brings it lateral and ventral to the lateral recess
and up and into the more caudal part of the cerebello-oblon-
gata junction. It can readily be seen that such a course would
carry it dorsolateral to the nucleus vestibularis superior. As
noted, the direct vestibular root fibers follow a similar course,
and a considerable portion of them pass to the auricular lobe,
the rest of them distributing to other portions of the cere-
bellum. It must be emphasized that in the absence of de-
generation preparations it is not possible to be dogmatic
about the exact termination or the relative number of the
direct root fibers, since it would be very possible that a large
number of them terminate on the cells of the nucleus vestibu-
laris dorsolateralis and continue onward from there as
secondary fibers (which is probably the case, to some extent
at least). However, in addition to the component to the
auricular lobe (which is secondary in part, accompanied by
direct and secondary cochlear fibers, doubtless the greater
portion of the tractus cochleo-vestibulo-cerebellaris, see p.
151), the larger number of vestibular fibers swing up and back
distributing to the granular layer of the cerebellum, where
they spread laterally and caudally, as liarsell ( '32) has noted;
this agrees with I npar's ('19) observation on the crocodile.
Much of the tract is in relation with the medial cerebellar
nucleus, where some of its fibers appear to terminate-an
appearance due in part to the fact that the accompanying
cerebello-vestibular component originates here. Some of the
fibers pass into the more caudal portions of the cerebellar
commissure, and probably reach the opposite side, in part,
no doubt, cerebello-vestibular, with origin in the opposite
nucleus medialis cerebelli, but possibly contributed to by the
axones of Purkinje cells, both homo- and contralateral, de-
spite certain experimental observations on higher forms to
the contrary.
I n the smaller forms, such as the lizard and the snake, due
partially to the relatively smaller rostrocaudal extent of the
region, and, in the case of the lizard, to the eversion of the
cerebellum (which tendency in Alligator leads to similar rela-
tions there), there is a sharper ventrodorsal gradient to the
course of this tract, particularly near its termination, though
actually the significant relations as noted are very similar
throughout.
I n many respects the foregoing account is but little differ-
ent from previous descriptions of this tract (Edinger, '08
and earlier ; Holmes, '02 ; Beccari, '12 ; Ariens Kappers, '21
and elsewhere; Huber and Crosby, '26; Larsell, '26 and '32;
Papez, '29 ; Hindenach, '31 ; Shanklin, '30 ; Frederikse, '31 ;
and others).
FIBER CONNECTIONS OF THE CEREBELLUM
Af er ent cerebellar connections
Due to the many times this
tract has already been described (Edinger, '08 and earlier ;
de Lange, '16 ; Ariens Kappers, '21 ; Huber and Crosby, '26 ;
Larsell, '26, and others), another detailed description of it
would be superfluous, particularly in view of the complete
consideration Larsell ('32) has given to this system. Our
material agrees essentially with the greater part of his de-
scription at least so far as based on adult material.
Briefly, then (based on Chrysemys), the spino-cerebellar
system can be followed forward from the junction of the
medulla oblongata with the spinal cord, occupying a ventro-
lateral position near the periphery of the bulb. Slightly
rostra1 of the entering Xth roots some of the more dorsally
situated of these fibers begin to edge laterally around the
Tractus spino-cerebellaris.
154 J EAN K. WESTON
descending Vth root, a few at a time, mixed somewhat with
the more rostra1 nucleo-cerebellar fibers. It is these former
fibers which compose the dorsal spino-cerebellar tract. They
come to lie just dorsal to the descending Vth root, continu-
ally recruiting more from below, until about the level of the
entrance of the Vth roots, when they swing dorsalward into
the cerebellum, distributing much as Larsell (32) noted (figs.
17 to 23, 25, 26). Larsell (32, p. 336) says of it: I n its
course through the bulb the tract under consideration lies
just dorsal to the spinal Vth bundle and arches dorsally just
caudad of the level of the Vth root. The material studied
is in agreement with this only to about the level of entrance
of the Xth root caudally, caudal to which it lies more ventro-
laterally and in company with the ventral spino-cerebellar
tract (see also p. 159).
The more ventrally lying component of the spino-cerebellar
system continues forward, moving slightly more dorsolater-
ally as the dorsal spino-cerebellar fibers swing around the
descending Vth root, until at about the level of entrance of
the VI I I th roots it lies mostly ventral to the descending root
of the Vth (figs. 19 and 26). It maintains this relation as it
passes forward, until about the level of the entrance of the
Vth roots, when it curves dorsally and rostrally, in part pass-
ing through the more ventrally entering Vth roots (figs. 16 to
18, 26) to form a considerable part of the cerebellar com-
missure. This is the ventral spino-cerebellar tract which,
throughout much of its course, is very intimately related
ventrally and medially to the spino-mesencephalic system and
the lateral lemniscus.
Where the ventral spino-cerebellar tract curves dorsally
to enter the cerebellar commissure, several rather obvious
fascicles swing abruptly ventrofrontalward from its more
dorsal portion (shown best in Weigert sagittal series). No
especial effort has been made to identify these, but they seem
to originate mostly from the chief sensory nucleus of the Vth
and might be the homologue of Wallenbergs (1898, 03)
quinto-frontal path in birds.
I n subscribing to the terms dorsal spino-cerebellar and
ventral spino-cerebellar as Larsell ('32) uses them, no ho-
mologies are necessarily implied with higher forms (see also
Larsell, '32), and further no definite line of demarcation
exists between these two tracts morphologically, particularly
in the absence of evidence regarding their origin in the cord
(Huber and Crosby, '26), although such a general division
does make it easier to describe the system and is essentially
in accord with the anatomic relations encountered. These
fibers (particularly the ventral spino-cerebellars) are in very
close anatomic relation to the cerebellar nuclei, especially the
medial nucleus (as Huber and Crosby, '26, noted for this
tract in the alligator).
The alligator appears to show a considerable relative
diminution in the size of the spino-cerebellar system, par-
ticularly of the ventral spino-cerebellar tract. J t is this
relatively (as compared with turtle) smaller bundle of fibers
which Huber and Crosby ('26) described as accompanying
the spino-mesencephalic fibers and then entering the cere-
bellum in the alligator, and considered as the possible homo-
l ope of the mammalian ventral spino-cerebellar system.
Larsell ('32) verified that observation in these forms. Per-
haps this disparity in size is partially due to the tendency
toward eversion of the rostral part of the cerebellar plate.
This would force the spino-cerebellar fibers to adopt a steeper
ventrodorsal gradient near the rostral part of their course
in order to attain their objective, and consequently more of
them would pass dorsal to the entering Vth roots or through
these roots, thus causing the apparent marked decrease in
size, more particularly of the ventral spino-cerebellar com-
ponent. This emphasizes that the main distinction between
these two tracts in reptiles appears to be their relation to the
entering Vth roots-a variable relation largely associated
with differences in the form of the cerebellar plate. How-
ever, there still obtains a relative decrease in the whole
spino-cerebellar system in the alligator, considering the much
larger size of the cerebellum (and especially the fact that it
156 JEAN K. WESTON
is being compared with the turtle, a form where there is
probably little, if any, contribution from trunk regions), for
which another explanation must be sought. Perhaps the
alligator is approaching the avian condition, where many
intersegmental fibers are present in the cord, forming there
a considerable component of the spino-cerebellar system, but
being spino-spinal rather than spino-cerebellar, so that, de-
spite the relatively larger muscular development, the tract
in the alligator does not show the same relative progressive
increase from caudal to rostra1 portions of the cord as is
the case in the higher mammals, particularly. It should be
noted, however, that the spino-cerebellar system is relatively
much larger in birds, in spite of the loss of this interseg-
mental component throughout the cord, and this is doubtless
correlated with the larger size of the cerebellum, both being
central expressions of the necessarily highly developed pro-
prioceptive mechanisms incident to flight.
Nucleo-cerebellar system. Here are included all the con-
tributions to the cerebellum from bulbar centers caudal to
the vestibular complex. I n most of the literature these fibers
have been included with the spino-cerebellar system, under
the term tractus spino-cerebellaris (de Lange, '16 ; Ariens
Kappers, '21 ; and others), or tractus spino-cerebellaris dor-
salis (Larsell, '26 ; Hindenach, '31 ; and others) , these latter
observers in many cases considering the term dorsal spino-
cerebellar as synonymous with bulbo-cerebellar, being a
rather loose usage of terminology, which has led to some
confusion (p. 159). Here the term nucleo-cerebellar is applied
to the diverse fibers originating from various bulbar centers,
including ventral superficial arcuate, olivo-cerebellar (if pres-
ent), and reticulo-cerebellar fibers, as well as fibers from the
more caudal sensory nuclei of the brain stem which are re-
lated to the cerebellum (these last mediate impulses brought
in over the descending root of the Vth and the somatic sen-
sory vagal component, which, after synapse, as Huber and
Crosby, '26, observed, run to the cerebellum).
Ventral superficial arcuates. At the very caudalmost levels
of the medulla oblongata, particularly noticeable in trans-
verse Weigert series of Chrysemys, though quite well de-
veloped in all the forms studied, many fibers cross the midline
in the more ventral portions of the raph6. After decussation,
these become grouped into little fascicles, and pass dorsal-
ward and slightly frontalward, lying at the very periphery
of the medulla oblongata. Many of these fibers originally
swing ventromedially out of the nuclei funiculi posterioris
(which in these forms have been homologized with the mam-
malian nucleus gracilis and nucleus cuneatus, Zeehandelaar,
'21), curve down to cross in the raphe at varying levels and
continue as noted above, apparently being homologous to the
ventral superficial arcuates of higher forms. Added to these
are many fibers derived from the reticular nuclei.
