— ~ ee ee
2 a ae. == OA OA, - = .
= * Te FO Pixp
> oN a a eS
Et I ors: Oe er
SS,
Ms ode eS
Ce ws
58
oe
ess
e2.
pe.
Sv & Se 4/
Piette, Se ‘S
eae:
etae
-
AS
ee
Wet ey
ol
«
$4
‘ ~~ f
a os
*~
oe
f
a3
aw A
ey
wee
<7
tthe:
<
’
+
athe
Crane:
tats
eaters
fetes eo
were
@.
®
4
2
*
bee
*
f%,
%
Phas
4
ve oe,
i
*
94%,
Sok
+
48.
$a4ney:
1% oF
gate
+59,
WK
*
*
.
a.
TNS
v.46 3
tte 7”
* A
a ss fect gee bigt in aet
Tepito dierts Sats athe Get ‘ $'5
sO Pee . ‘ 4 nattiys hid}
Sua ia ane ANA A
THE JOURNAL
OF
COMPARATIVE NEUROLOGY -
EDITORIAL BOARD
Henry H. Donatpson ApDoLF MEYER
- The Wistar Institute : Johns Hopkins University
J. B. JoHNsTon Ouiver S. Strona
University of Minnesota Columbia University
C. Jupson HERRICK, University of Chicago
Managing Editor
THIS VOLUME IS DEDICATED TO
PROFESSOR CAMILLO GOLGI
VOLUME 30
DECEMBER, 1918-AUGUST, 1919
PHILADELPHIA, PA.
THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
CONTENTS
No. 1. DECEMBER, 1918
Otor LARsSELL. Studies on the nervus terminalis: Mammals. Forty-nine figures....... 1
Epwarp Puetrs Aus, Jr. The ophthalmic nerves of the gnathostome fishes......... 69
D. A. Rutnenart. The nervus facialis of the albino mouse. Fourteen figures......... 81
Krvoyasu Marur. On the finer structure of the synapse of the Mauthner Cell with
especial consideration of the ‘Golgi-net’ of Bethe, nervous terminal feet and the
‘nervous pericellular terminal net’ of Held. Fifteen figures....................... 127
No. 2. FEBRUARY, 1919
FRONTISPIECE. Portrait of Professor Camillo Golgi —
Wiuiiam F. Auten. Application of the Marchi method to the study of the radix
mesencephalica trigemini in the guinea-pig. Thirty-five figures.................... 169
Hovey Jorpan. Concerning Reissner’s fiber in teleosts. Ten figures.................. 217
Rosert 8. Exuis. A preliminary quantitative study of the Purkinje cells in normal,
subnormal, and senescent human cerebella, with some notes on functional locali-
maui s LWwOtmures Hn ONG Charten'.:). i... ., Va.kr: ka eeed eee ye uke os. b ae eee 229
None APRS
Kiyoyasu Marutr. The effect of over-activity on the morphological structure of the
ene sa, MUUT COR tim Uren i 2), ). 2 at) SV 2s tues 4 Aalgtigisls « s. « Smuaye eye abby cates eee 253
O. VAN DER Srricut. The development of the pillar cells, tunnel space, and Nuel’s
spaces in the organ of Corti. Eighteen figures............ “ft, SANG. SiS cE, «eee 283
No. 4. JUNE
Howarp Ayers. Vertebrate cephalogenesis. IV. Trnasformation of the anterior end
of the head, resulting in the formation of the ‘nose.’ Twenty-six figures........... 323
Lesiiz B. Arry. A retinal mechanism of efficient vision. Two text figures............ 343
D. OGatTa AND SwALE VINCENT. A contribution to the study of vasomotor reflexes.
RUT TEES. 1.5 aa Vax os AS x Siete Boze os ser nase hiec Soe McaN 68. 17s + eS RBA Fg Ae we a
No. 5. AUGUST
Suicreyvuk1 Komine. Metabolic activity of the nervous system. III. On the amount
of non-protein nitrogen in the brain of albino rats during twenty-four hours after
SCANT y's, AE a Rese IRLT ey ere SPS, Ba re A oa oP cake. Us NIRS oF HOE peg STATS 397
JAMES Stuart Puant. Factors influencing the behavior of the brain of the albino rat
are TeR LLG? Sa eh, Name. ONY Ae ar Ge ooh. Semen Sines ny at ale eB Stone e Sicane 411
O. LarsELL. Studies on the nervus terminalis: Turtle. Sixteen figures............... 423
C. G. MacArruur anv E. A. Dotsy. Quantitative chemical changes in the human brain
rime ree ey Cli,” sk RIMM Shay", Facial et suse cscs vip Piso eaten dean ea lew eed Wa vk oan o ORS 445
ie he tet ee
yr f wif a Sn
lodges ia
oan att *
Ayia we
>
' F
‘
i (Pode a he”
TO
CAMILLO GOLGI
PROFESSOR OF PATHOLOGY AT THE UNIVERSITY OF PAVIA, SAVANT AND CITIZEN,
GUARDIAN OF PUBLIC HEALTH AND STUDENT OF HISTOLOGY, TO WHOM
THE WORLD IS INDEBTED FOR A METHOD WHICH GAVE A DEEPER
INSIGHT INTO THE ARCHITECTURE OF THE CENTRAL NERVOUS
SYSTEM—THIS VOLUME OF THE JOURNAL OF COMPARATIVE
NEUROLOGY IS DEDICATED AS A TOKEN OF HIGH
ESTEEM FOR THE SCIENTIST AND THE MAN
WHO, HONORED AND FULL OF YEARS,
NOW WITHDRAWS FROM
HIS PROFESSORIAL
RESPONSIBILITIES
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, No. 1
DECEMBER, 1918
Resumido por el autor, Olof Larsell.
Estudios sobre el nervio terminal. Mamfferos.
El autor describe con algtin detalle el nervio terminal en el
gato, buey y mulo,,dando también descripciones mds breves de
dicho nervio en el caballo, perro, ardilla y en el hombre. En el
buey, mulo y ardilla se describe este nervio por primera vez.
En los mamiferos esta’ formado principalmente por fibras del
simpatico; las del fasciculo principal del nervio tienen con las
que se distribuyen periféricamente una relacién semejante a la
que existe entre las fibras preganglionares y postganglionares.
Esta semejanza esta aumentada por la estructura de los racimos
ganglionares y por la presencia de cestas pericelulares en muchas
de las células ganglionares. También existen redes intercelu-
lares y células tfipicas del simpatico. Por la arteria cerebral
anterior y susramas y por el plexo vascular del tabique nasal se
distribuyen fascfculos de fibras mielinicas y amielfnicas proce-
dentes del tronco principal y ganglios. Estas fibras terminan
en las paredes de los vasos sanguineos por medio de termina-
ciones nerviosas sensitivas y motrices. Hay algunas pruebas de
que las terminaciones nerviosas libres en el epitelio nasal estan
relacionadas con el nervio terminal, pero la presencia indudable
de fibras del trigémino en el plexo del tabique, junto con las pro-
cedentes del terminal, no permite una afirmacién rotunda sin
previo trabajo experimental. Estas terminaciones, Junto con las
de ciertas células ganglionares parecen indicar la existencia de
un componente sensitivo en el mervio, distinto de las fibras
aferentes del simpdtico. Parece claro que la inervacién del
érgano de Jacobson por parte del nervio terminal es incidental y
secundaria. El autor expone la posibilidad de que el nervio
terminal represente una divisién del sistema del simpdtico, rela-
cionada con el cerebro anterior.
Translation by Dr. José Nonidez,
Columbia University
AUTHOR’S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 9
STUDIES ON THE NERVUS TERMINALIS: MAMMALS!
OLOF LARSELL
Department of Anatomy, University of Wisconsin
FORTY-NINE FIGURES
CONTENTS
If. JEN ys IGE) ae ia eae ieee ees ey eho Shey 6. riryic’'3'eecit tact che Bea Ree Rote cea bey d Ge acer 3
SIS GOMES SINC TLATNL IT 2 222 aye eee ote te gees elo, hs Or eke 8 sie es hate oa? 3
Conditions in the different classes of vertebrates..................... 5
APSE NP UEV GE UIACU 7510's: 2 cha cyank bs a. 4.3 3 ace, se uaeie oars Lames at la hts Rey Us 12
MeiienlaleanGemevhod ses. *.42:.7 casero Cee eee ere oe ee eter. 12
iL, Aad Tvsanvsy iGiaemhar bE) Oru CMtno ann onas oacosdassocntescusgcaavor 16
istolo orca lie | techs cls oeh ool hide ce en Se eee a ome 26
iynes at.gamg lion Celle... s25:.. xy. 18 tee bleed ake eee ede ENS oi 26
Kiberstand stibernnebwOrksic.30-,.... 0tadtonsee teeter eee ae 33
INET Ve RU CIEL AULOMS foes sccce cio sie Bare aeicce © oe MMe ts Bete 37
Hehe menvus terminalis Of; thie bee. .y 1... 55 scl: «ss » pele cee elas r= 43
uaa sical s seme tte. BASS OU 28h Re ee ee 48
Sin chuLe of menverOundlesiq cots ccios os (eee eee 48
INGIVcaLeLminatlOnsiyes piycinck ies cake acs» seein ae ee ee 50
3. The nervus terminalis of the mule and the horse.................. Bl
AU oVevaen0ll (ein Nae eae s e GA SoA can NES Go, ee A A Se 8 oe c 51
LIST OLO OTC AIM nahn. arth ates et Soa ora io > SA Es ee 55
TRIDENT EL aoa le eee Re hac ee ee a Rey. ae 57
4. The nervus terminalis of the dog, the squirrel, the human, and of
embpryosrof pig, sheep, andirabpit fica: < $2 4. Veen eee tle 61
Se eer LEMIAA ATV CATEG COTITMA CTI Bi (cc x os) pci e/si yak cvaaal gl wale «e's Sieldbd a Pabede had oe BM 62
hs TEI OUT CHATS 01620 eS ec ie denM IR ERIE ok eu Ue a nM PRs ely ae Pr 64
>
I. INTRODUCTION
Discovery and naming. The cerebral nerve now known as the
nervus terminalis first began to attract the attention of morphol-
ogists in 1894. In that year Pinkus described in the dipnoan
fish, Protopterus, a hitherto unrecognized nerve of the forebrain.
1 Contribution from the Zoological Laboratory of Northwestern University,
William A. Locy, Director.
4 OLOF LARSELL
This nerve had previously been figured by Fritsch in the sela-
chian, Galeus, in 1878, and had been mentioned in 1893, by C. L.
Herrick in the urodele amphibian, Necturus. The observations
of Fritsch and Herrick, however, were merely incidental, in con-
nection with other work, and the significance of the observations —
escaped them.
After these anticipatory glimpses the nerve remained unno-
ticed until Pinkus described it in Protopterus in 1894. In 1895,
in a more extended paper, Pinkus figured and described it as ly-
ing ventral to the olfactory nerve and extending caudad over
the ventral surface of the forebrain to the recessus praeopticus,
_its peripheral terminations being in the olfactory sac. Thus,
although first figured in selachians, it was first described with
sketches, in the Dipnoi.
Shortly afterward, Allis (97) described and figured in the
ganoid fish, Amia, a strand which he traced centrally to the
bulbus olfactorius and peripherally to the olfactory capsule.
He considered this strand to be homologous with the nerve de-
scribed by Pinkus. In his comments regarding the possible
function of the nerve, he suggested that it might be of sympathetic
type. This is interesting in view of the position taken by Brook-
over and others and of the demonstration of the presence of
sympathetic fibers in the bundle of the nervus terminalis of
mammals.
Now appeared the first study of the embryological history of
the nerve (Locy, ’99) together with a description of its adult
condition in Squalus acanthias. In this form the nerve was
described and figured as possessing a compact ganglion, as con-
nected with the brain in the fissure between the lobes of the tel-
encephalon, and as distributed anteriorly to the lateral part of
the olfactory capsule and entering between the folds of the nasal
epithelium. It was claimed that the nerve arises from the neu-
ral crest before the appearance of the olfactory fibers. At that
time Locy considered as doubtful its homology with the nerve
described by Pinkus, and provisionally designated it as a median
‘accessory olfactory strand.’ In 1903, after observing the same
nerve in six genera of selachians, Locy reversed his earlier opin-
NERVUS TERMINALIS: MAMMALS 5
ion and concluded that the nerve of selachians is homologous with
that described by Pinkus in Protopterus and later by Allis in
Amia. The same author in his paper of 1905, based on the ex-.
amination of twenty-seven species of adult selachians and the
puibsveluees history of the nerve in Squalus, proposed the name
of ‘nervus terminalis’ for this new nerve. This name has been
generally adopted.
In the interval Sewertzoff oy had described the nerve in
embryos of Ceratodus, finding it ganglionated and terminating
peripherally in the mucous membrane of the anterior nasal
chamber—not in the sensory epithelium. On account of its
point of central connection with the brain he suggested for it
the name of ‘nervus praeopticus.’ This nerve had also been
cited in the adult of Ceratodus by K. Firbringer in 1904, as
well as by Bing and Burckhardt in 1904 and 1905.