Followed forward, the ventral superficial arcuates eventu-
ally occupy the very dorsolateral angle of the lateral limb
of the medulla oblongata (at about the level of entrance of
the Xth nerve, fig. 20), retaining that position until the caudal
part of the cerebello-oblongata junction is reached, when they
turn dorsalward into the lateral part of the cerebellum, appar-
ently distributing, along with the reticulo-cerebellar, olivo-
cerebellar, part of the trigemino-cerebellar, and dorsal
spino-cerebellar systems, to the more lateral (including the
auricular lobe) and caudal parts of the cerebellum (figs. 17,
18, 26). This whole complex that swings up into the cere-
cellum is homologous to the inferior cerebellar peduncle or
restiform body of higher forms, as has been observed pre-
viously (Edinger, '08; Ariens Kappers, '21; Larsell, '26 and
'32; and others).
Banchi ('03) was the only observer who noted a ventral
superficial arcuate system among reptiles, although his brief
description is couched in such general terms that it is difficult
to be certain that he did not include with it the whole of the
nucleo-cerebellar system of this account.
Olivo- and reticulo-cerebellar fibers. It has been impossi-
ble to demonstrate satisfactorily an inferior olive and a con-
sequent olivo-cerebellar tract. True, it is possible to identify
158 JEAN I<. WESTON
a more or less well-defined group of cells in the region where
the inferior olive is alleged to be, but so far as can be deter-
mined, on the basis either of cell type or connections, this
mass can equally as well (and probably even better), be con-
sidered as a group of reticular cells; like the other reticular
cells they contribute fibers to the cerebellum. Practically all
of the workers on the reptilian cerebellum and bulb have
noted an inferior olive and olivo-cerebellar fibers (not, how-
ever, with the same degree of assurance), though practically
all are agreed that the olivo-cerebellar fibers become inter-
mingled with the nucleo-cerebellar and dorsal spino-cerebellar
component as herein described, and cannot be traced sepa-
rately to the cerebellum. Certainly if any one of the reticular
cell masses located in the caudoventral portion of the medulla
oblongata is considered an inferior olive, such a course for
its fibers as just noted can be readily verified in the material
studied. However, it is of interest that Kooy ('17) noted
that the inferior olive in reptiles is not evident, and that Tuge
('32) was unable to identify it satisfactorily in Chrysemys.
Among those who maintain the existence of an inferior olive
and olivo-cerebellar fibers in these forms may be mentioned
Ariens Kappers ( '21), Larsell ('26 and '32), and others. It
appears that an inferior olive and olivo-cerebellar fibers have
yet to be conclusively demonstrated in reptiles. It is difficult
to determine what criteria can be used in these forms to con-
clusively differentiate an inferior olive from the various
reticular gray masses scattered throughout this part of the
bulb. Possibly one way this might be definitely settled would
be by means of a study of the embryology of this region in
reptiles, side by side with a similar study in mammals.
Certain other nucleo-cerebellar connections. From brain
stem levels caudal to the entrance of the VI I I th nerve and
rostra1 to the XI I th nucleus, fibers can be observed still cross-
ing the midline and following the superficial path noted above
for the ventral superficial arcuates. These seem to be chiefly
contributions from the nucleus of the descending root of the
Vth and the reticular gray, as well as other sensory nuclei
of the region (p. 156). The contribution from the nucleus of
the descending root of the Vth nerve can be considered as the
caudal portion of the ventral trigemino-cerebellar system of
Huber and Crosby ('26), though here it is included with the
nucleo-cerebellar system (p. 156).
The above, then, make up the strictly bulbo- or better,
nucleo-cerebellar system. A fairly good indication of its size
can be obtained at about the level of the Xth nerve entrance,
at which point but few (if any) dorsal spino-cerebellar fibers
have become intermingled with it (fig. 20). The whole system
runs forward with the tractus spino-cerebellaris dorsalis and
distributes as noted previously (figs. 17 to 26). It seems to
be homologous with the bulbo-cerebellar system of Herrick
('14, '14a, '30) and Larsell ('20, '23, '25, '31, '32) for
Amphibia.
Tractus trigemislo-cerebellaris. The various secondary
trigemino-cerebellar connections have been well described by
Huber and Crosby ('26) and their interpretation has been
verified in the material studied. There is a dorsal trigemino-
cerebellar tract (fig. 26, tr.trig.cer.rect.), from the homo-
lateral chief sensory nucleus of the Vth and passing
dorsomedially, in relation with the cerebello-tegmental and
vestibulo-cerebellar tracts, to the region of the cerebellar
commissure, where its fibers become lost. Fibers of similar
origin appeared to pass dorsolaterally and frontally in com-
pany with the nucleo-cerebellar and dorsal spino-cerebellar
fibers already discussed (fig. 26). Following this course,
these fibers necessarily come into fairly close relationship
with the cerebellar nuclei.
The ventral or cruciate trigemino-cerebellar path (figs. 16,
17) is crossed and is contributed to both by the nucleus of the
descending root and by the chief sensory nucleus of the Vth
nerve. The more caudal portion of this tract is considered
with the nucleo-cerebellar system (p. 156). To this is added
more rostrally the fibers from the chief sensory nucleus of the
Vth nerve. The contribution from the more dorsal part of the
trigeminal lemniscus (Huber and Crosby, '26) passes into the
160 JEAN K. WESTON
base of the cerebellum, where it becomes lost (fig. 17). This
last contribution seems relatively larger in the snake, but is
so intermingled with other fibers to the cerebellum that it
is difficult to truly evaluate its development.
Larsell ( '26) noted no secondary trigemino-cerebellar fibers
in snakes and lizards. Shanklin ('30) briefly noted a small
bundle of secondary Vth fibers to the cerebellum. However,
Larsell ( '32) mentioned trigemino-cerebellar fibers in the
turtle and the alligator, but stressed more their commissural
nature. These appear to be homologous to the dorsal
trigemino-cerebellar system of Huber and Crosby ( '26) and
of this account, and possibly in part to the contribution from
the trigeminal lemniscus. Larsell's results are based to a
considerable extent on embryological material, and hence
could not be verified in the material available ; consequently,
although unquestioned trigeminal fibers can be traced into
the region of the cerebellar commissure, we are not prepared
to state whether or not they are merely commissural in
nature.
Tract us tecto-cerebeEZaris. As a prelude to taking up the
tecto-cerebellar system, the possibility of an isthmo-cerebellar
connection might well be considered. The nuclei isthmi,
lying one on each side at the junction of the medulla oblon-
gata and the mid-brain, and at the periphery of the brain
(except in snakes, where this nucleus is not obvious), are
interconnected by finely medullated commissural fibers, pass-
ing both through the anterior medullary velum and the cere-
bellar commissure. These proceed from the more dorsomedial
portion of the pars magnocellularis of the nucleus isthmi
(fig. 27). Fibers can also be noted swinging into the cere-
bellar commissure from more ventral and caudal levels of
the nucleus. Consequently, there is a strong probability of
an isthmo-cerebellar connection, which, if present, might be
considered an indirect tecto-cerebellar connection (i.e., a
tecto-isthmo-cerebellar connection). Some of the fibers of
the tecto-isthmal system which pass to this region appear
to turn caudomedially, without synapse, and add to the dorsal
VESTIBULAR. AND CEREBELLAR SYSTEMS 161
part of the cerebellar commissure, hence probably adding to
the tecto-cerebellar system (fig. 27).
Tecto-cerebellar fibers were first definitely demonstrated in
reptiles by Huber and Crosby ( '26) in the turtle. Since then
they have been observed among lizards (Larsell, '26 ; Shank-
lin, '30) ; in Sphenodon (Hindenach, '31), in the turtle
(Larsell, '32), and again noted by Huber and Crosby ( '33),
who verified, in lizards, Shanklin's ( '30) observation for
Chameleon that there exists a crossed as well as an uncrossed
component. Due to the rich collection of exceptionally well-
impregnated pyridine-silver material, it has been possible to
verify this tract in all the forms examined (fig. 27).
The tecto-cerebellar fibers are of two main types, crossed
and uncrossed. The uncrossed component (chiefly thinly
medullated) can be seen swinging out of the stratum album
centrale, passing beneath the decussating fibers of the IVth
nerve, and coursing on the dorsal surface of the anterior
medullary velum to swing up and into the cerebellum (fig.
27D). Certain of them turn deeper into the cerebellar com-
missure; others reach the region of the nucleus medialis
cerebelli, where they are lost ; still others pass into the granu-
lar layer. A few have been followed rather far into the
molecular layer, a fact suggestive of their termination as
climbing fibers (which have been demonstrated in the rep-
tilian cerebellum; Ram6n y Cajal, '11; Larsell, '32). These
fibers are difficult to see at all in any but sagittal sections,
and from such sections a good conception of their number
and extent cannot be obtained. 'When it is considered that
they can be identified from near the midline to the plane of
the dorsomedial extremity of the nucleus isthmi pars magno-
cellularis, even though but few fibers appear in any one
section, the sum total makes up a rather large tract. A con-
siderable number of the coarsely medullated fibers of the
stratum album centrale are also observed following a similar
course (fig. 27). I n addition, the fine, superficially placed
fibers which Huber and Crosby ('33) observed in Weigert
series and traced into the stratum griseum et fibrosum
THE JOUENAL OF CO?~PARATIVE NEuROLoaY, vofi. 65
162 JEAN I<. WESTON
superficiale (thus forming a cerebello-tectal path) have been
verified, in the silver material of the turtle, the alligator, and
the larger lizards, but were not demonstrable in the smaller
lizards nor in the snake (fig. 27). However, none of these
fibers could be positively traced to their cells of origin or to
their exact termination.
The uncrossed tecto-cerebellar tract has been known in
Amphibia for some time. Herrick ('25, fig. 39, p. 4771, espe-
cially, noted it and considered that it had origin from the
corpus posticum (the amphibian homologue of the inferior
colliculus). Golgi material of turtle demonstrates, as Huber
and Crosby ('33) have shown, that fibers from the reptilian
inferior colliculus add to the tecto-cerebellar system, but this
component does not appear to be very large, since silver
preparations show by far the larger number of fibers swing-
ing out of the stratum album centrale and passing over the
top of the inferior colliculus.
Regarding the relative contribution from the two colliculi,
the following facts are very interesting. It is noteworthy
that in the bony fishes, particularly, where the optic tectum
is very highly developed, the tecto-cerebellar system from
the optic tectum is very large. I n Amphibia, where the optic
tectum is relatively less developed and the corpus posticum
of good size, the tecto-cerebellar tract is largely, if not en-
tirely, a contribution from the corpus ,posticum. I n reptiles,
where the optic tectum is well developed and the inferior
colliculus developed to a variable degree, both types of tecto-
cerebellar fibers are present, the main contribution coming
from the optic tectum. I n mammals, with the relative de-
crease in importance of the optic tectum as a sensory correla-
tion center, the main contribution of tecto-cerebellar fibers is
apparently from the inferior colliculus. From the above it
can be seen that the term tecto-cerebellar has been used for
two entirely different sets of connections, very often inter-
changeably, which has led to some confusion in the literature.
Finely medullated fibers from the stratum album centrale
can be observed crossing in the caudal part of the tectum and
then almost immediately turning downward and caudalward
over the top of the anterior medullary velum, mixed with the
more medial part of the uncrossed component and distribut-
ing similarly (fig. 27). A larger number of fibers of this
crossed component insinuate themselves between the decus-
sating IVth nerve fibers and then pass into the cerebellar
commissure more medially.
The very rich periventricular connection between cerebel-
lum and tectum (fig. 27) has been observed by Huber and
Crosby ( '26 and '33), who noted ( '33) that this system is
afferent with respect to the tectum, thus composing a cere-
bello-tectal system of connections. The origin of such fibers
is still in doubt. Some have been traced into the region of
the medial cerebellar nuclei, but a great many more have been
followed well up into the cerebellar plate ; a few to very near
the layer of Purkinje cells. This system of fibers is closely
interwoven with those periventricular fibers noted beneath
the ependymal lining of the IVth ventricle throughout its
extent. Not enough evidence is at hand regarding this system
to attempt an interpretation of its functional significance at
this time.
Here should be mentioned the quite obvious mesencephalic
Vth fibers which appear to run into the cerebellum (see fig.
27D). These are very clear in some of the alligator and
snake material. I n Sternotherus a few of them have been
traced into the medial cerebellar nucleus, and others far into
the granular layer before they were lost. The strong possi-
bility of such a connection was noted by Weinberg ('28) and
others, but the earlier evidence was not very conclusive.
Ef ereBt cerebellar paths
Cerebello-tegmerztal system. The efferent cerebellar path-
ways are described here under a terminology slightly differ-
ent from that in current use. This appears the more advisa-
ble and justifiable since in certain instances the same termi-
nology does not appear to include the same group of fibers
to different observers, and also since some of the current
164 JEAN K. WESTON
terminology seems to imply at the same time both too much
and too little in the way of homology with mammalian forms.
Here all the efferent cerebellar tracts are considered as
the cerebello-tegmental system excepting the cerebello-tectal,
the cerebello-vestibular, and the cerebello-spinal fibers, which
latter two seem to correspond to the hooked fasciculus of
Russell (1894) for mammals. This includes the fibers which
curve ventrally out of the inferior portion of the cerebellar
commissure, fanning out rostralward and caudalward as
rather discontinuous, small bundles which lie at various levels
from near the periphery to near the ventricle (figs. 16, 17,
21, 22). The cerebello-tegmental system is then arbitrarily
subdivided into a portion passing rostrally to mid-brain
regions (the tractus cerebello-motorius et tegmentalis mesen-
cephali), and a portion passing caudally to bulbar regions
(the tractus cerebello-motorius et tegmentalis bulbaris).
Each of these may be rather imperfectly separated into a
tractus cerebello-motorius (either bulbaris or mesencephali)
and a tractus cerebello-tegmentalis (either bulbaris or mesen-
cephali) for these two regions, wherever such a subdivision
serves to make the description more comprehensible. The
term cerebello-motorius is restricted to that component more
strictly concerned with the motor nuclei of the cranial nerves,
insofar as such a component is capable of differentiation,
and the term cerebello-tegmentalis is restricted to that com-
ponent (the larger) which distributes largely to the more
highly specialized gray of the tegmental region, such as the
red nucleus. These components overlap to a considerable
extent. Such a terminology leaves much to be desired, but
appears to be sufficiently general to permit of ready adapta-
tion to other forms, and at the same time specific enough so
that there can be little question of the portion of this large
system under discussion at any time.
Tractus cerebello-motorius et tegmentalis mesencephali.
This portion is less easily separated into motor and tegmental
components than is the bulbar part of the system. Certainly
if there is a mammalian homologue of the brachium con-
junctivum in these forms it is this portion of the cerebello-
tegmental system. However, just as certainly this tract is
not homologous to the brachium conjunctivum as the term
has apparently been used in much of the literature on rep-
tiles, where it is restricted to the strictly cerebello-rubral
part of the system, which is only a part of the tractus
cerebello-tegmentalis mesencephali. However, if the term
brachium conjunctivum is used in the sense of the literature
on mammalian forms, notably human, it is in most respects
homologous to the tractus cerebello-motorius et tegmentalis
mesencephali of this account, varying in the relative develop-
ment of its components. I n other words, here is a superior
cerebellar peduncle surprisingly homologous to the mam-
malian superior cerebellar peduncle.
Leaving the more frontoventral part of the cerebellar com-
missure and passing ventrolaterally and frontally, chiefly
dorsolateral to the mesencephalic root of the Vth nerve
(which separates this component from the tractus cerebello-
motorius mesencephali) and ventromedial to the nucleus
isthmi pars magnocellularis, a rather thin, discontinuous
sheet of fibers (chiefly medullated) can be readily observed,
particularly in sagittal sections (excepting in snakes, in which
the nucleus isthmi is inevident, where these fibers simply
course through the fiber tracts in that region at an angle,
having otherwise the same relations). Following this
course, the tractus cerebello-tegmentalis mesencephali passes
through, and largely dorsal to, the bundles of the tractus
cerebello-tegmentalis bulbaris (fig. 16, 17, 24). I n exception-
ally favorably cut sections of young Chelhydra and of the
alligator, the whole sweep of these fibers can be seen and are
found to be more concentrated rather far laterally in the
mid-brain (figs. 21 and 22). They curve ventrally and medi-
ally, mostly crossing the midline rather far ventrally and at
varying rostrocaudal levels, after which many turn forward
and certainly come into relation with the large reticular cells
of this region (which in part lie in the relation de Lange,
'12, noted for the red nucleus in Testudo), though it has not
166 JEAN K. WESTON
been possible to find an unqestioned termination of such
fibers on these cells. Others seem to disappear in the general
reticular gray of the region. Some uncrossed fibers might
easily get into relation with the large reticular cells above
noted (figs. 21 and 22). Many of the fibers of this tract do
not swing so far laterafward, but occupy a more medial posi-
tion in the tepentum of the mid-brain at variable distances
from the ventricle, and in part (the more medially lying
fibers) come into relation with the I I I rd and the IVth nerve
nuclei, either homolateral or contralateral, as well as with
the tegmental gray. Some of them appear to pass into the
medial longitudinal fasciculus and probably run in that bundle
to the motor nuclei of this region or possibly to higher cen-
ters. This more media1 portion overlaps to some extent the
tractus cerebello-motorius mesencephali.
The tractus cerebello-motorius mesencephali (figs. 21 and
22) is quite small and consists of that part (largely unmedul-
lated) of the tractus cerebello-motorius et tegmentalis mesen-
cephali which courses medial to the mesencephalic root of
the Vth nerve (and in part through it), lying fairly close to
the ventricle, and runs medially, crossing high up in the
raphe, either to end directly on the nucleus of the IVth nerve
or to run forward, probably to the nucleus of the I I I rd nerve,
by way of the medial longitudinal fasciculus. I n part it
passes to the tegmental gray either crossed or uncrossed and
overlaps the more medial (and dorsal) portion of the tractus
cerebello-te,gmentalis mesencephali. This system is quite
small as compared to the tegmental component and is directly
continuous caudally with the tractus cerebello-motorius bul-
baris, which may also contribute to the nuclei of the I I I rd
and IVth nerves by way of the medial longitudinal fasciculus.