Conditions in the different classes of vertebrates. Since 1905 a
considerable literature has accumulated regarding the nervus
terminalis. Its presence.has been demonstrated in all cases of
vertebrates except the cyclostomes (and possibly the birds)
and various suggestions regarding its function have been made
from time to time.
Inasmuch as the present paper deals with the nervus terminalis
chiefly in mammals, it is not necessary to enter into a review of
the rather extensive literature of the nerve in the lower verte-
brates, But, since the nerve presents some modifications and
some erdries: in the mammals, it is advantageous both for
comparison and for discussion of results, to have a brief state-
ment of the chief structural features which have been observed
in other classes of the phylum. ‘The nervus terminalis appears
to be more generalized in the fishes, especially in the selachians,
and in discussing its relations in mammals it would be a mistake
to disregard the findings in the lower vertebrates.
Fishes: a) Selachians. In a paper on the telencephalon of the
selachians, Johnston (’11) shows that typically the nervus ter-
minalis enters the brain substance near the recessus neuroporicus
internus. He remarks: ‘“‘Some evidence has appeared recently
(Burckhardt, ’07, p. 340; Brookover, ’10) that the nervus ter-
6 OLOF LARSELL
minalis is a mixed nerve, containing in some fishes peripheral
sympathetic fibers distributed to the blood-vessels. These ef-
ferent fibers make their exit in a dorsal root (N. terminalis) as
the viscero-motor fibers typically do in the spinal region of
lower vertebrates.”’
Belogolowy (12), from a study of young selachian embryos,
concludes that the nervus terminalis is derived from the terminal
portion of the neural crest. This was also claimed by Locy (99)
and (’05 a). \
McKibben (’14) studied the histological structure of the gan-
glion terminale of Mustelus by intravitam methylene-blue stain-
ing. He found the great majority of the cells to be multipolar
and ‘‘few if any bipolar cells.”” Landacre (16), from observa-
tions on Squalus embryos, holds that the terminalis is combined
with the olfactorius as a possible general cutaneous component
of the latter. .
b) Ganoids. Through the studies of Brookover (’08 and ’10)
we have very complete reports of the terminalis for the ganoid
fish Amia. He concludes that the ganglion of the terminalis
arises from the olfactory placode a little later than do the
fibers of the olfactory nerve. He doubts its mdependence
and is disposed to consider it a part of the olfactory system.
In histological observations he finds ganglion cells chiefly of sym-
pathetic type. He also finds fibers along the blood-vessels and
a connection between the terminalis fibers and the ciliary gan-
glion, and suggests that the circumstantial evidence leads one to
ascribe to it a vasomotor function, in part. The same author
gets similar results from Lepidosteus (14) where the ganglion
terminale appears to arise from the olfactory placode after the
formation of the olfactory fibers. ‘‘The disposition of the cells
in Lepidosteus in a more compact central and a diffuse peripheral
ganglion allows of its falling quite naturally into. the morpho-
logical relations of the typical autonomic system.”
This shows the tendency toward interpreting the terminalis
as sympathetic in nature (or at least as containing sympathetic
fibers) which becomes more marked in studies of the mammals.
NERVUS TERMINALIS: MAMMALS ri
c) Teleosis. The presence of the nervus terminalis in bony
fishes was first reported by Sheldon and Brookover (’09) in the
carp (Cyprinus carpio). Sheldon (’09) independently takes up
the central course of the nerve. The tract is composed of un-
myelinated fibers. Numerous scattered ganglion cells were
found on the ventromedial aspect of the olfactory nerve, from
some of which coarse fibers were traced to the olfactory epithel-
ium where they were distributed with the olfactory nerve fibers.
Centrally the fibers for the most part decussate at the anterior
commissure, but no exact nuclear connection could be found.
Brookover and Jackson (’11) studied the development of the
terminalis in Ameiurus, and also its adult relations by means of
the silver-impregnation methods. They find the nerve to be
closely related in its development to the olfactory nerve and
are inclined to consider it a part of this rather than as an inde-
pendent nerve. They point out the proximity of fibers of the
nervus terminalis to blood-vessels, but find only a single in-
stance where the blood-vessel definitely appeared to be inner-
vated by terminalis fibers. A vasomotor functon is suggested.
Amphibia. C. Judson Herrick (’09) found in larval and adult
frogs a bundle of unmyelinated fibers corresponding so. closely
to the nervus terminalis of the fishes in its central course that he
considered it to be homologous with the latter. He was unable
from his material to determine peripheral terminations. The
nerve is not exposed as in selachians. It runs along the ventral
border of the olfactory nerve and becomes imbedded in the brain
substance just caudad to the glomerular formation. Within the
brain substance it passes caudally (in one case showing arboriza-
tions in the lamina terminalis) and the fibers cross in the middle
part of the anterior commissure.
McKibben (’11) traces the course of the nervus terminalis in
Necturus and a number of other tailed amphibians. As in the
frog, it is mainly imbedded in the brain substance. The principal
central distribution is to the preoptic nucleus. The central
bundle undergees a partial decussation in the anterior commis-
sure, but groups of direct fibers extend further backward from
this point, giving off branches at intervals. The continuity of
8 OLOF LARSELL
these fibers with those of the nervus terminalis was not clearly
demonstrated. The fiber groups were lost in the brain substance
without evidence of definite terminations, though the fascicles
reach backward into the mesencephalon (hypothalamus and
interpeduncular region), a condition not yet noted in other
forms. Peripherally are ganglion cells with fibers going appar-
ently to the nasal capsules.
C. Judson Herrick (’17), in connection with other studies,
has completely confirmed in Necturus the findings of McKibben.
Reptilia. Among reptiles the nervus terminalis has been
found by Johnston (’13). In a paper, which embraces also a
consideration of the nerve in pig, sheep, and human embryos,
he describes the terminalis in embryos of the turtle (Emys
lutaria). In Emys the nerve emerges from the rostral end of
the median wall of the brain hemisphere, caudal to the olfactory
bulb. From this point ‘‘It descends over the medial surface of
the bulb and olfactory nerve and bears clumps of ganglion cells
at several points of its course. It comes into close relation with
the dorsal division of the olfactory nerve, but is distinguished
from it.’ Peripherally the fibers of the nervus terminalis are
distributed with those of the dorsal. division of the olfactory
nerve to the extreme lateral portion of the nasal sac, which is
interpreted by Johnston to correspond with the vomeronasal
organ of mammals.
According to McCotter (’17), the dorsal division of the olfac-
tory nerve of the turtle is to be considered homologous with the
vomeronasal nerve of mammals, and the dorsal portion of the
formatio olfactoria of the bulb, as illustrated in Johnston’s fig-
ure 11, corresponds to the accessory olfactory bulb of mammals.
Birds. Whether or not the ganglion cells observed by Ruba-
schin (’03) in the chick embryo represent cells of the nervus termi-
nalis is problematical. This writer describes a ganglionic mass
related on the one hand to the trigeminus nerve and on the other
to the olfactory mucosa. Axones from this mass were traced
to the Gasserian ganglion. Two types of cells were found in the
ganglionic knots: 1) bipolar cells resembling those of the inter-
vertebral ganglia, and 2) multipolar cells with numerous proc-
NERVUS TERMINALIS: MAMMALS 9
esses, of which one in each case enters the ‘ramus olfactorius
nervus trigemini.’ Cells of this type are relatively few in num-
ber. These observations have been interpreted by some writers
to indicate the presence of the nervus terminalis in birds.
Mammals. Since 1905 the nervus terminalis of mammals has
been dealt with in no less than nine scientific memoirs. In some
cases it has been confused with the fibers of the vomeronasal nerve
(Devries, ’05; Déllken, ’09), but in most cases the fibers of the
terminalis are distinguished from those of the vomeronasal.
Notwithstanding these investigations, the nervus terminalis in
the mammals is very imperfctly known and its relations are
obscure.
The first published notice of this nerve in the mammals was
made by DeVries in 1905. He found in the human fetus of three
to four months a transitory ganglion which he regarded as cor-
responding to the ganglion of the nervus terminalis, and which
he designated ‘ganglion vomero-nasale.’ He also found similar
conditions in the guinea-pig. His assumption that the vomero-
nasal nerve of mammals represents the nervus terminalis of
selachians and other fishes is not substantiated by more recent
work.
Dollken (09) studied embryonic stages of rabbit, mouse, guinea-
pig, pig, and human. His account of the central connections of
what he describes as the nervus terminalis is extended. He finds
roots which enter the brain and reach the cortex, the gyrus forni-
catus, the hippocampus, and the septum pellucidum. The pe-
ripheral distribution he describes as being by four or five strands
‘to the vomeronasal organ. It seems clear from his description
and figures that he is dealing almost entirely, if not completely,
with the vomeronasal nerve, which Read (’08) and McCotter (12)
have clearly differentiated from the olfactory fibers proper in
mammals, and McCotter (’17) in the turtle and the frog.
In a paper already referred to in connection with the nervus
terminalis in reptiles, Johnston (’13) also describes the nerve in
embryos of pig, sheep, and human. In pig embryos he finds the
root of the terminalis entering the brain at the ventral end of the
fissura prima. ‘The fibers are traceable for some distance within
10 OLOF LARSELL
the brain toward the anterior commissure. The peripheral
course of the nerve is described as being in the wall of the sep-
tum nasale, along which it passes by several strands to the wall
of Jacobson’s organ and to a small area of the nasal sac immedi-
ately adjacent.
In the human embryo Johnston found essentially the same
central relations as in the pig, but the peripheral distribution is
by a network of nerve bundles in the nasal septum.
He concludes that ‘‘the evidence at present in hand seems to
establish beyond doubt the presence in all vertebrates of a re-
ceptive component in the nervus terminalis supplying ectodermal
territory. This component is derived either from the terminal
part of the neural crest (Johnston, ’09b; Belogolowy, 712) or
from the olfactory placode (Brookover, ’10). The nerve is dis-
tributed to the nasal mucosa, or to a specialized part of it, the
vomeronasal organ.”
McCotter (713) demonstrated by dissection the main central
bundle of the terminalis in the adult dog and eat. He also found
the typical ganglion cells of the nerve distributed along the
vomeronasal strands. No differentiation of fibers from those
of the vomeronasal nerve was obtained by the staining methods
used.
The application to the problem of a modified pyridin-silver
technique by Huber and Guild (’13), served to clearly differenti-
ate the terminalis from the vomeronasal nerve. These investi-
gators used rabbit fetuses and young rabbits. They were
fortunate in securing a differential stain which made it possible
to follow the fibers of the two sets of nerves individually. They
conclude that
this nerve is not a component of the olfactory and vomero-nasal com-
plex, but an independent nerve, with central connections by means of
several small roots to the ventro-mesial portion of the forebrain, caudal
to and independent of the olfactory stalk, and courses in the form of a
loose plexus along the ventro-mesial surface of the olfactory bulb,
reaching the nasal septum and the mesial surface of the vomero-nasal
nerve, which nerve it follows to the vomero-nasal organ, and is further
distributed to the septal mucosa anterior to the path of the vomero-
nasal nerve, in which region especially it is joined by terminal branches
of the trigeminus, mainly from the naso-palatine bundles.
NERVUS TERMINALIS: MAMMALS Nail
Numerous ganglionic masses of various sizes are found. One
group of relatively large size located near the most caudad bundle
of the vomeronasal nerve a short distance from where the latter
leaves the accessory olfactory bulb is regarded as the ganglion
terminale of authors. These groups of ganglion cells present the
appearance of small sympathetic ganglia, and the authors state
that the nerve fibers have more the appearance of sympathetic
and preganglionic fibers than of neuraxes and dendrites of sen-
sory neurones. A comparison of these cells with the cells of the
Gasserian ganglion of the same animals revealed.the fact that
the terminalis cells are of smaller size.
Distribution of terminalis fibers to blood-vessels and septal
mucosa was considered probable from the observations, but the
authors hesitated to assert such distribution because of the com-
mingling of fibers from the trigeminus with those of the nervus
terminalis.
Johnston (’14) describes the central relations of the terminalis
in the adult human, in the horse, porpoise, and the sheep. Nu-
merous rootlets were found in some, especially in the horse, while
in other forms only two or three are enumerated. In the horse,
a large, compact ganglion is described. In the other mammals
the ganglionic masses are described as being smaller but more
numerous, and more or less scattered along that portion of the
nerve which it was possible to examine.
Brookover (’14) independently of Johnston discovered the cen-
tral portion of the nerve in the brain of the adult human, and
reached substantially the same conclusions as to central connec-
tions as did Johnston, namely, that its intracranial course lies
over the middle of the gyrus rectus and appears to enter the
brain substance in the region of the medial olfactory striae.
Both central and peripheral distribution in the human fetus
is described by McCotter (’15). He indicates the peripheral
course in the nasal mucosa, where it resembles in general the
conditions found by Huber and Guild in the rabbit. Centrally
the majority of the fibers form a single strand.
The latest paper which has come to notice has appeared since
the greater part of the observations recorded in the present ar-
i OLOF LARSELL
ticle were made. In this paper Brookover (’17) describes the
peripheral distribution ofthe terminalis in the nasal septum of the
human fetus at full term. This material was prepared by the
pyridin-silver method.