The origin of the tractus cerebello-motorius et tegmentalis
mesencephali seems to be largely homolateral, chiefly from
the nucleus lateralis cerebelli, though there is a considerable
addition, particularly to the tegmental portion, from the
nucleus medialis cerebefli as well as from the cerebellar plate,
probably from Purkinje cells. There is a strong possibility
of a small contralateral component, more particularly as re-
gards the cerebello-tegmental part.
The large reticular cells of the mid-brain (wherein has been
located a red nucleus) stand out exceptionally well in the
series of young Chelhydra. I n practically all cases where
an evident neuraxis could be observed, it crossed the midline
(about the middle of the raph6) almost immediately after
leaving the cell body. Such fibers then turned caudally and
became collected into small fascicles, just lateral and ventral
to the medial longitudinal fasciculus, which soon became lost
among the other longitudinally coursing paths in this re,' wion.
Such fibers probably constitute a rubro-spinal tract, a system
which has been assumed to exist among reptiles, but one
which needs further documentation.
The tractus cerebello-motorius et tegmentalis mesencephali
can be identified with little variation in all the forms studied,
varying in size with the size of the brain. I n the alligator
and the lizard (to some extent in the snake) the larger part
of the tegmental portion crosses rather more dorsally than
the above description would indicate. It is quite significant
that such facts agree well with the relative dorsoventral posi-
tion of the red nucleus as de Lange ('12) has located it in
these various forms.
The above system appears to have been included (in large
part at least) by Huber and Crosby ('26) under the term
tractus cerebello-tegmentalis mesencephali. They homolo-
gized this to Ariens Kappers' ( '21) brachium conjunctivum
anterius of Crocodile and to Larsell's ('23) brachium con-
junctivum of the frog. Larsell ('26) briefly noted a brachium
conjunctivum in lizards and snakes. Papez ( '29)' Shanklin
( '30), Hindenach ( '31)' and Frederikse ( '31) also noted a
similar tract as the brachium conjunctivum. Larsell ( '32)
observed such a tract both in the turtle and the alligator.
Tractus cerebello-motorius et tegmentalis bulbaris. Cours-
ing in the ventral and caudal portion of the cerebellar com-
missure, apparently in part contralateral in origin, a large
bundle of medullated fibers swings out of that commissure
168 J EAN K. WESTON
and turns ventrally, passing over the dorsal part of the
mesencephalic Vth root and spreading out into numerous
fascicles which course ventrally and caudally (certain of them
rostrally). Here these fascicles lie medial to the afferent
cerebellar systems and the lateral lemniscus, though laterally
they intermingle with them. Continuing this course, they
become lost among the other longitudinally coursing fibers
of the region and can be followed no farther than to about
the level of entrance of the VI I I th nerve, although they
probably do reach more caudal levels. A small portion of
the fibers seem to swing more medially and cross the raph6,
and the more medial part of this bundle certainly comes into
relation with the motor nucleus of the Vth nerve (figs. 16 to
18, 23 to 25).
This tract contains a considerable contralateral component
the origin of which could not be definitely determined. It
receives a large contribution from the homolateral nucleus
lateralis cerebelli and probably some from Purkinje cells.
A few questionable fibers appear to join it from nucleus
medialis cerebelli.
This system appears to be in part the efferent fibers lying
in the inferior cerebellar commissure noted by Ariens
Kappers ( '21). It is included in the cerebello-tegmental
system of Huber and Crosby ('26) and others. Larsell ('32)
verified the findings of Huber and Crosby ('26) to which he
apparently added the tractus cerebello-motorius of this
account, considering it the representative of Russell's fascicu-
lus in these forms.
The tractus cerebello-motorius bulbaris is that portion
which remains mostly medial to (in part passing through)
the mesencephalic Vth root as the tract leaves the cerebellar
commissure. It carries many thinly medullated fibers, par-
ticularly at its more caudal levels of origin. Coursing ventro-
medially, this tract turns more medially toward the raph6,
where it crosses just beneath and to some extent through the
medial longitudinal fasciculus (figs. 17, 18, 26), to which it
apparently contributes. It spreads into the tegmental regions
of the opposite side, but is a very large contributor to the
motor nucleus of the Vth nerve. Undoubtedly, as Huber and
Crosby ('26) suggest, it also places the other bulbar motor
nuclei under cerebellar control, in part through its connec-
tions by way of the medial longitudinal fasciculus.
This tract originates chiefly from the homolateral lateral
cerebellar nucleus. It receives a considerable number of
fibers which have crossed in the cerebellar commissure, and
a few direct fibers out of the cerebellar plate. More rostrally
it may receive a few fibers from the medial cerebellar nucleus.
The tract lies close to the ventricle throughout its extent,
but cannot be definitely followed any farther caudalward than
can the tegmental portion. It appears to correspond to the
tractus cerebello-motorius of Ariens Kappers ( '21) for
Crocodilia and to a part of the cerebello-tegmental system of
Huber and Crosby ('26) and of Larsell ( '32). It is probably
included, at least in part, when the term Russell's fasciculus
or the fasciculus uncinatus has been used for reptiles (Ariens
Kappers, '21 ; Papez, '29 ; Larsell, '32).
Tractus cere.bello-spinalis ( et cerebello-uestibzclaris). This
tract could be considered with the cerebello-tegmental system,
but it appears to answer most nearly to the description of
Russell's (uncinate or hook bundle) fasciculus in higher
forms, especially when considered together with the cerebello-
vestibular fibers coursing in company with the vestibulo-
cerebellar system.
The tractus cerebello-spinalis (et cerebello-vestibularis)
appears to be principally a contribution from the medial
cerebellar nucleus, receiving no additions from the homo-
lateral lateral cerebellar nucleus. I n addition, it carries a
large component from the cerebellar plate as well as fibers
which have crossed in the cerebellar commissure. These com-
ponents become grouped together at the more caudodorsal
part of the cerebellar commissure and immediately hook
around the lateral recess of the I Vth ventricle in the
caudomedial part of the cerebello-oblongata junction, lying
considerably dorsal to the main part of the tractus cerebello-
motorius bulbaris, and very close to the ventricle (fig. 24).
170 J EAN K. WESTON
At this point the tract lies in close juxtaposition to the in-
coming vestihulo-cerebellar tract and its cerebello-vestibular
component, but as it passes caudally, remaining near the
ventricle, it becomes grouped by itself as a noticeable bundle,
distinct even in transverse series (figs. 18 to 20, 24, 25). It
is best followed in a sagittal silver series of Chrysemys. I n
the vestibular region it lies near the ventricle, ventral to the
cochlear gray and dorsolateral to the prevagal component of
the fasciculus solitarius ; being in relation laterally with
ascending vestibular fibers, anteriorly, particularly with the
cochleo-vestibulo-cerebellar system ; and with descending ves-
tibular fibers, posteriorly (vestibulo-spinal tract). As it
reaches the caudal parts of the vestibular region, it loses its
compactness and, a few at a time, its fascicles become inter-
mingled with the vestibulo-spinal (and spino-vestibular ) path,
until about the level of entrance of the Xth nerve, when it
cannot be further distinguished from that tract. I n its course
through the vestibular region it is very obvious that fascicles
leave this tract and break up in the vestibular gray, and some
can even be followed farther ventrally into the reticular gray
of the region. I n passing through the vestibular region the
cerebello-spinal tract comes into particularly close relation
to the pars dorsalis of the nucleus vestibularis ventrolateralis,
and seems to receive a small vestibulo-spinal component from
that nucleus. It could not be traced separately into cord
levels, but there was a suggestion in the material that it did
reach such levels, hence the name.
This tract is most distinct in the turtle, although in none
of the other reptiles does it constitute B large component. I n
the smaller forms and in those with a marked eversion, it
does not form the compact bundle noted in the turtle.
From the literature it does not appear that this path has
received the consideration due it as a distinct efferent system.
Either it has been included under the terms cerebello-motorius
or cerebello-tegmentalis (Ariens Kappers, '21 ; Larsell, '32),
or it accompanies some ascending vestibular system (such as
the cochleo-vestibulo-cerebellar or the vestibulo-cerebellar
tract), or a descending vestibular system such as the
Deiterso-spinal tract (Shanklin, '30). I n the alligator, as a
result of the enormous cochlear development, this tract is
crowded somewhat ventrolaterally and easily confused with
the ascending vestibular systems in that region. I n snakes
it does not lie so near the ventricle as in turtles, and is con-
siderably mixed with the vestibulo-cerebellar systems due to
the small size of the cochlear region in these forms. (The
tract so labeled for the snake, fig. 25C, contains a large com-
ponent of vestibular fibers.)
Connections of t he cerebellar pl at e. At present the amount
of Golgi material available does not permit any added con-
tribution to the interconnections of the cerebellar plate com-
ponents in reptiles. Consequently no attempt will be made
here to consider such connections, though in this respect
reference might well be made to Larsell's ( '32) paper, where
a considerable contribution to the present knowledge of such
connections is contained. However, it is felt that a study of
pyridine-silver material of all the reptilian forms available
warrants the impression that a considerable component of all
of the efferent cerebellar paths is contributed by fibers, evi-
dently not commissural, passing out of the cerebellar plate,
which very likely are the neuraxes of Purkinje cells, though
this conclusion could not be satisfactorily verified.