A large plexus of fibers is found anastomosing over the nasal
septum deep to the main arteries. The writer states that this
network ‘‘is so large that it may be considered as hypertrophied
as compared to the known development in other mammals,
without apparently increasing the central root.’’ There are
indications of a sympathetic chain connection with the spheno-
palatine nerve and ganglion.
It seems clear that in selachians, Dipnoi, ganoids, teleosts,
Amphibia, reptiles (turtle at least), and mammals there is com-
mon to all a nerve with central connections with the brain near
the embryonic anterior neuropore, and having a primary periph-
eral distribution to some part of the lining of the nasal cavity.
II. DESCRIPTIVE PART
Material and methods. ‘The original design of this investiga-
tion was to make a comprehensive analysis of the nervus termi-
nalis in the various classes of vertebrates. More difficulties of
analysis and interpretation are encountered in mammals than
in the other groups, so that the present contribution is limited
in its scope to the conditions found in certain mammals, with
the expectation of extending these observations in a subsequent
paper to other classes.
The studies were carried on in the Zoological Laboratory of
Northwestern University from 1915 to 1918, inclusive, under the
direction of Prof. William A. Locy, to whom I express my sense
of indebtedness for helpful advice and criticism. My thanks are
also due Prof. 8S. W. Ranson, of Northwestern University Medical
School, for valuable suggestions.
The mammalian material used consists of the following:
1 longitudinal and 1 transverse series of 10 mm. kitten embryos, fixed in 10 per
cent formalin and stained with haematoxylin.
1 sagittal series through forebrain and nasal septum of kitten one day old, treated
by the pyridin-silver process.
NERVUS TERMINALIS: MAMMALS 13
1 sagittal series through forebrain and nasal septum of kitten two weeks old,
pyridin-silver method.
2 sagittal series of nasal septum of kittens two weeks old, pyridin-silver method.
1 sagittal series through forebrain and nasal septum of kitten one day old, fixed
and decalcified in Zenker’s fluid and stained by the Weigert method.
2 septal mucosae of kitten one day old, stained with methylene-blue.
Numerous series of pig embryos at various stages, stained with haematoxylin or
with Mallory’s connective tissue stain.
In addition to these sections, the following materials were also
studied according to the method indicated: Numerous dissec-
tions of kittens, of puppies, and of adult dogs and cats were made.
The method described by McCotter, by which the head is fixed
in’ Miiller’s fluid to which acetic acid has been added, was used
with good results. Fixation in 10 per cent formalin, followed by
decalcification in 10 per cent nitric acid, was also adopted in
many instances. A light surface stain with borax-carmine was
found advantageous in differentiating the nervus terminalis
from the neighboring tissues.
Frozen beef brains were obtained from the Chicago Stock
Yards, and were found to be very favorable for the dissection of
the delicate strands of the nervus terminalis. These refriger-
ated brains were allowed to thaw in a solution of 10 per cent
formalin at room temperature. Besides being relatively easy to
dissect, this material responded well to the gold-chloride tech-
nique and gave excellent histologic preparations. A number of
beef fetuses of 110 mm. to 140 mm. greatest length were also
dissected. These had been preserved in formalin. Supple-
mentary studies were made on beef material from which the
meninges and a portion of the brain beneath the region of the
nervus terminalis were removed and fixed in 1 per cent osmic acid,
immediately after the brain was taken from the cranial cavity.
Most of the material thus obtained was still warm when fixed.
It was made possible to obtain this through the courtesy of
Swift & Company.
The brain of a full-term mule, freshly removed and fixed in
10 per cent formalin, was obtained. Another mule fetus of 121
mm. greatest length was also studied. The brain, in situ, and
the nasal septum of an adult horse was obtained through the
14 OLOF LARSELL
courtesy of the Chicago Veterinary College. This was fixed in
10 per cent formalin and decalcified in nitric acid.
A number of dissections of pig, rabbit, and sheep embryos
were made. Most of these were treated according to the method
given by Prentiss (715).
Opportunity to examine twelve human brains was obtained
through the courtesy of Prof. 8. W. Ranson. It was attempted
to ascertain if the relations of the fibers of the nervus terminalis
to the cerebral blood-vessels is the same in the human as in the
cat, the mule, and the ox. The nerve was identified in five of
the brains, but no definite light was obtained on the point in
question.
ABBREVIATIONS
ant.cer.art., anterior cerebral artery
art.w., arterial wall
ax., axone
az.cyl., axis cylinder
bi.c., bipolar cell
bl.v., blood-vessel
bu.olf., bulbus olfactorius
bu.olf.ac., accessory olfactory bulb
cen., centripetally
cer.hem., cerebral hemisphere
co.ant., anterior commissure
c.pr., central process
cri.pl., eribriform plate
d.n.ter., dorsal main bundle of nervus
terminalis in nasal septum
fi., nerve fiber
gn., main ganglion (‘ganglion termi-
nale’) of nervus terminalis
gn’., accessory ganglia
in.pl., intracranial plexus of nervus
terminalis
lam.ter., lamina terminalis
m.n.ter., median main bundle of nervus
terminalis in nasal septum
my., myelinated nerve fiber
my.sh., myelin sheath
n.eth.ant., anterior ethmoidal nerve
n.op., optic nerve
no.R., node of Ranvier
n.ter., nervus terminalis
n.vom., hervus vomeronasalis
olf.fi., olfactory fibers
op.chi., optic chiasma
or.vom., vomeronasal organ
per., peripherally ;
p.pl., peripheral (septal) plexus of ner-
vus terminalis
p.pr., peripheral process
pr., nerve process
R., fiber of Remak
r1jr2.r3,r4, central roots of nervus ter-
minalis
ram., ramus
r.dor., ramus of dorsal main bundle of
nervus terminalis
resp.epith., respiratory epithelium
r.med., ramus of middle main bundle
of nervus terminalis
r.ven.,ramus of ventral main bundle of
nervus terminalis
sp., spiral
un.c., unipolar cell
unmy., unmyelinated nerve fiber
v.n.ter., ventral main bundle of nervus
terminalis
NERVUS TERMINALIS: MAMMALS 15
Turtle embryos and young of Amia were prepared by various
methods and a number of dissections were also made.
For nerve terminations the gold-chloride method was em-
ployed, according to the modification given by Hardesty. Meth-
ylene-blue was tried, but with less success in demonstrating the
sensory terminations to be described, although it brought out
Fig 1 Dissection of nervus terminalis of kitten of two weeks, illustrating
plexiform appearance of nerve, as seen through the binocular microscope. The
blood-vessels, along which the majority of the nerve strands course, are not
represented.
the motor endings quite clearly. Both sensory and motor ter-
minations were shown in pyridin-silver preparations.
A modification of the pyridin-silver method given by Huber °
and Guild (’13) was used with good results. This modification
consisted essentially in lengthening the periods during which
the preparation was kept in the various fluids. Briefly summar-
ized, the procedure was as follows: Ammoniated absolute alcohol
16 OLOF LARSELL
(after injection with same) seven days; decalcification in 7 per
cent nitric acid; washing; 80 per cent, 95 per cent, and absolute
alcohols with 1 per cent ammonia added to each, ten days alto-
gether, to insure thorough dehydration; pyridin four days; silver
nitrate, after washing, ten days; four per cent pyrogallol in 5
per cent formalin two days. All of these fluids except the last,
were changed several times. In several of the preparations the
strength of the silver solution was varied, beginning with a solu-
tion of 2 per cent for several days, then reducing to 1 per cent,
to 0.75 per cent, and finally back to 2 per cent. The results in
the way of details of structure and of staiming the finer fibers,
which were obtained by this modification were superior to those
given by’the unmodified procedure.
Fig. 2 Right nervus terminalis from same kitten of which the left nerve is
shown in figure 1. Removed and mounted entire. X 32.
1. The nervus terminalis of the cat
The nervus terminalis of the cat is found on the medial side
of the olfactory stalk. The main trunk runs parallel with the
ventral border of this stalk, between the fissura prima of the
forebrain and the vomeronasal nerves. As shown in figure l,
which represents the mesial aspect of the forebrain and nasal
septum of a kitten of two weeks, the nerve is connected with the
brain by three strands (r},r?,r?). They follow closely parallel to
blood-vessels of small calibre, which enter the brain near the fis-
sura prima. When these vessels were cut at their points of en-
trance into the brain, the nerve strands also became detached.
This was due to the minuteness of the strands which it was not
found possible to sufficiently disentangle from the connective tis-
NERVUS TERMINALIS: MAMMALS 17
sues surrounding both vessels and nerves. Sections (figs. 3 and
23) indicate that the larger strands at least enter the brain inde-
pendently of the blood-vessels.
A fourth strand (r4), which joins the nerve trunk, appears not
to enter the brain. This strand was traced caudally for a short
distance along the anterior cerebral artery, but became so atten-
uated by separating into minute bundles of fibers that it was
not possible to follow the divisions far.
In one of the specimens examined, a kitten one day old, cells
similar to ganglion cells were observed along the course of the
roots for a little distance within the brain (fig. 23).
An elongated ganglionic swelling (fig. 1, gn’) is shown rostrad
to the point where the nerve strand es the anterior cerebral
artery unites with the main trunk of the nerve. Further rostrad
a larger ganglion (gn) is seen in close proximity to the most cau-
dal of the three principal vomero-nasal bundles.
Between these two ganglionic masses the main trunk of the
nerve breaks up into a plexus of nerve strands (a, b,c). Many of
these follow the larger blood-vessels of the region and send twigs
into their walls. Three of the largest strands of the plexus con-
verge distally, uniting with vomeronasal bundles. A number of
strands, finer than any represented in the figure were found, but
were torn in dissection. They were composed of relatively few
fibers each, and uniting with the larger strands, formed a loose
plexus over the medial surface of the olfactory bulb, as shown
in figure 3.
The more ventrally located (c) of the larger pimlles divides
into two strands which unite, one with the ventral bundle of
the vomeronasal nerve, the other with a more dorsally lccated
bundle of the same nerve. i
While the three strands (a, b, c), already noted, diverge at
various angles from the principal axis of the nerve, the other
divisions do not depart so widely. As shown in figure 1 (e),
they form a secondary plexus which reunites, with the exception
of one small strand, into the ganglionic swelling (gn) already
noted. The single bundle which continues rostrally from this
ganglion crosses two of the vomeronasal bundles to become en-
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL, 30, NO. 1
OLOF LARSELL
18
-umpukg *09}4TY PlO-syooM-omy Jo pvoy Ot} JO 4aed pie
auvd uo uo pozoford stjeurUL1e} SnALoU ey} JO snxe]d
‘onbruyoe} TOATIS
MIOJ YSNOLY} SUOTIIOS [V}IIGVS JO SOLOS & WLOdF
[BIUBIOVIPUT OY} JO UOTPNAYSMOSOY & ‘SI
du)
NERVUS TERMINALIS: MAMMALS 19
cased within the same connective tissue sheath which surrounds
the third, more rostrally situated, vomeronasal bundle. The
small strand which fails to reunite with the plexus appeared. to
have been torn from its course distally along a blood-vessel
which passes between the dorsal surface of the olfactory bulb
and the cerebral hemisphere.
Essentially the same relations to the vomeronasal nerves and
to the cribriform plate were found in a dissection of a half-grown
kitten, not figured. In this specimen the left olfactory bulb was
removed, and it was attempted to trace strands of the terminalis
into the nasalseptum. Six strands which clearly belong to the
nervus terminalis were present. Four of these became related
to the vomeronasal nerves, and one of these four remained suf-
ficiently separated from the dorsal bundle of this nerve so that
its course could be followed distinetly through the cribriform
plate. Most of the strands became enclosed by the sheaths of
the vomeronasal bundles in such a manner that it was not possible
to distinguish them from the strands of the latter nerve in their
course peripherally by the method of dissection. Two of the
strands which passed more dorsally did not converge with the
vomeronasal bundles. One of these passed through one of the
more dorsally situated foramina of the cribriform plate, together
with a large olfactory bundle, and its course on the nasal septum
was traced for some distance. ‘The other continued dorsally and
became attached to an artery which lay in the furrow between
the olfactory bulb and the cerebral hemisphere.
Figure 3, which represents a graphic reconstruction of the
terminalis plexus between the vomeronasal nerves and the
point where its roots enter the brain, shows essentially the same
relations. This figure represents a composite of thirteen sections
of the region of the forebrain and nasal septum of a kitten two
weeks old, prepared by the pyridin-silver method. It supple-
ments figure 1 by showing the finer strands of the plexus to
which reference was made, and by bringing out numerous small
ganglia which could not be seen in the dissection represented in
figure 1. A comparison of figure 3 with figure 2, which repre-
sents an in toto amount of the right nervus terminalis of the same
20 OLOF LARSELL
kitten from which figure | was drawn, is interesting in showing
the same manner of distribution of the larger strands as is seen
in the reconstruction. It also indicates, somewhat more clearly,
the position of the ganglion cells along the main trunk of the
nerve. This specimen differs from the one illustrated in figure
3 in that the main ganglionic mass (gn) has fewer ganglion cells
than are present in the corresponding ganglion (gn, fig. 3) of the
other kitten of the same age. More numerous cells, however,
are scattered along the nerve trunk, so that the total number is
approximately the same in the two specimens, if the smaller
ganglia, not observed in the dissected animal, are left out of
consideration in both.
The finer strands, which radiate in various directions from
what may be designated the central bundle, follow along or soon
reach, blood-vessels of various sizes. Many similar strands,
consisting of but three or four fibers could not be seen with the
low magnification of the projection apparatus used in plotting
the figure, and are not included in this reconstruction. These,
if represented, would make the plexus much more intricate,
especially in its rostral part, and would cause it to extend further
rostrally over the olfactory bulb than is figured.
The course of the nervus terminalis in the nasal septum was
followed to best advantage in methylene-blue preparations, fixed
in ammonium-picrate and mounted in a mixture of ammonium
picrate and glycerine. The silver preparations brought out
more clearly the finer strands, but the distortion of the septum
produced by this technique made it difficult to follow the general
course of the various branches by the method of reconstruction.
The nasal septum of a kitten one day old was removed and was
kept moistened for forty minutes in a 0.25 per cent solution of
methylene-blue in physiological salt solution. The preparation
was examined from time to time until a differentiation was ob-
served between the main bundles of the vomeronasal nerves and
the smaller bundles which course parallel to them and which
had previously appeared to be part of them. These smaller
bundles assumed the blue color characteristic of this stain, while
the vomeronasal nerves remained practically unstained. After
NERVUS TERMINALIS: MAMMALS 21
fixation over night in ammonium-picrate, the mucosa was re-
moved from the bony septum and was mounted whole. This
was done with the mucosa from both sides of the septum. The
two sides showed essentially the same. picture, but the right
side was somewhat clearer, and is illustrated in figure 4.
In this specimen the vomeronasal nerves (n.vom.) consist of
three principal bundles in their proximal course on the septum.
About midway toward the organ of Jacobson, which could not
be removed with the mucosa without too great danger of injury
to the latter, two of these bundles divide into secondary strands.
These strands continue to the vomeronasal organ. Parallel
with the vomeronasal bundles and in close proximity to them
for some distance are the main strands of the nervus terminalis.
These strands pass through the cribriform plate, as previously
shown in the dissections and as verified by pyridin-silver prepa-
rations, in company with the vomeronasal bundles. They con-
tinue parallel with them for some distance (fig. 4, n.fer.) and
then divide, forming an intricate plexus in the deeper part of
the mucosa (fig. 4, p.pl.). Comparison with silver preparations
of the nasal septum makes it evident that only a portion of the
plexus was stained in this specimen. This portion was derived
chiefly from the most dorsal (d.n.fer.) of the principal terminalis
strands present. The median of these strands (m.n.fer.) gives
off some small twigs of fibers which anastomose with the main
trunk of the dorsal bundle, and more rostrad it breaks up into
branches, one of which (r.med.) forms a portion of the plexus.
A larger branch of this median bundle continues parallel with
the median branch of the vomeronasal bundle, but could be fol-
lowed for only a short distance rostrally. The most ventral
bundle of the terminalis (v.n.ter.), which is also the largest, was
lost distally because of the idiosyncrasy of the stain. A small
twig (r. ven.) given off in the more proximal part of its course
passes beneath the ventral vomeronasal bundle and is soon lost
in the mucosa ventral to this bundle. A large branch (r.dor.)
from the dorsal terminalis bundle also courses ventrally, but
this could not be traced beyond the dorsal ramus of the ventral
vemeronasal bundle. All other branches which were stained
OLOF LARSELL
N
N
‘02 X “QuNoU sjoY AA “UTe4S
anjq-ousAyPL “Plo ABp ouo u0zTy B jo wNgdes [eseU UO ST[VUTUTIOZ SNAIOU Jo uorngqiaysip jeroydiueg fF “Bi
NERVUS TERMINALIS: MAMMALS a
coursed toward that part of the mucosa which lay dorsal to the
vomeronasal bundles, where the greater part of the plexus is
located. Examination of the figure indicates that many of the
nerve strands follow quite closely the paths of the blood-vessels
represented by broken lines. Owing to the complexity of the
vascular network, only the larger of these vessels were seen
clearly enough in the preparations to make it possible to trace
their courses. No ganglion cells or small ganglia were seen,
but this was laid to the peculiarity of the stain. Silver and Wei-
gert preparations revealed the presence of such cells in the nasal
septum, but in much smaller numbers than are indicated in the
rabbit by Huber and Guild (7138) or in the human by Brookover
(17). The ganglion clusters are, however, very numerous in-
tracranially, especially on the mesial sides of the olfactory bulb.
While it seems likely that most of the plexus formed by the
nervus terminalis on the nasal septum was seen, it is doubtless
true that the rostral part of the septum, which unfortunately
did not take the stain, also contains a continuation of this plexus.
Pyridin-silver preparations indicate this beyond question in the
cat, and it has been shown to be true in the rabbit by Huber and
Guild, and in the human by Brookover, in the papers above
cited.
The observations of the olfactory region of the mucosa are
more dubious. In the methylene-blue preparations, the region
in which olfactory fibers were present was stained a diffuse dark
blue-green, which made it impossible to see any portion of the
plexus if it were present. The pyridin-silver material shows oc-
casional fibers in this region which may belong to the nervus
terminalis, but this cannot be stated with any degree of certainty.
It is possible that they are fibers from the anterior ethmoidal
nerve, the main trunk of which lies in close proximity to many
of the fibers found. .
So far as the methylene-blue material indicates, there is no
connection of the nasal plexus of the terminalis with either the
anterior ethmoidal or the nasopalatine branches of the trigemi-
nal nerve. The silver preparations, however, showed such a
confusion of trigeminal and terminalis fibers in the rostral end
24 OLOF LARSELL
of the septum that it seems certain that there is some comming-
ling of fibers from the nasopalatine nerve in the terminalis plexus.
This was also indicated, although somewhat less clearly, in
Weigert preparations. The Weigert material showed very
clearly that fibers from the anterior ethmoidal nerve take some
cen.
Fig. 5 A small cluster of cells slightly posterior and ventral to the ganglion
(gn.) of figure 3, illustrating some of the types of ganglion cells, together with
myelinated and unmyelinated fibers. Pyridin-silver technique. X 825:
part in the formation of the peripheral plexus of the nervus
terminalis. As shown in figure 22, which represents a portion
of this plexus in a kitten one day old, a few myelinated fibers
are present. Some of these were traced into a strand of the
anterior ethmoidal nerve, which lay in close proximity to the
NERVUS TERMINALIS: MAMMALS 25
portion of the plexus figured. This nerve shows development of
the myelin sheaths, while the intracranial portion of the nervus
terminalis of this specimen (fig. 23) gave no indication of myelin
sheaths as brought out by the Weigert treatment. It does not
seem likely that such sheaths would be formed in the peripheral
portion of the nerve at an earlier date than they are formed in
the part of the nerve nearer the brain. It is therefore assumed
that they are fibers belonging to the already myelinated anterior
ethmoidal nerve.
So far as this material indicates, the central roots of the ter-
minalis, which are easily followed in the sections to their points
Fig. 6 Two bipolar eells and some of the nerve fibers from periphery of main
ganglion (fig. 3, gn.) of the nervus terminalis in kitten of two weeks. Pyridin-
silver technique. X 1266.
of entrance into the brain, are composed entirely of unmyelinated
fibers. In kittens of two weeks, myelinated fibers are found in
the intracranial plexus, although in these also no clear evidence
of such fibers was found among the strands which enter the
brain.
To avoid as far as possible the entrance of fibers from the
fifth nerve as a factor, the greater part of the histological studies
to be described was confined to the intracranial plexus and
ganglia of the terminalis.
Histological. The plexiform character of the nerve in the cat,
together with the structure of the ganglion cells to be described
26 OLOF LARSELL
in the mule, and the relation of the nerve to the adjacent blood-
vessels shown in cat, mule, and beef, suggested strongly that
the nervus terminalis is composed, at least in part, of sympa-
thetic fibers. There remains the possibility, which could not be
adequately tested in the available equine or bovine material,
that there may be a general or special sensory component, in
addition to the motor and sensory sympathetic fibers which
were found. It was accordingly deemed advisable to make as
thorough a study of the terminalis ganglia and of the fibers con-
nected with them in the cat as the material available would
permit.
The fact that the nerve is situated in a position so difficult
of access, together with its small size, and the further circumstance
of its relation to the cribriform plate, made necessary a technique
permitting of decalcification, so that nerve and ganglia might
be studied in situ. For this reason, chiefly, the pyridin-silver
method, as previously described, was employed.
There was considerable variation in different parts of the same
section in the intensity and clearness of the impregnation. It
was also found that, in general; the cells of the smaller ganglionic
clusters were much better differentiated from the background
than were those in the more crowded larger ganglia. Because
of this fact the majority of the cells figured are from the small
clusters of cells, but for comparison, considerable attention was
paid to the large ganglionic mass (figs. 1, 2 and 3, gn.) which
appears to correspond with the ‘ganglion terminale’ of authors.
Types of ganglion cells. Figure 5 represents a typical small
ganglionic mass which had its position in the meninges covering
the mesial side of the olfactory bulb. It lay slightly caudad to
the ganglion terminale and a little more ventrally. This clus-
ter was similar to numerous others scattered throughout the
plexus. Such clusters of cells are usually situated at the meet-
ing point of several small strands of fibers which converge from
various directions. The group of cells figured represents only
a portion of this particular ganglionic mass. The remaining
portion was to be seen in the next section of the series.
NERVUS TERMINALIS: MAMMALS 27
It will be noted that both myelinated (my.) and unmyelinated
(unmy.) fibers are present. The axis cylinders of the myelinated
fibers were stained a darker orange color than were the myelin
sheaths surrounding them. The processes of the unmyelinated
fibers were quite black and stood out distinctly. One of the
latter (a), which comes from the direction of the central connec-
tion of the nerve, divides into two smaller fibers which in turn
subdivide into terminations with small varicosities, and which
form simple pericellular baskets on two of the nerve cells shown.
Several classifications, both of ganglion cells and of sympa-
thetic cells, have been made by investigators, notably by Cajal
(05) Dogiel (08), and Ranson (12) for the former; and by .
Cajal (05), Carpenter and Conel (14), Dogiel (’96), Michailow
(11), and others for the latter type. There is considerable in-
dividual variation among the ganglion cells observed in the ter-
minalis clusters, and they come within one or the other of these
classifications. Still, for purposes of description it is convenient
to designate the types observed according to the number of proc-
esses they possess as unipolar, bipolar, and multipolar. A few
binucleated cells were seen in the cat, but aside from the num-
ber of nuclei, they resembled the other cells of the several varie-
ties and will not be treated separately.
(a) Unipolar cells. Most of the unipolar cells observed re-
semble those usually considered characteristic of the spinal
ganglia. The body of the cell (figs. 5, 10, 18) is ovoid or spheri-
cal, with a rather large nucleus. The single process, which usu-
ally stained brown near the cell body, becomes darker as it
assumes a smaller diameter in its course away from the peri-
karyon. In those cases in which it was possible to follow it for
any distance, it divides into two processes, one directed in the
general direction of the central connection of the nerve, the
other peripherally as respects this connection. No marked dif-
ference in size of these two divisions was noted, although that
which appeared to be the central process was usually slightly
smaller in diameter. It must be understood that only the gen-
eral direction of the course of these fibers is indicated, because of
the various directions different strands of the plexus assumed.
28 OLOF LARSELL
While myelinated fibers were present in the same bundles
which included processes from the unipolar cells, in no case was
a myelin sheath found in connection with processes from such
cells.
Three distinct sizes of unipolar cells were observed. The
predominating size is represented in figure 5 (un.c.), which also
Fig. 7 Bipolar cell and accompanying unmyelinated fiber on the wall of a
blood-vessel between the cerebral hemisphere and the olfactory bulb. Kitten
two weeks old. Pyridin-silver technique. X 450.
Fig. 8A Characteristic small bundle of myelinated and unmyelinated fibers,
with a single bipolar cell in its course. Fig. 8B Portion of a myelinated fiber
more highly magnified, showing a node of Ranvier, and some large varicosities.
Both from kitten of two weeks. Pyridin-silver technique. Figure 8A, X 660;
figure 8B, X 950.
indicates very clearly the bifurcation of the single process of the
cell. Figure 5 (wn.c’.) also shows one of the smallest of the uni-
polar cells seen. The process of this cell could not be followed
for any great distance and no bifurcation was seen. The proc-
ess was directed peripherally. Attention may again be directed
to the pericellular basket surrounding this cell. The relation of
the cell to the nerve fiber of which this basket is the termination,
NERVUS TERMINALIS: MAMMALS 29
together with the peripherally directed process of the cell, sug-
gests that it pertains to the sympathetic system. Its small size
favors this interpretation.
The largest unipolar cell observed (fig. 10) was also the only
one of this type found which was binucleated. The process was
of large diameter and did not stain so dark as did the majority
of fibers. It was not possible to follow it beyond its point of
entrance into the bundle of smaller black fibers of which it
became a part. The large size, both of cell and of cell process,
resembles somewhat the large unipolar cells found by Carpenter
(12) in the ciliary ganglion of the sheep. .
Unipolar cells were not numerous and were found only in the
smaller ganglia. Whether their presence in the larger ganglia
was hidden by the crowded condition of these could not be de-
termined. It seems, however, that the latter are composed
principally of multipolar cells, with fairly numerous bipolar cells
near their peripheries.