SUMMARY AND DISCUSSION
So intimate a relation exists between the cerebellar and
vestibular regions in reptiles that an adequate analysis of
one region is dependent upon a thorough understanding of
the other.
The vestibular nuclear complex in reptilian forms may be
divided into the following nuclei : 1) the nucleus vestibularis
ventrolateralis (Deiters ') , the very obvious, large-celled,
highly developed nucleus situated at the level of entrance of
the VI I I th nerve ; 2) the nucleus tangentialis, a variably de-
veloped nuclear mass lying intercalated among the entering
vestibular fibers, external to the nucleus vestibularis ventro-
172 JEAN K. WESTOK
lateralis ; 3) the nucleus vestibularis ventromedialis, a small-
celled, poorly delimited cell group situated internal to the
nucleus vestibularis ventrolateralis, near the inferolateral
angle of the ventricle ; 4) the nucleus vestibularis descendens,
another small-celled mass, variably developed, and situated
most caudally in the vestibular region ; 5) the nucleus vestibu-
laris dorsolateralis, the small-celled, rostra1 continuation of
the nucleus vestibularis ventrolateralis, very intimately re-
lated rostrally to the cerebellar nuclei; 6) the nucleus vestibu-
laris superior, a differentiation of the last named nuclear
mass, obvious only in turtles. Associated with these nuclei
is the nucleus laminaris, which may be partly vestibular but
appears to be mainly cochlear in function. While all these
vestibular nuclei, excepting the nucleus vestibularis superior,
are present in all the reptilian forms, they present certain
interesting variations. Although well developed in all rep-
tiles, the nucleus vestibularis ventrolateralis reaches its high-
est development in the alligator, where it forms one of the
most conspicuous cell masses of the medulla oblongata. I n
all these forms its constituent neurons are of the large, multi-
polar, efferent type, with a rich dendritic network spreading
to all parts of the vestibular region. The largest neurons,
and relatively the smallest number of them, are found in the
snake. These two facts presumably are correlated with each
other and related to the type of locomotion found here, for
it seems probable that fewer larger neurons with more numer-
ous collaterals might well supply adequately a larger group
of the ventral horn neurons related to trunk muscles than of
those related to the functionally more specifically specialized
extremity muscles. The tangential nucleus is least developed
in turtles, best developed among lizards, and reaches an inter-
mediate stage of development in the snake and the alligator.
There seems to be an interesting correlation between the
relative development of this nucleus and the degree of trunk
movements encountered in these forms (p. 176). For an ade-
quate analysis of the nucleus vestibularis ventromedialis,
degeneration preparations are required, for while this nucleus
certainly receives vestibular impulses, it is also certain that
it receives fibers of passage, chiefly trigeminal, cochlear and
vagal. The degree of development of the consistently small-
celled nucleus vestibularis descendens is in inverse relation
to the development of the nucleus tangentialis, since it is
relatively largest in turtles and smallest in lizards. The
nucleus vestibularis dorsolateralis foreshadows the compara-
ble nuclear complex in avian forms, and, as is to be expected,
is best differentiated in lizards and crocodiles, but in no rep-
tile is it highly specialized. The nucleus vestibularis superior
is present only in turtles, which suggests that its development
is in inverse relation to the development of the nucleus ves-
tibularis dorsolateralis, and that it is included in this latter
mass as that grows larger.
The gross morphology of the cerebellum is of considerable
interest in reptiles since it affects to a marked degree the
relations of closely associated cell masses and fiber tracts.
The smaller cerebella are found among the small lizards and
snakes, though due to marked differences in shape it is diffi-
cult to evaluate properly the relative differences in cerebellar
size between these forms and to determine which of the forms
has the smaller. The Crocodilia possess the largest cerebella,
those of the turtles being intermediate in size. These facts
agree with Larsell's ('26 and '32) observation that the auricu-
lar lobe is least developed in the snake and much better de-
veloped in the reptiles which make use of extremities in looo-
motion, although this agreement in the observed facts does
not necessarily imply an agreement with his interpretation
of these facts. The cerebella of certain lizards show a
marked eversion, while indications of such an eversion are
found in other reptilian forms, particularly in Sphenodon,
and to a lesser extent among the crocodilians. The cerebel-
lum consists of a cerebellar cortex and a cerebellar nuclear
complex. Its nuclear complex is incompletely divided into
a larger-celled, more medial and a smaller-celled, more lateral
cell mass, the medial and lateral cerebellar nuclei, both of
which are directly continuous caudally with the vestibular
174 JEAN K. WESTON
gray. The differences in form relations of these nuclei are
related to the degree of eversion of the cerebellar plate, the
effect of such an eversion being to crowd the nuclei rostro-
caudally and slightly mediolaterally, and to spread them out
dorsoventrally, causing also a relatively greater intermingling
of the medial and lateral nuclei of the same side. The cere-
bellar cortex shows the usual division into a molecular, a
Purkinje, and a granule cell layer, the latter two, particu-
larly near the junction of the cerebellum with the medulla
oblongata, enclosing a variably developed stratum fibrosum.
The vestibular connections may be considered under four
headings ; entering vestibular root fibers, internuclear con-
nections, secondary afferent connections, and secondary effer-
ent connections. The root fibers of the vestibular nerve
distribute : 1) to the homolateral cerebellum terminating
chiefly in the auricle, and passing to the corpus cerebelli and
the cerebellar nuclei, as well as entering the cerebellar com-
missure in part, presumably to reach contralateral cerebellar
areas ; 2) to all of the homolateral vestibular nuclei ; 3) proba-
bly to certain of the contralateral vestibular nuclei by way
of decussating fibers at the level of entrance of the root;
4) to the spinal cord in company with the homolateral
vestibulo-spinal tract and possibly also in company with both
the medial longitudinal fasciculi. The proportionate distri-
bution to the nucleus vestibularis ventrolateralis varies with
the degree of development of the tangential nucleus. All the
vestibular nuclei appear to be interconnected by short inter-
nuclear fibers, and there is likewise a close commissural
internuclear connection relating them. The secondary aff er-
ent connections of the vestibular complex fall into two general
groups. To the first group belong those paths which termi-
nate within the nuclear complex either by way of pericellular
synapses or by synaptic relations with short, intranuclear
dendrites. To this group belong : 1) the cerebello-vestibular
paths, chiefly from the medial cerebellar nucleus to the
nucleus vestibularis ventrolateralis and the nucleus vestibu-
laris dorsolateralis ; 2) the spino-vestibular tract, which ends
primarily in relationship with the neurons of the nucleus
vestibularis descendens, but also has connections with the
nucleus tangentialis and the nucleus vestibularis ventrolater-
alis ; 3) certain secondary connections from various sensory
centers of the medulla oblongata, such as cochlear, trigeminal,
and vagal centers, to the nucleus vestibularis ventromedialis,
giving to this nucleus an indeterminate functional character.
The second group is formed by the spread of the dendrites
of the vestibular nuclei into neighboring nuclear masses or
along the course of various fiber paths and is well developed
in all the forms studied. These constitute a very important
source of excitation of certain of these nuclear groups in
reptiles. To this group belong the various dendritic ramifi-
cations of the nucleus vestibularis ventrolateralis into the
tegmental and cochlear regions and along the course of all
of the afferent paths to the cerebellum, and a similar, although
less widespread ramification of the dendrites of the nucleus
vestibularis dorsolateralis into the tegmental region and
along the afferent cerebellar paths. The major efferent paths
of the vestibular area consist of: 1) connections with the
medial longitudinal fasciculus, largely crossed and descend-
ing from the nucleus vestibularis ventrolateralis, the nucleus
tangentialis, and the nucleus vestibularis descendens, and
probably largely ascending and crossed from the dorsolateral
vestibular nucleus ; 2) vestibulo-spinal connections, chiefly
from the nucleus vestibularis ventrolateralis and possibly
added to by the nucleus vestibularis descendens ; 3) vestibulo-
cerebellar connections, largely, though possibly not exclu-
sively, from the dorsolateral vestibular nucleus.
I n all the forms studied there are two general types of
vestibulo-spinal connections in addition to the one passing by
way of the medial longitudinal fasciculus. The first of these,
a more dorsolateral vestibulo-spinal path, is developed in all
the forms. It lies in a dorsolateral position in the caudal
part of the vestibular region, and becomes largely inter-
mingled with the spino-vestibular path, these two together
being the vestibulo-spinal path referred to in the literature.
176 J EAN K. WESTON
The second is, strictly speaking, a vestibulo-spinal path only
in lizards. It lies slightly ventrolateral to the inferolateral
angle of the ventricle, at least near its place of origin from
the large Deiters cells. I n this position, in all the other
forms, there is a large number of vestibulo-reticular fibers
from Deiters nucleus (not so numerous in the snake), and
these may, after a synapse on reticular cells, reach the motor
centers of the cord as reticulo-spinal fibers. It is interesting
to relate these facts to the relative development of the ex-
tremities in reptilian forms. The lizard-particularly the
smaller lizard-has by far the most use (and the most rapid
movements) of its limbs. The snake, of course, has no limbs,
while the turtle is very clumsy in respect to limb movements
on land, such movements being relatively slow and studied.