(b) Bipolar cells. Cells of this type are quite numerous, both
in the smaller ganglia and in the larger ones. A number of such
cells isolated from other nerve cells were also found. Figure 6
represents two typical bipolar cells from the periphery of the
largest ganglion (fig. 3, gn.) of the left nervus terminalis of a
kitten of two weeks. The processes of these, which are of large
size, took only a brownish tinge from the impregnation. A few
fibers of small size which stained black were present in the im-
mediate neighborhood of these cells and are indicated in the
figure, as are also some large unmyelinated fibers. These are
doubtless fibers of similar bipolar cells. A few relatively small
bipolar cells were found, one of which is shown in figure 5 (67.c.).
The processes of these were quite slender and stained black.
They were followed for some distance, but no conclusive evi-
dence as to their terminations was obtained.
The isolated bipolar cells above noted were found in the course
of small strands of fibers, between the nodal points where several
such strands converge (fig. 8). A few were found on the walls
of blood-vessels or near them. One of these is illustrated in fig-
ure 7, in which also is shown an unmyelinated fiber of small size
30 OLOF LARSELL
Figs. 9 to 15 Ganglion cells from intracranial clusters of nervus terminalis
of kitten two weeks old. Figures 12 and 13 were drawn from cells which lay in
the trunk of the nerve between its points of entrance into the brain wall and the
main portion of the plexus; figures 14 and 15 were drawn from cells which lay
near the center of the main ganglion (gn.) and the other figures were drawn from
cells in various portions of the plexus. Figures 9, 10, 11, and 14 magnified 990
times; figures 12 and 13 magnified 1020 times, and figure 15 magnified 675 times.
Pyridin-silver technique.
NERVUS TERMINALIS: MAMMALS ol
which runs parallel with the processes of the cell. The position
of this cell was well within the crevice between the olfactory bulb
and the cerebral hemisphere, on the wall of a blood-vessel which
passes between the bulb and the hemisphere. The processes of
the cell were stained rather lightly by the silver and it was not
possible to follow them far. Their course so far as visible was
parallel with the wall of the artery. Two slender processes may
be seen to issue from the cell, in addition to the polar processes
of much larger size. Because of these slender offshoots the cell
should possibly be classed with those of multipolar type, but it
is included with the bipolar cells because of its greater similarity
in other respects.
In figure 8A is shown a cell whose position was on the medial
surface of the olfactory bulb, posterior and ventral to the main
ganglion of the terminalis. The larger process (p.pr.) is directed
toward the ganglion, while the very slender process from the
opposite pole of the cell is turned in the general direction of the
central connection of the nerve with the brain, although the
group of three fibers of which it forms one, turns at right angles
to the principal axis of the terminalis plexus. ‘This process
was followed for some distance, but it was lost in the plexus
centripetally.
(c) Multipolar cells. The multipolar type of cell predomi-
nates in the ganglia. These cells vary in size from the relatively
small ones shown in figure 16, to the large cells drawn to the
same scale represented in figures 9, 11 and 17. The number of
processes varies. The cells illustrated in figures 12 and 13,
which were found on the main trunk of the nerve, show but three
processes. The majority of multipolar cells included in the
small ganglia have at least five offshoots. Typical examples of
such cells are shown in figures 11 and 16. The axone could not
always be determined with certainty, but in the cells shown in
figures 11, 12, 14, and 15 the process marked az. appeared to be
the axone. In each of these cases it was directed centripetally.
That which appeared to be the axone of the cell shown in figure
13 (ax.) was directed peripherally. This cell lay in the course
of the main trunk of the nerve.
cen.
JA
ioe
LEE
Y,
\7
Fig. 16 Small ganglionic cluster and intercellular network from portion of
intracranial network of nervus terminalis near lower posterior border of olfactory
bulb. Kitten two weeks old. Pyridin-silver technique. X 990.
Fig. 17 Large multipolar cell with extracapsular network from intracranial
plexus of nervus terminalis of kitten two weeks old. Pyridin-silver technique.
x 990.
32
NERVUS TERMINALIS: MAMMALS oe
Very few binucleated cells of the multipolar type were ob-
served. These, as illustrated by the one represented in figure
9, were similar in every other respect to other multipolar cells.
Fibers and fiber networks. A few of the individual cells were
surrounded by a reticulum of delicate fibers which suggest an
extracapsular network (fig. 17). The capsule itself was not easy
to see in such cases, but the position of the capsule nuclei within
the network seems to justify the interpretation given to these
reticula. In the example illustrated, a small number of threads
extend from the region of a neighboring bipolar cell and take
part in the formation of the pericellular plexus described. The
main fiber from which these threads have their origin, runs par-
allel to the larger process of the bipolar cell, and appears to ter-
minate in a spiral (fig. 17) about this process. The fibers form-
ing this network are intertwined in the most confusing manner
in every direction. They are of very small size. ‘
Intercellular networks also were found in those parts of the
preparations where the impregnation was most favorable. A
peculiarity of the impregnation revealed itself in the fact that
those parts of the sections which showed the fibers most clearly
did not serve so well to differentiate the outlines of the cell
bodies and of the large processes from the perikaryon.
The intercellular networks were found in every case at nodal
points where a number of fiber strands converge. Usually the
ganglion cells enclosed by such a network were of small size, but
occasionally a larger cell was also included. In the example il-
lustrated (fig. 16), which was situated near the lower margin of
the posterior portion of the olfactory bulb, four strands of fibers
converge about a small group of cells. Many of the individual
threads may be followed from one strand through the cell cluster
and into one of the other converging strands, without any ap-
parent connection with the cells. The majority of fibers were
lost in the network. A few may be seen to connect with nerve
cells of the cluster.
This intercellular network resembles to a considerable degree
structures of a similar nature found by Dogiel (’95) in the di-
gestive tract of the dog. It bears an even more striking resem-
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 1
34 OLOF LARSELL
ape NCE CD KC
ioe “
ESO Ren AFR
E3
!
j
ae
7 Be
eS
ZR
Fig.18 Bundle of fibers with two cells in its course, and showing the charac-
teristic method by which small strands of fibers leave the larger bundles. Kitten
two weeks old. Pyridin-silver technique. X 675.
Fig. 19 A somewhat isolated strand of myelinated fibers and a single unmye-
linated thread of relatively large size as compared with the centrally directed
process of the nerve cellshown. Kitten two weeks old. Pyridin-silver technique.
x 990.
NERVUS TERMINALIS: MAMMALS 30
blance to networks found by Ranson and Billingsley (718) in
the cervical ganglion of the dog.
An attempt was made to analyze the various strands of the
plexus, in order to determine, if possible, the types of component
nerve fibers. As previously stated, both myelinated and un-
myelinated fibers are present. Of the latter type, both varie-
ties, namely, fibers of Remak and naked filaments, are abundant.
While for short stretches some of the strands appear to be com-
posed exclusively of one type or the other (figs. 5, 6, 8, 19), the
rule is mixed strands. The fibers of Remak predominate as to
number when the entire plexus is considered. They are particu-
larly numerous near the main ganglion, and appear to include
the majority of fibers which enter this ganglion.
In figure 20 are shown two converging strands consisting prin-
cipally of Remak’s fibers, which approach this ganglionic mass.
The bundle resulting from their union was one of the most com-
pact of the entire plexus. Naked filaments of several sizes are
also included in it. One of the most slender of these (fi) is seen
to approach the nerve cell represented in the figure and to follow
what appeared to be the axone of the cell, to end on the perikar-
yon in a simple pericellular termination. A similar strand, but
with fewer naked fibers, is shown in figure 21, which was drawn
from near the ventral border of the posterior part of the olfac-
tory bulb. This bundle of fibers runs parallel with one of the
arteries of this region. An offshoot (ram.) consisting of four or
five threads leaves the main strand and passes to the wall of a
branch of the artery.
Two fibers which show no neurilemma sheath leave the strand
to pass to the wall of the main artery. These strands so closely
resemble others, the terminations of which are described below,
that it seems certain that they represent fibers which ramify
to form the type of nerve-endings shown in figures 24 to 29.
' Whether or not the fibers of Remak terminate in the sensory end-
ings represented in figures 28 and 29 could not be determined
with certainty. It seems unlikely, in view of the fact that the
fibers leading to the sensory terminations show myelin sheaths
near their endings. The naked filaments show considerable
36 OLOF LARSELL
variation in size, the finer fibers greatly outnumbering those of
coarser diameter.
The myelinated fibers are relatively few in number. They
usually occur in strands of three to five filaments, accompanied
C3her
RK NZ
es SOSA WN ge
Sis)
sre
Ais ators
PRAT SA
art.w.
21
Fig. 20 A bundle of fibers from near the dorsal margin of the olfactory bulb,
caudad to the main ganglionic mass. The strand divides centrally. Kitten two
weeks old. Pyridin-silver technique. X 600.
Fig. 21 A bundle of fibers from near ventral margin of olfactory bulb in
about the same vertical plane as figure 20, showing relation of some of the fibers
to a blood-vessel. Kitten two weeks old, Pyridin-silver technique. X 600.
NERVUS TERMINALIS: MAMMALS 37
by naked threads. In some cases these naked filaments repre-
sent fibers which have lost their myelin sheaths (fig. 8A, az.
cyl.). Many of them show very large varicosities (fig. 8, A and
B), the most pronounced of which are often found near the
nodes of Ranvier. These varicosities appear to have been pro-
duced by unequal shrinkage of the axis cylinder, probably dur-
ing the process of fixation of the tissue. At other points, as
shown in the figures, the cylinders are extremely attenuated.
Some show spiral formations of the axis cylinder within the
myelin sheath (fig. 8A, sp.), In diameter the intracranial
myelinated fibers varied from 1.5 yu to.2.6 yu.
In the septal plexus of the terminalis, myelinated fibers are
found intermingled with the unmyelinated threads (fig. 22).
As previously stated, these belong, in part at least, to the tri-
geminus. It is possible that the myelinated fibers of the intra-
cranial plexus are related to those found in the septal plexus,
and may therefore be trigeminal in orig. This does not seem
likely, but can only be adequately tested by degeneration experi-
ments, which the writer hopes to perform.
The roots which enter the brain, in both of the extra-uterine
stages of growth of the cat in which this point was examined,
appear to be composed exclusively of unmyelinated threads (fig.
23). Fibers of Remak predominate, but a few naked filaments
are mingled with them.
Nerve terminations. There are present in the cat two kinds of
nerve terminations in the walls of the cerebral blood-vessels,
which are connected with fibers from the nervus terminalis.
A third type consisting of free endings in the epithelium of the
nasal septum appears also to be related to this nerve.
The nerve terminations in the walls of the anterior cerebral
artery and its branches, for convenience of description, will be
designated as type I and type II.
Type I (figs. 24 to 27) consists of delicate varicosed fibers
which penetrate the muscular walls of the ‘blood-vessels from the
nervous plexus surrounding these vessels. At varying depths
in the muscular layer, the fibers which penetrate the arterial coat
ramify into very fine arborizations which pass between the smooth
38 OLOF LARSELL
muscle cells and end on the latter. The nerve fibers which end
in this manner are in every case unmyelinated. ‘They show very
slight typical varicosities. The twigs of the terminations are
Fig. 22 Portion of septal plexus showing the presence of myelinated fibers.
Midway between principal vomeronasal foramen and rostral end of nasal septum.
Kitten one day old. Weigert technique. X 230.
Fig. 23 Central roots of nervus terminalis at point of entrance into brain
wall. Kitten one day old. Weigert technique. X 75.
also varicosed. It will be noted from the figures that the man-
ner of distribution of these terminations varies considerably.
Those represented in figures 25 and 27 end in relatively short,
NERVUS TERMINALIS: MAMMALS 39
stout twigs. Others (figs. 24 and 26) have long, very delicate
branches, which sometimes continue their course parallel with
the plane of the fibers from which they spring, sometimes di-
art.w.
25
Fig. 24. Motor nerve termination from muscular coat of anterior cerebral
artery of kitten two weeks old. Gold-chloride technique. 1020.
Fig. 25 Motor nerve termination from branch of anterior cerebral artery of
kitten two weeks old. Gold-chloride technique. > 1020.
Fig. 26 Motor nerve termination from one of the vessels on the mesial sur-
face of the olfactory bulb (in pia mater) of kitten two weeks old. Pyridin-silver
technique. X 1425.
Fig. 27 Motor nerve termination from one of the blood-vessels of the nasal
septum of kitten one day old. Pyridin-silver technique. X 1425.
Fig. 28 Sensory nerve termination from anterior cerebral artery of kitten
two weeks old. Gold-chloride method. X 1020.
Fig. 29 Sensory nerve termination from a small vessel in the meninges near
the olfactory bulb of kitten two weeks old. Pyridin-silver method. X 1425.
verge from it at various angles. These terminations appear to
be similar to those found by Huber (’99) in the cat, and consid-
ered by him to be motor endings.
4() OLOF LARSELL
The terminations represented in figures 24 and 25 were stained
by the gold-chloride method. Those shown in figures 26 and 27
are from pyridin-silver preparations. Similar endings were also
found by the molybdenum methylene-blue process, in the an-
terior cerebral artery and its branches, from which vessels all
of the preparations were made.