The crocodilians resemble the turtle in this respect. From
this it is logical to assume that the vestibulo-reticulo-spinal
path may mediate the relatively slow limb responses made by
the turtle and the alligator on land, while in the lizard, with
the much more rapid locomotion in these forms, the reticular
phase of the path has been short-circuited, so to speak. Thus
an obviously more efficient and direct vestibulo-spinal path is
developed controlling these limb movements, probably more
truly homologous to the ventrolateral vestibulo-spinal path
of higher forms. It should be emphasized that the tangential
nucleus appears to be developed coordinately with the de-
velopment of the trunk musculature and contributes directly
and almost entirely to the medial vestibulo-spinal path which
runs as a component of the medial longitudinal f asciculus.
I n the lizard and the snake, where this nucleus is relatively
best developed, there is marked coordination of the body
musculature of the two sides in connection with locomotion.
I n the alligator the more sluggish movement is associated
with a relatively smaller development of this nucleus and its
descending tract, and in the turtle, where there is practically
no trunk movement, there is likewise practically no tangential
nucleus.
The cerebellar connections, whether afferent or efferent,
appear to be related largely to both the cerebellar cortex and
the underlying gray. Various observers have considered that
the neuraxes of Purkinje cells in mammals do not pass out-
side of the cerebellum, although there is not unanimous agree-
ment on this. I n reptiles there is an increasing body of
evidence, though it is not conclusive, that the neuraxes of
Purkinje cells do enter directly into the formation of the
efferent cerebellar paths (see also Larsell, '32).
The afferent cerebellar fiber tracts demonstrated in the
material were :
I . A dorsal spino-cerebellar. This tract is continued di-
rectly forward from the spinal cord along the periphery of
the medulla oblongata and swinging dorsal to the descending
Vth root to enter the cerebellum, where it terminates laterally
and caudally in the body of the cerebellum, and possibly to a
slight extent in the homolateral auricular lobe, also contribut-
ing to the cerebellar commissure.
2. A ventral spino-cerebellar tract. This bundle is accom-
panied by the foregoing tract along its dorsolateral border
throughout the bulb and continuing forward to the level of
the entering roots of the Vth nerve, where it turns dorsally,
curving rostral to (and partly through) the Vth nerve roots,
to pass to the homolateral medial cerebellar nucleus and, in
large part, into the cerebellar commissure to distribute chiefly
to the rostral portion of the contralateral part of the body
of the cerebellum.
This fiber system is com-
posed of various tracts relating sensory centers of the
medulla oblongata to the cerebellum, notably ventral super-
ficial arcuate fibers from the contralateral nucleus funiculus
posterior, a reticulo-cerebellar system from the general re-
ticular gray of the medulla oblongata (which includes any
olivo-cerebellar component that may be present), and a
trigemino-cerebellar system from the nucleus of the descend-
ing root of the Vth nerve of the opposite side-this whole
complex distributing to the lateral and caudal portions of
3. A Izucleo-cerebellar system.
THE JOURNAL OF COMPARATIVE NEUROMMY, VOL. 85
178 JEAN I{. WESTON
the corpus cerebelli and possibly in small part to the auricu-
lar lobe.
4. A uestibulo-cerebellar syst em. This system consists of
fibers arising chiefly in the nucleus vestibularis dorsolater-
alis, joined by fibers arising chiefly from the nucleus laminaris
(cochleo-vestibulo-cerebellar or lamino-cerebellar) , as well as
by some direct vestibular root fibers.
5. A trigemino-cerebellar system. This system is composed
of paths taking origin in the contralateral and homolateral
chief sensory nuclei of the Vth nerve, both paths passing to
the homolateral corpus cerebelli and into the region of the
commissure.
6, Direct root fibers of cranial qzerves. These consist
largely of fibers from the vestibular and trigeminal roots,
which in the turtle, at least, are supplemented by certain root
fibers running in the cochlear root, all of these passing chiefly
into the auricular lobe.
7. A tecto-cerebellar syst em. This system is a continuation
of the stratum album centrale of the optic tectum which
swings caudalward into the cerebellum through the region of
the anterior medullary velum, supplemented by fibers from
the inferior colliculus, and, in the turtle at least, possibly by
a few fibers from the nucleus isthmi. This system is partly
crossed and partly uncrossed in the cerebellar commissnre,
and can be traced in part to the homolateral medial cerebellar
nucleus, in part into the granular layer, and in small part
into the homolateral molecular layer of the cerebellum. This
connection from the optic tectum deserves special notice in
reptiles, since it appears to be more particularly developed
here, where the optic tectum plays so dominant a role, than
in subreptilian or even mammalian forms.
The efferent cerebellar fiber tracts demonstrated in the
available material were :
I. A tractus cerebello-spinalis. This tract is presumably
the reptilian equivalent of Russells fasciculus or the hooked
bundle, originating largely from both medial cerebellar nuclei,
supplemented by Purkinje cell axones, and immediately after
leaving its cells of origin hooking sharply around and under
the lateral recess of the I Vth ventricle. It runs caudally at
about that same level, lying very near the ventricle, at least
in turtles where it forms the most compact bundle. This
tract distributes to the nucleus vestibularis ventrolateralis
as well as to the other vestibular nuclei and tegmental gray
of the region, though to a lesser degree, and continues caud-
ally to become lost in the vestibule-spinal tract, probably
accompanying that tract into the spinal cord.
2. A tractus cere;bello-vestibzllaris. This tract has an origin
similar to the cerebello-spinal tract, together with which it
might be included as Russells fasciculus, and distributes to
the more medial part of the nucleus vestibularis dorsolater-
alis and to the nucleus vestibularis ventrolateralis throughout.
3 and 4. A tractus cerebello-motorius et tegme&& bulbaris
and a tractus cerebello-motoritis et tegmentalis mesemephali.
These tracts really form a continuous cerebello-tegmental
system of fibers, supplying the motor nuclei of the cranial
nerves (the tractus cerebello-motorius bulbaris and the
tractus cerebello-motorius mesencephali) and other tegmental
regions of the bulb and mid-brain (the tractus cerebello-
tegmentalis bulbaris and the tractus cerebello-tegmentalis
mesencephali). Of these the tract of particular interest from
a phylogenetic standpoint is the tractus cerebello-tegmentalis
mesencephali, which carries within it fascicles connecting the
cerebellum with the reptilian primordial red nucleus, and is
the forerunner, then, of the cerebello- or dento-rubro-thalamic
path of higher forms, which is to play a conspicuous part in
the development of forebrain centers.
This fiber complex shows a
significant differentiation into a superficial and a periven-
tricular portion. The superficial portion apparently is very
small, being a direct connection forward through the anterior
medullary velum to the superficial layers of the optic tectum.
The periventricular portion is surprisingly well developed ;
its fibers, apparently arising in part from Purkinje cells of
the cerebellum, come into close relation with the ventricle,
5. A cerebello-tectal system.
180 JEAN K. WESTON
coursing beneath the inferior colliculus (to which they may
contribute) to the periventricular fiber layer of the tectum,
where they break up around the dendrites of the cells of the
periventricular gray. There is considerable evidence from
the literature (Huber and Crosby, '26 and '33) that these are
joined by fibers from the hypothalamic region and from the
inferior colliculus, all of which distribute to the neurons con-
stituting the periventricular gray. Since such neurons,
through their long dendrites, are in relation to numerous
exteroceptive impulses entering the tectum near its periphery,
it is obvious that this periventricular bundle provides a means
of correlating proprioceptive impulses from the cerebellum
with exteroceptive impulses from other sources.
I n the arrangement of the paths to and from the cerebellum
there is a tendency toward a grouping which is comparable
to the formation of the superior and inferior cerebellar
peduncles of higher forms. The term superior peduncle or
brachium conjunctivum should not be limited to the cerebello-
rubral path, as is sometimes done, but should be used to in-
clude those systems to the tegmental regions and motor nuclei
of the upper cranial nerves which are also recognized con-
stituents of the brachium conjunctivum of mammalian forms.
We are not ready as yet to enter into a detailed comparison
of the reptilian vestibular and cerebellar regions with those
regions of lower or higher forms; yet some brief statement
of the relative development of these areas as compared with
those of amphibians and mammals may be of interest. Com-
pared with the reptilian cerebellum the amphibian cerebellum
(Herrick, '14, '24, '24 a; Larsell, '20, '23, '25, '31) is much
smaller and much more generalized in type. The cerebellar
nuclear gray is represented by a single cell mass which ex-
tends relatively far down into the peduncle. I n correlation
with its less highly differentiated cerebellar structure, the
amphibian vestibular area is smaller and shows less nuclear
differentiation; a large celled nucleus comparable to the rep-
tilian nucleus vestibularis ventrolateralis and usually other
nuclear gray nearer the cerebellar gray have been recognized.
The fiber connections of the amphibian and reptilian cere-
bellar and vestibular areas are comparable in type but are
smaller and show less specialization in the former animals.