The type II endings are strikingly different in appearance
from those of type I. As shown in figures 28 and 29, the termina-
tions are by somewhat spindle-shaped structures, composed of
short, thick branches from the main fiber. These rami end with
terminal knobs. In the gold-chloride preparations (fig. 28) a
spindle-shaped clear space appears to be enclosed by the short
processes which are derived from the nerve fiber. In the silver
material no such clear space is evident, although the general
contour of the termination is the same as in the gold preparations.
The pyridin-silver slides showed the presence of delicate myelin
sheaths on the fibers leading to these terminations. Such sheaths
are not clearly evident in the gold chloride material of the cat,
although similarly prepared slides of the corresponding blood-
vessels of the beef indicate their presence in that animal. No
capsules are present around these terminations in any of the
animals in which they were examined. The methylene-blue
staining did not clearly demonstrate this type of endings, al-
though suggestions of them were visible by this method also.
In general appearance these end-organs resemble to some ex-
tent the corpuscles of Ruffini, but are much smaller. Both in
shape, however, and in the absence of a capsule, they bear a
stronger sunilarity to a type of sensory ending found by Dogiel
in the heart of the cat (Dogiel, ’96, fig. 2, D). Smirnow de-
scribes terminations of somewhat similar appearance in the
atrial endocardium of the cat (Smirnow, 95, fig. 6), and Michailow
(08) has also described non-capsulated sensory terminations in
the myocardium.
In the anterior cerebral artery and its branches of the cat and
of the beef, they lie not in the loose connective tissue, but seat-
tered at various levels in the muscular coat itself.
NERVUS TERMINALIS: MAMMALS 41
So far as the writer is aware, similar structures have not
heretofore been described in the walls of the cerebral blood-
vessels. Huber (’99) noted myelinated fibers along the walls of
the cerebral arteries of the cat, and considered them to be sen-
resp. epith.
3|
Fig. 30 -Free nerve terminations in respiratory portion of nasal mucosa of
kitten one day old, also a portion of the nervous plexus. Pyridin-silver tech-
nique. > 600.
Fig. 31 Portion of the septal plexus of the nervus terminalis of kitten one
day old, showing one of the fibers bifurcated and ramifying into slender twigs
which appear to have been cut off at their ends. Pyridin-silver technique. X
600.
sory as distinguished from the unmyelinated motor fibers which
he found in company with them.
The compact form of this type of endings, differing to so
marked a degree from the other type which has been described
as motor, and closely resembling sensory terminations in other
42 OLOF LARSELL
parts of the vascular system, seems to justify the assumption
that they are sensory in function.
The third type of ending to which reference was made was
found among the epithelial cells hLning the mucosa of the nasal
septum. These endings consist of very delicate arborizations
which pass between the columnar cells of the epithelium and
approach the surface of the membrane (fig. 30). They are ter-
minal twigs of fibers which appear to be unmyelinated. These
fibers, as shown in the figure, approach the epithelium in small
strands of three or four fibers to spread out at its base, where
the terminal threads which form the free end fibers are given
off. No.varicosities or end-knobs were seen.
Such terminations are present in both sensory and respiratory
regions of the septal mucosa and in the epithelium of the vomero-
nasal organ. Similar endings, but with varicosities or end-
knobs, have been described and figured in these regions by von
Brunn (’92), von Lenhossék (’92), Retzius (92), Cajal (94),
Read (’08), and others. Most of these writers tend to ascribe
them to the trigeminal nerve, although von Lenhossék suggests
the possibility that they represent olfactory fibers whose cells
of origin do not have the same position as others, but le within
the centripetal olfactory tract, enclosed in the course of the
olfactory bundle.
Figure 31 represents what appears to be the centripetal con-
tinuation of a fiber which gives rise to free endings such as those
just described. This figure was drawn from a section which lay
just below the epithelium, in a sagittal series through the nasal
septum. As shown in the figure, the fiber (fi) divides at its
extremity into four slender twigs which appeared as if they had
been cut near their tips. Centripetally this fiber unites with a
similar one (fi’). The nerve process of which these fibers are
branches is part of a small bundle (p.pl.) which forms a portion
of the terminalis plexus of the nasal septum shown in figure 4.
While the fibers which terminate in the manner indicated re-
semble in size and distribution those of the nervus terminalis,
there remains the-possibility that they are the continuation of
the more delicate threads which are present in the nasopalatine
NERVUS TERMINALIS: MAMMALS 43
and anterior ethmoidal branches of the trigeminal nerve. As
already noted, there is a commingling of fibers of this nerve and
of the terminalis, and fibers of the trigeminus enter the bundles
which constitute the septal plexus of the terminalis. The intri-
eacy of this plexus made it impossible to follow any individual
fiber very far. Accordingly it was not possible to determine
with certainty whether the free terminations of the septal mucosa
are from terminalis fibers or from the trigeminal nerve.
Other fibers of larger size are also present in the mucosa. These
terminate on or near the septal glands, and show varicosities on
their finer twigs. They also enter into the plexus of the termi-
nalis to some extent. They are given off from the fifth nerve.
It is usually stated that the septal glands are innervated by the
trigeminus, and these fibers appear to be the ones by which
this is accomplished.
2. The nervus terminalis of the beef
The nervus terminalis of the beef, as shown in figure 32, lies
median to the olfactory nerves, between the meninges and the
ventral brain surface. Running parallel with it are branches
of the anterior cerebral artery. In the specimen figured the
greater portion of the left nerve is a compact bundle, while the
right nerve is composed of several strands for the greater part of
its length. In the numerous brains examined there was con-
siderable variation in the relation of the different strands which
by their union form the main nerve bundle. This variation was
found not only when comparing one brain with another, but, as
just indicated, on comparing the two nerves of the same speci-
men. In. some cases the strands were independent up to a
short distance from the forward margin of the hemispheres, in
other cases the bundle was formed much further caudad.
For the greater part of its course the portion of the nerve pres-
ent in the specimens was covered by the meningeal membranes.
It emerges to the outer surface of the arachnoid coat at about
the point where the cerebral hemispheres begin to curve upward.
At this point the nerve was always compact in a single bundle.
44 OLOF LARSELL
This bundle after emerging lay free on the surface of the arach-
noid. The appearance of the free end indicated that it had been
stretched and broken in removing the brain from the cranial
cavity.
Several ganglionic swellings are visible, the largest one in the
brain from which the figure was drawn, at the point (fig. 32,
gn.) where the nerve crosses the artery. Just caudad to this
ganglion, as more clearly shown in figure 33, the bundle divides
into two strands. A short distance rostral to this ganglionic
bg arse eS ant.cerart. (a
Fig. 32. Ventral view of anterior portion of the beef brain, showing the nervus
terminalis and a portion of the anterior cerebral artery and some of its branches.
Nerve strands traced along arteries to points marked z.
mass, a similar strand is given off from the main trunk. These
strands follow the branches of the anterior cerebral artery, as in-
dicated in figures 32 and 33, ramifying to the secondary branches
of these vessels, and at intervals they give off fine twigs which
penetrate into the muscular walls of the arteries.
Several of the strands were traced continuously to the points
marked «x in figure 32, where the respective arteries dipped into
the fissures. On the same brain, and on others, similar strands
were found on all of the arteries examined in this region of the
Fig. 33 Portion of left nervus terminalis and related vessels shown in figure
32, enlarged to show the distribution of strands to the walls of the blood-vessels.
45
46 OLOF LARSELL
brain, both on the ventral surface and in the sagittal fissure.
In all cases in which their connection could be determined, when
followed toward the ventral surface of the brain such strands led
to the principal branches of the anterior cerebral artery, and
connected with the main nerve bundle of the terminalis. When
the overlying connective tissues were successfully removed the
white nerve strands stood out clearly against the reddish brown
of the blood-vessels. The larger bundles were followed with
comparative ease even with the unaided eye, and were not easily
torn. It is unlikely that all of the finer bundles were left intact.
The relatively long stretches on some of the arteries shown in
figure 33 where no twigs are represented probably indicate areas
where they were inadvertently torn in dissection. Those shown
in other parts of the vessels could be raised upon the point of a
needle and stretched in such a manner as to clearly show that
the finer twigs into which they ramify enter the walls of the
blood-vessels.
Caudally, the principal roots which by their union form the
nerve trunk, follow along the larger vessels as far as the latter
could be traced without cutting into the brain. In most of the
specimens examined the internal carotid artery had been severed
so close to the brain in removing the organ that it was not pos-
sible to determine with certainty that any of it was present.
There seems little doubt, however, that the continuation of
strands from the main trunk of the terminalis unites with the
plexus surrounding the internal carotid.
A branch from the main bundle was also followed along the
posterior ramus of the anterior cerebral artery as far as the genu
of the corpus callosum, and similarly on other rami of this vessel
nerve strands were observed. —
Evidence of direct connection with the brain was difficult to
obtain. Delicate branches from the nerve strands on the arteries
were found occasionally to enter openings in the anterior per-
forated space. These branches were extremely difficult to dis- .
entangle from the mass of small blood-vessels, connective tissue,
and elastic fibers among which they were found. Many ap-
peared to be related to the larger vessels which enter the brain
NERVUS TERMINALIS: MAMMALS 47
substance in this region. In the material examined, only one
relatively large strand (fig. 32, c) was seen to enter the brain.
This compared in size with the strands which enter the mule’s
brain (figs. 40, 41, 42).
Although the peripheral relations in the beef could not be
determined in the available adult material, dissection of a num-
ber of ox fetuses of 110 mm. to 140 mm: greatest length brought
ge
Py
Fig. 34 Part of the nasal septum and the cerebral hemisphere of fetal ox of
121 mm. greatest length, showing the nervus terminalis, the vomeronasal nerves,
and a portion of the vomeronasal organ. This drawing was combined from sev-
eral dissections and is to that extent diagrammatic. ™X 4.
out the fact that in the beef, as in the other mammals studied,
the terminalis passes through the cribriform plate in company
with the vomeronasal nerve, and doubtless spreads out on the
septum in the characteristic plexiform manner observed in other
mammals.
As shown in figure 34, which represents the relations in a
fetus of 121 mm. greatest length, the main nerve bundle divides
48 OLOF LARSELL
on approaching the cribriform plate into three strands, each of
which unites with one of the bundles of the vomeronasal nerve.
No differentiation between the terminalis and the vomeronasal
nerve could be seen after the two had joined. Since they are
separate in the cat, rabbit, horse, and human, there can be little
doubt that differential staining methods would show the two as
distinct in the septal region of the beef also. Attention may be
directed at this point to the two ganglia which are visible on the
Fig. 35 Longitudinal section of two nerve strands accompanying anterior
cerebral artery of the beef, showing myelinated fibers. Formalin fixation, iron-
hematoxylin stain. Sections 10 yu. X 450.
Fig. 36 Transverse section of relatively large nerve strand parallel to anterior
cerebral artery of beef, showing myelinated fibers. Formalin fixation, iron-hae-
matoxylin stain. Sections 10 yu. X 450.
main central bundle of the nerve in the fetus and to the two roots
at its central end. These roots appeared to enter the brain
substance near the fissura prima.
Histological. The composition of the main nerve bundle in
the adult, and of the principal strands which run parallel with
the several branches of the anterior cerebral artery, was deter-
mined by the study of sections. As shown in figure 36, which
represents a cross-section of one of the larger strands on the
internal frontal branch of the anterior artery cerebral, sixteen
NERVUS TERMINALIS: MAMMALS 49,
very delicate myelin sheaths are present, as brought ‘out by
iron-haematoxylin staining. More slender nerve strands in the
same preparation showed a smaller number of myelinated fibers.
In longitudinal sections (fig. 35) the myelin sheaths are shown
to continue without interruption for considerable stretches.
They measure from 1.5 » to 2 » in diameter.
Sections of the main trunk peripheral to the ganglionic masses
revealed a much larger number of myelin sheaths. In the sec-
tion illustrated (fig. 37) which was stained by the Weigert method,
sixty-four delicate sheaths are visible. Attention may also be
called in this connection to the many fasciculi, fifteeen in num-
ber, which enter into the formation of the larger bundle.
Fig. 37 Transverse section of main trunk of nervus terminalis of beef, show-
ing distribution of myelinated fibers. Formalin fixation, Weigert stain. 360.
Great care was exercised to prevent decolorizing any of the
sheaths during the process of differentiation, when those meth-
ods were employed which required caution. In view of the par-
tial degeneration which had taken place in some of the sheaths,
it is possible that this process had reached a stage in some fibers
at which the staining methods employed were no longer effective
in differentiating the myelin. It is believed, however, that the
majority of myelinated fibers were stained. The remaining por-
tion of the nerve strand was assumed to be made up of unmye-
linated fibers. These observations were subsequently confirmed
in material which was fixed in osmic acid shortly after the animal
was slaughtered.