The cerebellum and the cerebellar nuclei show a marked in-
crease in specialization in avian (Ingvar, '19 ; Sanders, '29)
as compared with reptilian forms. This is especially evident
with regard to the cerebellar nuclei, which are particularly
large and show marked cellular differentiation and character-
istic folding in birds. Again in correspondence with this
cerebellar development, is the increased specialization of the
vestibular nuclear group, particularly noticeable in the avian
region homologous to the nucleus vestibularis dorsolateralis
of reptiles. I n the present account of reptiles, however, it
has been possible to recognize in these forms all the major
nuclear divisions present in birds. There is difference of
opinion with regard to the extent to which the mammalian
cerebellar hemispheres find representation in the reptilian
cerebellum. Many observers agree that the auricular lobe is
the forerunner of the flocculus. Larsell ('26, '32) was of the
opinion that a portion of the so-called neocerebellum, as
represented by the lobulus ansiformis, found representation
in his pars lateralis of the reptilian cerebellum. Ingvar ('19)
considered this portion of the cerebellum to make its first
appearance in mammals. It is felt that further evidence is
necessary before judgment can be passed regarding the inter-
pretation of Larsell in this respect. Perhaps the most that
can be said in an.attempted comparison of reptilian and mam-
malian cerebellar nuclei is that these nuclei in reptiles are
differentiating toward the nuclear configuration seen in mam-
malian forms, but that such differentiation is neither struc-
turally nor functionally complete. A comparison of the
vestibular area of reptiles and mammals is a matter of some
difficulty. It is probable that in a general way the reptilian
nucleus vestibularis ventrolateralis, the nucleus vestibularis
descendens, and the nucleus tangentialis are homologous to
the similarly situated vestibular nuclei of mammalian forms.
Whether either or both the ventromedial vestibular nucleus
182 JEAN K. WESTON
or the dorsolateral vestibular nucleus are homologous to the
mammalian medial vestibular nucleus, or merely substitute
nuclear groups, it is impossible to state at present. From
position and general relations it seems that the superior
vestibular nucleus of turtles marks the beginning of a mam-
malian superior vestibular nucleus.
The foregoing description of the reptilian vestibular and
cerebellar systems and the resume of its nuclear masses and
fiber connections illustrate the similarity of pattern which
underlies the diverse differentiation within the nervous
system, and emphasize again the possibility of understanding
the more specialized vertebrate nervous systems through a
study of the relations and connections in the simpler forms.
LITERATURE CITED
All references quoted i n the text which are not included i n the following
bibliography may be found in the bibliographies at the ends of the appropriate
chapters of the first reference listed below.
ARIENS KAFTERS, C. u., G. CARL HUBER AND ELIZABETH c. CROSBY 1936 The
comparative anatomy of the nervous system of vertebrates, including
man.
BANCHI, A. 1903 Sulle vie di connessione del cervelletto. Arch. ital. di anat.
e di embriol., vol. 2, p. 426.
BECCARX. N. La constituzione, i nuclei terminali e le vie di connessione
del nervo acustico nella Lacerta muralis, Merr. Arch. ital. di anat.
e di embriol., vol. 10, p. 646.
DITMARS, R. L. 1910 Reptiles of the world. Xturgis and Walton Co., New
York.
FREDERIKSE, A. 1931 The lizards brain: An investigation of the histological
structure of the brain of Lacerta vivipara. C. C. Callenbaeh, Nijkerk
(Holland).
GADOW, H. 1901 Amphibia and reptiles. Cambridge Natural History. The
Macmillan Company, New York.
HINDENACH, J. C. R. 1931 The cerebellum of Sphenodon punctatum. J. Anat.,
vol. 65, p. 283.
HOLMES, G. M. 1903 On the comparative anatomy of the nervus acusticus.
Tr. Roy. Irish Acad., vol. 32, sec. B, p. 101.
HUBER, G. CARL AND ELIZABETH C. CROSBY On thalamic and tectal nuclei
and fiber paths in the brain of the American alligator. J. Comp.
Neur., vol. 40, p. 97.
Das Hinterhirn, das Nachhirn, and das Riickenmark der
Reptilien. 301. Neurobiol., Bd. 10, S. 385.
The Macmillan Company, New York.
1912
1926
DE LANGE, S. J. 1916
LARSELL, 0. 1926 The cerebellum of reptiles; lizards and snake. J. Comp.
1932 The cerebellum of reptiles: chelonians and alligator. J.
REESE, A. M. 1908 The development of the American alligator. Smith. Misc.
1910 The development of the brain of the American alligator.
Smith. Misc. Coll., no. 1922, vol. 54,
Neur., vol. 41, p. 59.
Comp. Neur., vol. 56, p. 299.
Coll., no. 1791, Washington, D. C.
The paraphysis and hypophysis.
Washington, D. C.
1915 The alligator and its allies. Putnam, New York.
VAN HOEVEI~L, J. L. D. 1916 The phylogenetic development of the cerebellar
nuclei. KGn. Akad. v. Wetenseh. te Amsterdam, Proc. see. so., vol. 18,
p. 1431.
LI ST OF ABBREVIATIONS
a, commissural fibers of nucleus isthmii
a.coch., area eochlearis
ang.lat.v.IV, angulus lateralis ventrieuli
aq.S., aqueductus Sylvii
b, decussation of tractus eerebello-
motorius bulbaris
c, afferent cerebellar fiber systems
cell.Purk., Purkinje cells
col.inf ., eolliculus inferior
comxer., eommissura eerebelli
d, efferent cerebellar fibers (in large
dec.r.mo.N.V, decussatio rami motorii
dec.N.V, decussatio nervi trigemini
dec.tr.eer.teg.mes., decussatio tractus
cerebello-tegmentalis mesencephali
e, vestibulo-reticular fibers
f, mixed nueleo-cerebellar, dorso-spino-
cerebellar, vestibulo-cerebellar, and
direct V and VI I I fibers
fascsol., faseieulus solitarius
fib.arc.dors., fibrae arcuatae dorsales
fib.arc.vent.superf., fibrae arcuatae ven-
fibxoch., fibrae cochleares
fib.coch.sec., fibrae eochleares secundae
fib.com., fibrae commissurales
fib.perivent., fibrae periventriculares
f .I.m., fasciculus longitudinalis medi-
g, extension of ventral horn gray
gang.vest., ganglion vestibulare
quarti
part at least)
nervi trigemini
trales superficiales
alis
gr.perivent., griseum periventriculare
gr.ret., griseum reticulare
h, vestibulo-reticular fibers
junct.cer.ob1, junctura cerebello-oblon-
1, canalis centralis of spinal cord
Lam., lobus aurieularis
l edat., lemniscus lateralis
lemtrig., lemniscus trigeminalis
mes., mesencephalon
N.IV, nervus trochlearis
N.V, nervus trigeminus
N.VI, nervus abducens
N.VII1, nervus aeustieus
N.IX, nervus glossopharyngeus
N.X, nervus vagus
n.111, nucleus oculomotorius
n.IV, nucleus trochlearis
n.VI, nucleus abducens
n.eoch., nucleus eochlearis
n.coeh.dors., nucleus cochlearis dorsalis
n.dors.magnocell., nucleus dorsalis mag-
n.isth.p.magnocell., nucleus isthmii pars
n.isth.p.parvocell., nucleus isthmii pars
darn., nucleus laminaris
n.lam.p.dors., nucleus laminaris pars
n.lnt.cer., nucleus lateralis eerebelli
n.med.cer., nucleus medialis cerebelli
n.mo.N.V, nucleus motorius nervi tri-
gata
nocellularis
magnocellularis
parvocellularis
dorsalis
gemini
184 JEAN K. WESTON
n.mo.N.VI1, nucleus motorius nervi
n.precer., nucleus precerebellaris
n.r.d.N.V, nucleus radicis descendentis
n.sens.prin.N.V, nucleus sensibilis prin-
n.tang., nucleus tangentialis
n.vest.d., nucleus vestibularis descend-
n.vest.dors.lat., nucleus vestibularis dor-
n.vest.sup., nucleus vestibularis superior
n.vest.vent.lat., nucleus vestibularis ven-
trolateralis
n.vest.vent.lat.p.dors., nucleus vestibu-
laris ventrolateralis pars dorsalis
n.vest.vent.lat.p.lat., nucleus vestibularis
ventrolateralis pars lateralis
n.vest.vent.lat.p.vent., nucleus vestibu-
laris ventrolateralis pars ventralis
n.vest.vent.med., nucleus vestibularis
ven tromedialis
olmp., oliva superior
pl.ch., plexus chorioideus
r.N.V, radix nervi trigemini
r.N.VI, radix nervi abducentis
r.N.IX, radix nervi glossopharyngei
r.N.X, radix nervi vagi
r.dors.N.VII1, radix dorsalis nervi
r.mes.N.V, radix mesencephalica nervi
r.mo.N.VI1, radix motoria nervi facialis
r.mo.N.V, radix motoria nervi trigemini
r.sens.N.V, radix sensibilis nervi tri-
r.sens.N.VI1, radii sensibilis nervi
r.N.trig.ad cer., radix nervi trigemini
r.vent.N.VII1, radix ventralis nervi
r.asc.N.vest., ramus ascendens nervi
r.N.coch.ad cer., ramus nervi cochlearis
r.d.N.V, ramus descendens nervi tri-
facialis
nervi trigemini
cipalis nervi trigemini
ens
solateralis
acustici
trigemini
gemini
f acialis
ad cerebellum
acustici
vestibularis
ad cerebellum
gemini
r.d.N.vest., ramus descendens nervi
r.N.fac.ad cer., ramus nervi facialis ad
rec.lat., recessus lateralis
str.alb.cent., stratum album centrale
&.fib., stratum fibrosum
str.gran., stratum granulosum
str.mol., stratum moleculare
str.Purk., stratum Purkinje
tect.op., teeturn opticum
tr.cer.mo.bulb., tractus cerebello-motori-
us bulbaris
tr.cer.mo.et teg.mes., tractus cerebello-
motorius et tegmentalis mesencephali
tr.cer.mo.mes., tractus cerebello-motorius
mesencephali
tr.cer.teg.bulb., traetus cerebello-teg-
mentalis bulbaris
tr.cer.teg.mes., tractus cerebello-teg-
mentalis mesencephali
tr.cer.spin., tractus cerebello-spinalis
tr.cer.spin. (et tr.cer.vest.), tractus cere-
bello-spinalis (et tractus cerebello-
vestibularis)
tr.cer.vest., tractus cerebello-vestibularis
tr.nuc.cer., tractus nucleo-cerebellaris
tr.spin.cer.dors., tractus spino-cerebel-
lark dorsalis
tr.spin.cer.dors.et tr.nuc.cer., tractus spi-
no-cerebellaris dorsalis et tractus
nucleo-cerebellaris
tr.spin.cer.vent., tractus spino-cerebel-
lark ventralis
tr.tect.cer., tractus tecto-cerebellaris
tr.trig.cer.cruc., tractus trigemino-cere-
bellaris cruciatus
tr.trig.cer.reot., tractus trigemino-cere-
bellaris rectus
tr.vest.cer. (et cer.vest.), tractus ves-
tibulo-cerebellaris (et cerebello-ves-
tibularis )
vestibularis
cerebellum
tr.vest.spin., tractus vestibulo-spinalis
tr.vest.spin. (et spin.vest.) , tractus ves-
tibulo-spinalis (et spino-vestibularis)
v.IV, ventriculus quartus
v.IV (ang.inf J at.), ventriculus quartus
vel.med.ant., velum medullare anterius
(angulus inferolateralis)
VESTIBULAR AND CEREBELLAR SYSTEMS
L
185
Fig. 1 Transverse section through the cerebellum at the rostralmost level of
the cerebellar nuclei. Chrysemys marginata. Toluidh blue preparation. X 10.