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL, 30, NO. 1
50 OLOF LARSELL
An attempt was made to analyze those roots which penetrate
into the brain, using the osmic-acid material. The few roots
which were successfully isolated were of small size. They were
removed and mounted whole in glycerin. One or two myelin
sheaths were observed in each case, the remainder of the root
being evidently composed of unmyelinated fibers. They thus
resemble, except in size, the strands on the arterial walls.
art.w.
my.sh;
38A 398
Fig. 388 A and B- Motor terminations in muscular layer of anterior cerebral
artery of beef. A illustrates the typical appearance, B showed slight terminal
knobs. Gold-chloride technique. X 733.
Fig. 39 Aand B_ Sensory nerve terminations in wall of anterior cerebral artery
of beef, illustrating typical spindle-like outline and myelinated fibers. A repre-
sents a typical termination, B shows two smaller ones attached to a common
branch. Gold-chloride technique. X 733.
Nerve terminations. The frozen beef brains responded well to
the gold-chloride treatment, and terminations of the twigs which
enter the muscular walls of the blood-vessels were found. These
nerve endings, as in the cat, are of two types. Type I (fig. 38,
A and B) consists of delicate varicosed fibers which ramify and
run for varying distances among the unstriated muscle cells.
NERVUS TERMINALIS: MAMMALS 51
These fibers are unmyelinated, and correspond in appearance
with those previously described in the cerebral vessels of the
cat. They are present at various levels of the muscular coat
of the arterial wall, as shown in cross-sections. They send their
twigs between the smooth muscle cells in the typical manner of
motor terminations in this type of muscle. As previously stated,
it seems probable that they should be regarded as motor endings.
The terminations which have been designated as Type II are,
as in the cat, strikingly different in appearance (fig. 39, A and B)
from those of Type I. The nerve fiber leading to these is mye-
linated, as indicated by a reddish cylinder surrounding the pur-
plish-black of the central axis. The myelin sheath in most
cases continues almost to the spindle-shaped termination. The
axis, after emerging from the sheath, divides into two or three
main rami which give off a varying number of short processes.
These processes terminate invariably in rounded knobs. The
portion of the spindle-shaped organ which was not occupied by
these processes appeared nearly clear or had a slight bluish tint
and was slightly granular. It stood out in strong contrast to
the purplish red of the surrounding muscle cells.
The spindles in the beef showed considerable variation in size.
Of those measured, the smallest were 4 » in diameter and 12 u
in length. The largest were 6 u to 7 uw in diameter and 35 » in
length. The majority appeared to be about 5 » by 25 u to 30
uw. Figure 39 illustrates one of the largest observed (A) and also’
an example of the smallest size (B) drawn to the same scale. It
will be noted that the two end-organs represented in figure 39 B
are the terminations of what appeared to be a branch of the
larger fiber shown in the figure. A node is seen in the myelin
sheath, through which the branch to these terminal organs passes.
In a few instances it was possible to follow the myelinated fiber
to the external surface of the artery and for a little way beyond.
3. The nervus terminalis of the mule and the horse
The mule. As represented in figure 40, the nervus terminalis
of the mule runs parallel to the olfactory tracts, between them
and the sagittal fissure of the brain. The nerve in both the mule
52 OLOF LARSELL
and the horse, as in the beef, may be easily seen with the unaided
eye. As shown in the figure, the left nerve had two compact
ganglia. The right nerve had a single ganglionic mass of larger
size. The larger of the two ganglia of the left nerve, which was
the one more rostrally situated (figs. 40 and’ 43), was about 2
mm. long and 4 mm. in diameter. The smaller ganglion was of
about one-half the volume of the other.
The peripheral ends of both nerves were frayed and had the
appearance of having been broken, doubtless where they sub-
divided into strands similar to those observed in the cat and in
the horse. The anterior ends of the olfactory bulbs had been
torn off in removing the brain from the cranial cavity, so it was
not possible to study the relation of the terminalis to the vomero-
nasal nerves in this specimen. The main bundles of the nerve
lay outside the pia mater as far caudally as several millimeters ros-
trad to the ganglia. Here they pierced the pia and in their fur-
ther course centrally lay between this and the brain surface.
Immediately caudal to the posterior ganglion of the left nerve,
and slightly more caudad to the single ganglion of the right nerve,
the main bundle divides into ‘a number of strands or rootlets.
For convenience, the description will be confined to the left
nerve, but in all essential respects, unless indicated to the con-
trary, the same statements apply to the right nerve also.
As shown in figures 40 and 48 the main trunk of the nerve is
formed by the union of three strands. Two of these, the mesial
one and the lateral (figs. 40 and 41), follow a course closely parallel
to two of the branches of the anterior cerebral artery, and give
off delicate rami to them. These twigs enter the walls of the
vessels in the same manner as has already been described in the
OX.
The relation of the medial of the three roots to another branch
of the anterior cerebral artery is shown in figures 40 and 41 at a.
Figure 40 represents the left hemisphere of the brain as viewed
from the medial side. In figure 41 a portion of the same surface
is represented on a larger scale, as seen through the binocular
microscope. It will be noted that strand a turns so as to run
nearly at right angles to the main nerve bundle and follows the
NERVUS TERMINALIS: MAMMALS 53
course of the artery. The nerve strand soon divides into two
smaller strands, one of which had been inadvertently cut in.
separating the hemispheres. The severed portion may be seen
on the stump of the artery, which it was ascertained belongs to
the right hemisphere of the brain.
a
co.ant. nopt. Se:
lamter.
Fig. 40 Mesial view of left hemisphere of full-term mule fetus, showing the
nervus terminalis. Slightly enlarged.
The middle root of the left nerve, as also shown in figures 40
and 41, runs toward the perforated area in front of the lamina
terminalis. At first it is roughly parallel with the anterior cere-
bral artery, then turning upwards, it gives off a small ramus which
enters the brain through a perforation in the furrow of the small
suleus (posterior parolfactory) shown in the figure. Another
ramus of slightly larger size is given off a little further caudad.
This, and the caudally continuing main strand, also enter the
54 OLOF LARSELL
co.ant.
lam.ter.
Fig. 41 Central connections of the left nervus terminalis of full-term mule
fetus, illustrating relations to neighboring blood-vessels and points of entrance
into brain substance. X ca. 5.
Fig. 42 Central relations of right nervus terminalis of mule fetus. X ca. 9.
NERVUS TERMINALIS : MAMMALS 55
brain substance, the one slightly rostrad to the sulcus previously
mentioned, the other at a point about 2 mm. anterior to the
lamina terminalis, and somewhat more than midway between
the anterior commissure and the ventral border of the brain.
As already stated, essentially the same conditions are found
on the right side of the same specimen (fig. 42). Here, however,
only two roots were found which entér the brain substance. The
larger one (b) enters through the same opening as does one of the
blood-vessels of medium size. This was the only case observed
in the mule where a nerve strand of the terminalis and a blood-
vessel enter the brain together. Other rootlets enter very near
the openings through which other blood-vessels pass. The main
nerve trunks on both sides are very clearly formed by the union
of the roots described, with the possible addition of other smaller
ones which may have been torn and lost during the dissection.
The left nerve was removed from its attachments to make
possible a closer study of its structure. As represented in figure
43, the three principal roots merge intothesmallerganglion. Ros-
trally from this ganglion the nerve is compactly enclosed in a
sheath of connective tissue. A few millimeters rostrad from the
small ganglion is the larger one previously noted. Continuing
forward from this point, the nerve remains as a compact bundle
as far as the place where it had been broken in removing the
brain from the cranial cavity.
A very slender blood-vessel (fig. 48, 6/.v.) winds around the
nerve for the greater part of its course. At intervals this vessel
gives off branches which penetrate into the nerve bundle.
Histological. Sections of the ganglia stained with thionin
reveal a considerable variety in the form of the ganglion cells.
The sections were cut 18 » and 20 u thick, so that the processes
could be followed in many instances for some distance. As il-
lustrated in figures 44 to 47, all variations of type from simple
bipolar, to cells having at least five processes occur. Numerous
binucleated cells (fig. 47) were observed. Of the 292 ganglion
cells which were found in the single ganglion of the right nerve,
twenty-seven were binucleated, or 9.25 per cent. Sections cut
at 10 » and stained with haematoxylin showed the same types
of cells in the two ganglia of the left nerve.
rminalis of mule fetus at full term, removed from its
Fig. 43 Left nervus te
connections. Only the ganglia, with portions of the main trunk and of the cen-
tral roots are represented.
56
NERVUS TERMINALIS: MAMMALS o7
Some attention was given to the arrangement of the chroma-
tophile granules in the cells. Carpenter and Conel (’14) have
pointed out the characteristic peripheral arrangement of this
substance in ganglion cells of the sympathetic system, and hold
this to be a distinguishing mark which differentiates them from
cells of the cerebrospinal system. The granules are said to be
scattered throughout the body of the perikaryon in ganglion
cells of the central system. No very satisfactory results were
obtained with the material available. Many of the cells (figs.
44 and 45) are seen to have a somewhat peripheral arrangement
of the granules, while the other cells figured do not give suffi-
cient indication of such distribution of granules to be noticeable.
It should be stated, however, that the sections stained with
thionin, which would affect the Nissl bodies, were too thick to
be favorable for such a study.
An effort was made to demonstrate myelinated fibers, if
present, and to learn, if possible in the large central roots of the
terminalis of the mule, the relative number of such fibers to the
number in the main peripheral trunk. The osmie acid, iron-
haematoxylin, Weigert, and Stroebe methods were each tried,
with variations, a number of times. No success was had in
demonstrating myelin sheaths. In some of the preparations
delicate fibers of small diameter were visible, but it was not
possible to trace them continuously through more than three or
four sections of a series. In the central roots from two to four
such sheath-like structures were present in some of the sections.
In others none could be seen. The peripheral bundle showed as
many as sixteen in certain sections. The results of this part of
the study were on the whole negative. Either the myelin
sheaths are not well developed in the nervus terminalis at this stage
of fetal life of the mule or the formalin preserved material did
not respond to the methods employed for their demonstration.
The Horse. The nervus terminalis of the horse brain examined
(fig. 48) consists centrally of four principal strands which unite
near the base of the olfactory stalk to form a broad, flattened
nerve trunk. This trunk continues rostrally as a compact bundle
as far as the posterior part of the .bulbus olfactorius, receiving
58 OLOF LARSELL
a number of delicate strands in this part of its course. Many of
these strands were traced caudally to neighboring blood-vessels.
The shrunken condition in which these vessels were found, how-
ever, did not favor an attempt to find twigs, such as were ob-
served in the beef, which pass into their muscular walls. At the
region where the olfactory stalk swells to form the bulbus olfac-
torius, the main trunk of the terminalis divides into five rela-
tively large strands and a number of smaller ones. A few milli-
meters caudally of this point a single strand is given off, which
appears to pass laterally and beneath the olfactory stalk. The
five larger strands continue rostrally, anastomosing with the finer
strands and with one another, and thus form a plexus very similar
Figs. 44-47 Ganglion cells from ganglion terminale of full-term mule fetus.
Formalin fixation, thionin stain. > 550.
to that observed in the eat (figs. 1 and 3) and in the-dog (fig.
49),
About 3 mm. rostrad to the point where these strands assume
a separate course, and lying in the path of the largest of the
five, is a large flattened ganglion (fig. 48. gn.) of irregular outline.
This ganglion receives fibers also from other strands of the
plexus than the one in whose apparent course it is placed, so
that it lies in the midst of the plexus.
A much smaller ganglion (gn’,) is present caudally, on the main
trunk midway between the larger ganglion and the point of
union of the central roots.
Rostrally of the main ganglion, two of the larger strands enter
the sheath of connective tissue which encloses the vomeronasal
NERVUS TERMINALIS: MAMMALS 59
nerve in its passage through the cribriform plate. One enters
dorsally of the latter nerve trunk and the other on its ventral
side. The vomeronasal nerve differs in the horse examined from
the condition found in the other mammals studied in that it
cerhem.
(
ll
Hi
(
,
yi
Fig. 48 Ventrolateral view of forward portion of the brain and of part of the
nasal septum of the horse, showing the relations of the nervus terminalis. The
olfactory bulb was removed to bring into view the ganglion terminale and the
intracranial plexus, from the angle at which the dissection was made.
does not divide into the characteristic number of -bundles until
it has passed through the bony plate as a single compact trunk,
surrounded by a heavy sheath. A relatively large strand of the
terminalis penetrates this sheath immediately distal to the point
60 _ OLOF LARSELL
of emergence of the vomeronasal bundle from its foramen. This
strand assumes an independent course, running dorsally a little
way, then divides into a number of smaller strands whose gen- .
eral course is rostrally. These become so attenuated by re-
peated division that it was not possible to follow ‘their finer
ramifications by dissection. Apparently they form a portion of
a plexus similar to that found in the cat and in other mammals.
aN
cerhem.
bu.olf.
bu.olf ac.
Fig. 49 Anterior portion of cerebral hemisphere and a portion of nasal septum
of a puppy two weeks old (estimated) showing nervus terminalis.
The course of the nerve strands in the septum was followed to
some extent by the expedient of stripping off the thick mucosa
from the bony septum and examining its deeper surface. This
required a minimum of dissection, as the nerve strands lay for
the most part in the deeper portion of the mucosa. The proxi-
mal portions of the first two divisions of the vomeronasal nerve
lay in rather deep furrows in the bony septum, and the strands
of the terminalis which accompany the larger nerve through the
cribriform plate, accordingly, emerge from the sheath very close
to the periosteum of the bony septum. In addition to the rami
NERVUS TERMINALIS: MAMMALS 61
of the bundle just noted, other strands were observed in the sep-
tum which also appeared to form a part of the terminalis plexus.