Fig. 2 Transverse section caudal to preyious figure. Chrysemys marginata.
Toluidin blue preparation. X 10.
Fig. 3 Transverse section just rostra1 to the entrance of the VIIIth nerve.
Uhrysemps marginata. Toluidin blue preparation. X 10.
186 JEAN K. WESTON
4
5
Fig.4 Transverse section through the level of entrance of the VI I I th nerve.
Note the cell grouping as well as the scattered small cells lying intercalated
among the fibers of the entering VI I I th nerve. These latter are not labeled
but are probably in part the representative here of the nucleus tangentialis.
Chrysemys marginata. Toluidin blue preparation. x 10.
Transverse section just caudal to the entrance of the VI I I th nerve.
This section is taken about at the transition level where the nucleus vestibularis
ventrolateralis i s passing over caudally into the nucleus vestibularis descendens
and where the nucleus laminaris is passing over into the nucleus dorsalis magno-
cellularis. Note the obvious Deiters cell mixed with the more medium sized
ones of the nucleus vestibular& descendens. Chrysemys maginata. Toluidin blue
preparation. x 10.
Fig.5
187
Fig. 6 Sagittal section just lateral. to the lateral angle of the I Vth ventricle
showing quite well the relations of the cerebellar and vestibular nuclei to each
other. Pseudoemys elegans. Toluidin blue preparation. X 10.
This shows
particularly the relations in the vestibular region. Pseudoemys elegans. Toluidin
blue preparation. X 10.
Fig. 7 Sagittal section farther medialward than the preceding.
188 JEAN I(. WESTON
8
9
11
Fig. 8 Transverse section through the cerebellar region showing the cerebellar
Fig. 9 Transverse section through the nucleus vestibularis ventrolateralis.
Fig. 10 Transverse section through the caudalmost portion of the entering
nuclei. Holbrookia. Toluidin blue preparation. X 20.
Holbrookia. Toluidin blue preparation. x 20.
roots of the VIIIth nerve. Holbrookia. Toluidin blue preparation. x 20.
11
189
Fig. 11 Transverse section through the cerebellar region and the cerebellar
nuclei of Thamnophis sirtalis. Note the resemblance of this figure to figure 15
of Larsells ( 32) account of the turtle. Note also how relatively slight is the
separation between the cerebellar nuclei and between the lateral cerebellar nucleus
and the nucleus vestibularis dorsolateralis. Toluidin blue preparation. -X 20.
Fig. 12 Transverse section through the vestibular region of Thamnophis
sirtalis. Note particularly the small development of the cochlear region, the
large nucleus tangential&, and the relatively large size of the cells of the nucleus
vestibularis ventrolateralis. Toluidin blue preparation. X 20.
190 JEAN X. WESTON
Fig. 13 Transverse section through the cerebellar nuclei of Alligator mississip-
piensis. Compare this with the figures of van Hoevell ( '16). Note the scattered
appearance of the nucleus medialis cerebelli-a result of the tendency toward
eversion of the anterior third of the cerebelluin here-and compare with the
figures of the turtle. Toluidin blue preparation. X 10.
Fig. 14 Transverse section through the vestibular region of Alligator mississip-
piensis slightly rostra1 to the entrance of the roots of the VI I I th nerve. Note
particularly the marked development of the nucleus vestibularia ventrolateralis
and the subgrouping of the cells of the same. Observe also how the much larger
cochlear region has increased at the expense of the I Vth ventricle. Toluidin
blue preparation. x 10.
192 JEAN K. WESTOM
Fig. 15 Transverse section through the eaudalmost level of entrance of the
VIIIth roots in Alligator mississippiensis. Note the relative development of
the nucleus tangentialis as compared with that of the lizard. Toluidin blue prepara-
tion. X 10.
VESTIBULAE AND CEREBELLAR SYSTEMS 193
Fig. 16 Transverse section through the cerebellar commissure. Chrysemys
marginata. Pyridine-silver preparation. X 10.
Fig. 17 Transverse section through the brain stem just caudal to the cere-
bellar eommissure. Chrysemys marginata. Pyridine-silver preparation. x 10.
Pig. 18 Transverse section through the brain stem at the caudal level of the
entrance of the roots of the Vth nerve. Note the motor root of the Vth nerve
quite apparently crossing the midline. Chrysemys marginata. Pyridine-silver
preparation. X 10.
TEE JOWBNAL OF O O M P m T m NEUROLOGY, VOL. 65
194 JEAN K. WESTON
Fig.19 Transverse section of the brain stem through the caudal levels of
entrance of the VI I I th nerve and through the emergence of the root fibers of the
VI th nerve. Chrysemys marginata. Pyridine-silver preparation. X 10.
Transverse section of the brain stem through the level of entrance
of the I Xth nerve. Note particularly the ventral superficial arcuates and the
vestibulo-spinal tract. Chrysemys marginata. Pyridine-silver preparation. X lo.
Fig.20
195
n
Fig.21 Horizontal section cut i n an oblique plane so that the rostra1 end is
at a considerably more ventral level than the caudal end. Note particularly the
direct ascending cochlear fibers to the cerebellum and how far laterally they
swing to pass by the lateral recess of the I Vth ventricle. Note also the course
and relations of the tractus cerebello-motorius et tegmentalis mesencephali.
Chelhydra serpentha (3 days after hatching). Pyridhe-silver preparation. X 20.
Horizontal section from same series as figure 21, but from a more
ventral level. Note the ascending and descending direct VI I I th fibers as well
as the bifurcating ones. Observe al so the course of the sensory VI I th root.
Chelhydra serpentina (3 days after hatching). Pyridine-silver preparation.
x 20.
Fig.22
196
JEAN X. WESTON
herKV;,:: ; /J ihtcer.
24 nuf/ f ti.m.mall
b.cer.mb& hmNV
Fig.23
Sagittal seetion through the brain stem of Chrysemys marginata at
approximately the same level as figure 6. Note particularly the relations of the
efferent cerebellar paths. Pyridine-silver preparation. X 10.
Sagittal section through the brain stem of Chrysrmys marginata
farther medial and through the vestibular region not far lateral to the ependpal
lining of the I Vth ventricle. Note particularly the relations of the efferent
cerebellar paths, particularly the tractus eerebello-spinalis (et eerebello-vestibu-
laris). Pyridine-silver preparation. X 10.
Fig. 24
Fig.25 Transverse Weigert sections through the level of entrance of the
VI I I th nerve. Observe the various relations of the roots of both the VI I I th
and the VI I th nerves. Note also as far as possible the secondary vestibular
connections illustrated. A, Alligator mississippiensis. X 10. B, Anolis earo-
inensis. Note tr.vest.spin. X 20. C, Natrix. Here the fibers labeled tr.cer.spin.
(et tr.cer.vest.) contain large numbers of secondary vestibular fibers, x 20.
D, Ghrysemys marginata. The heavy black curved line represents a blood vessel.
x 20.
198 J EAN K. WESTON
Fig. 27 Sagittal sections (except G and H, horizontal) through the region of
the anterior medullary velum to show the tecto-cerebellar system. A and B,
Alligator mississippiensis. X 10. C and E, Natrix. X 20. D, Sternotherns.
x 20. F, Horned toad. X 20. G, Alligator mississippiensis. X 20. H, Chrysemys
marginata. X 20.

Laboratory University of Michigan

Uploaded by

Copyright:

Available Formats

Laboratory University of Michigan

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Laboratory University of Michigan

Uploaded by

Copyright:

Available Formats

THE REPTILIAN VESTIBULAR AND CEREBELLAR

GRAY WITH FI BER CONNECTIONS

You might also like