Returning to the other intracranial strands which were noted
distally of the ganglion terminale, the most dorsal one was
somewhat torn in dissection, but appeared to have passed be-
tween the olfactory bulb and the cerebral hemisphere. The re-
maining fibers appear to pass through the cribriform plate in
company with bundles of olfactory fibres, ventral to the vomero-
nasal foramen. Such fibers could not be found distally of the
plate. Possibly the strands in the septum which were noted in
connection with those which emerge in company with the ner-
vus vomeronasalis are distal continuations of the strands now
under discussion. The connections in the bony plate might
easily have been inadvertently injured to such an extent that
the strands immediately distal to the plate were completely
destroyed.
4. The nervus terminalis in other mammals
The observations of the nervus terminalis of the dog, squirrel,
and human, and in embryos of the pig, sheep, and rabbit, were
less extended than those described in the preceding pages. So
far as they were carried, they agreed with the findings in the
other animals studied. The nerve was found in the typical re-
lation to the vomeronasal bundles, and Bases through the
eribriform plate in company with these.
Only one ganglion was found in the dog. This is situated
near the vomeronasal nerve, and is long and fusiform (fig. 49).
In the squirrel also (not figured) one large ganglion was present
at the point where the main trunk of the terminalis splits into
strands similar to those found in the eat and the horse, which
accompany the larger bundles of the vomeronasal nerve through
the bony plate. A number of smaller strands, which appeared
to correspond with the intracranial plexus found in the other
animals, were observed. It may be noted that in the squirrel
the accessory olfactory bulb lies on the dorsolateral side of the
bulbus olfactorius so as to be invisible when the brain is viewed
from the medial side.
62 OLOF LARSELL
As already stated, the attempt to find strands from the termi-
nalis to the blood-vessels of the human brains was not suecess-
ful. Nerve strands similar to those present on the anterior
cerebral artery of the beef and the mule were found and small
twigs were observed to enter the walls of the corresponding
vessels of the human material. Efforts by the gold-chloride and
Bielschowsky methods to demonstrate nerve terminations in
these vessels were not satisfactory with the material available.
The studies on the embryonic material did not serve to reveal
any features not already known. |
Ill. SUMMARY AND COMMENTS
1. The nervus terminalis of mammals is made up in part, at
least, of sympathetic fibers, and its ganglionic clusters contain
sympathetic cells. The wide distribution and large number of
fibers of the peripheral plexus (as noted by Brookover), in com-
parison with the small size of the central connections, resemble
the relation of preganglionic fibers to the postganglionic fibers of
the sympathetic system. This resemblance is strengthened by
the occurrence of pericellular baskets on many of the ganglion
cells of the intracranial clusters.
2. Two types of neurones are present in the terminalis, namely,
1) sensory and 2) motor.
3. Some of the sensory fibers end in the muscular walls of the
anterior cerebral artery and its branches by a type of nerve ter-
mination hitherto undescribed in cerebral blood-vessels.
4. There is some evidence that free nerve terminations in the
epithelium of the septal mucosa and of Jacobson’s organ are also
connected with afferent neurones of the nervus terminalis. For
the present it is assumed that these free sensory terminations
belong to a sensory component of the nervus terminalis which is
distinct from sympathetic afferent fibers which have terminations
in the walls of the blood-vessels.
The type of nerve endings and their position in the mucosa
would seem to indicate that this component is part of the general
visceral afferent system. The early embryonic history of the
nerve in ganoids and selachians might, however, point to a rela-
NERVUS TERMINALIS: MAMMALS 63
.tionship with the special visceral afferent group. As previously
noted, Brookover states that the origin of the nervus terminalis
in the ganoid fishes studied by him is from a portion of the ol-
factory placode. Locy, in describing its early development in
Squalus, attributes its origin to the neural crest, but states also
that “The new nerve has at first a fusion (placode) with the
thickened surface epithelium, located just above the shallow de-
pression that marks the beginning of the olfactory pit. This
connection between the surface epithelium and brain-wall, con-
sists of a group of closely packed cells in which I have failed at
this early stage [6 to 8 mm.] to recognize fibers.’ If this pla-
code described by Locy be homologous with that found by
Brookover, the embryonic evidence in the two groups on which
these writers worked points to an origin which in part at least
corresponds to that of other special sensory ganglia in the head
region.
In connection with the free nerve terminations described, this
embryonic origin of the nerve from a placode suggests the conclu-
sion that there is a sensory component of the terminalis which is
_ distinct from the sensory fibers which terminate in the walls of
the blood-vessels. The neuroblasts which have their origin in
the neural crest might well give rise to the sympathetic cells
' described and figured in the present article, and which have
been described in other groups of vertebrates than the mammals
in the nerve under consideration. The free nerve terminations
in the mucosa do not, however, seem to fit in with the view that
this sensory component belongs to the special visceral system.
5. Many physiologists find insufficient evidence of vasomotor
control of the cerebral blood-vessels. Wiggers (’05, ’08) and
Weber (’08) have presented experimental support of the histo-
logical evidence. Both conclude that there is direct physiological
proof of nerve control over the cerebral vessels. Weber, more-
over, finds indications that there is an accessory vasomotor
center further rostrad in the brain than that situated in the bulb.
This observation is suggestive, in connection with McKibben’s
findings in urodeles, of terminalis tracts extending to the inter-
peduncular region, with a probable center in the neighborhood
of the preoptic nucleus.
64 OLOF LARSELL
6. If connection of the nervus terminalis with such a center.
in the forward part of the brain should be demonstrated, the
forebrain should be included in the list of those divisions of the
central nervous system which are directly related to the sympa-
thetic system.
7. The evidence now at hand points to the conclusion that
the nervus terminalis of mammals is functional. Its relatively
small size in this group as compared with its development in
selachians, may indicate that its functional importance is reduced.
8. Innervation by the nervus terminalis of the vomeronasal
organ, which appears to occur to some extent, is merely incidental.
IV. BIBLIOGRAPHY
ALEXANDER, W.T. 1875 Bemerkungen iiber die Nervender Dura Mater. Archiv
fiir Mikros. Anat., Bd. 11, 8S. 281.
Aus, E. P. 1897 The cranial muscles and the cranial and first spinal nerves
in Amia calva. Jour. Morph., vol. 12, no. 3.
Batoau, C. 1860 Das Jacobson’sche Organ des Schafes. Sitzungsb. d. Kais.
Akad. d. Wissenschaften in Wien, Bd. 42, 8. 449.
BaRDEEN, C. R. 1903 The growth and histogenesis of the cerebrospinal nerves
inmammals. Am. Jour. Anat., vol. 2, pp. 231-257.
BawvEN, H. H. 1894 The nose and Jdcobson’s organ with especial reference to .
Amphibia. Jour. Comp. Neur., vol. 4, pp. 117-152.
Beprorp, E. A. 1904 The early history of the olfactory nerve inswine. Jour.
Comp. Neur., vol. 14, pp. 390-410,
BeLocotowy, G. 1912 Studien zur Morphologie des Nervensystems der Wir-
belthiere. I. Die Entwickelung des Nervus terminalis bei Selachiern.
Bull. Soc. Nat. Moscou, vol. 25.
Bina, R., AND BurcKHARDT, R. 1904 Das Centralnervensystem von Ceratodus
forsteri. Anat. Anz., Bd. 25, S. 588-599.
1905 Das Centralnervensystem von Ceratodus forsteri. Semon Zoél.
Forsch., in Jenaische Denkschr., Bd. 4.
BrRooKxoverR, Cuas. 1908 Pinkus’s nerve in Amia and Lepidosteus. Science,
N.S., vol. 27, p. 918.
1910 The olfactory nerve, nervus terminalis and preoptic sympathetic
system in Amia calva, L. Jour. Comp. Neur., vol. 20, pp. 49-118.
1914a The development of the olfactory nerve and its associated
ganglion in Lepidosteus. Jour. Comp. Neur., vol. 24, pp. 113-130.
1914b The nervus terminalis in adult man. Jour. Comp. Neur., vol.
24, pp. 131-135.
1917 The peripheral distribution of the nervus terminalis in an*
infant. Jour. Comp. Neur., vol. 28, pp. 349-360.
BRookover, Cuas., AND Jackson, T. 8S. 1911 The olfactory nerve and the ner-
vus terminalis of Ameiurus. Jour. Comp. Neur., vol. 21, p. 237.
NERVUS TERMINALIS: MAMMALS 65
Casa, 8S. R. 1893 Neue Darstellung vom histologischen Bau des Centralner-
vensystems. Arch. fiir Anat. u. Phys., Ant. Abtheil., p. 393.
1905 a Tipos celulares de los ganglios sensitivos del hombre y mami-
feros. Trab. del Lab. de Inves. Biol. de la Univ. de Madrid, T. 4, p. 1.
1905 b Las celulas del gran simpatico del hombre adulto. Trab. del
Lab. de Inves. Biol. de la Univ. de Madrid, T. 4, p. 79.
1911 Histologie du Systéme Nerveux de L’Homme, T. 2, p. 670.
CARPENTER, F. W. 1911 The ciliary ganglion of birds. Folia Neurobiologica,
Bd. 5, 8. 738-754.
1912 On the histology of the cranial autonomic ganglia of the sheep.
Jour. Comp. Neur., vol. 22, p. 447.
CarPENTER, F. W., AND Cone, J. L. 1914 A study of ganglion cells in the sym-
pathetic system, with special reference to intrinsic sensory neurones.
Jour. Comp. Neur., vol. 24, p. 269.
CarPEnTER, F. W., ann Main, R. C. 1907 The migration of medullary cells
into the ventral nerve-roots of pig embryos. Anat. Anz., Bd. 31, S.
303-306. :
Cote, Franx J. 1896 The cranial nerves of Chimaera monstrosa. Proc. Roy.
Soc. Edinburgh, vol. 21, pp. 49-56.
Devries, E. 1905 Note on the ganglion vomeronasale. K. Akad. van Weten-
schappen te Amsterdam, vol. 7, p. 704.
Doaiet, A. S. 1895 Zur Frage iiber den feineren Bau des sympathischen Ner-
vensystems bei den Sdugethieren. Arch. fiir Mikros. Anat. und
Entwick., Bd. 46, 8S. 305-344.
1896 Zwei Arten sympathischer Nervenzellen. Anat. Anz., Bd. 11,
S. 679-687.
1898 Die sensiblen Nervenendigungen im Herzen und in den Blutge-
fissen der Siugethiere. Arch. fiir Mikros. Anat. und Entwick., Bd.
52, S. 44-70.
1908 Der Bau der Spinalganglien des Menschen und der Siugethiere.
Jena.
DoéuixeNn, A. 1909 Ursprung und Zentren des Nervus terminalis. Monatsschr.
f. Psych. u. Neur., Erg. Heft, Bd. 26, S. 10.
Fritscu, G. 1878' Untersuchungen iiber den feineren Bau des Fischgehirns.
Berlin.
Firprincer, K. 1904 Notiz iiber einige Beobachtungen am Dipnoerkopf.
Anat. Anz., Bd. 24, S. 405-408.
Herrick, C. Jupson 1903 On the morphological and physiological classification
of the cutaneous sense organs of fishes. Amer. Nat., vol. 37, pp. 313-
318.
1909 The nervus terminalis (nerve of Pinkus) in the frog. Jour.
Comp. Neur., vol. 19, p. 175.
1916 Introduction to neurology. Philadelphia.
1917 The internal structure of the midbrain and thalamus of Nectu-
. ‘rus. Jour. Comp. Neur., vol. 28, p. 236.
Herrick, C. L. 1893 Topography and histology of the brain of certain reptiles.
Jour. Comp. Neur., vol. 3, p. 124.
Howe, W. H. 1915 Text-book of physiology, 6th ed., p. 632. Philadelphia.
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, No. 1
66 OLOF LARSELL
Huser, G. Carut 1897 Lectures on the sympathetic nervous system. Jour.
Comp. Neur., vol. 7, pp. 73-145.
1899 a Observations on the innervation of the intracranial vessels.
Jour. Comp. Neur., vol. 9, pp. 1-25.
1899 b A contribution on the minute anatomy of the sympathetic
ganglia of the different classes of vertebrates. Jour. Morph., vol. 16,
pp. 27-90.
1913 The morphology of the sympathetic nervous system. XVII
International Congress of Medicine, London.
Huser, G. C., anp Dewitt, Lyp1a_ 1898 A contribution on the motor nerve-
endings and on the nerve-endings in the muscle spindles. Jour.
Comp. Neur., vol. 7, pp. 169-230.
Huser, G. C., anp Guitp, Stacy R. 1913 Observations on the peripheral dis-
tribution of the nervus terminalis in Mammalia. Anat. Rec., vol. 7,
pp. 253-272.
Jones, W. C. 1905 Notes on the development of the sympathetic nervous sys-
tem in the common toad. Jour. Comp. Neur., vol. 15, p. 113.
JounstTon, J. B. 1906 The nervous system of vertebrates. Philadelphia.
1909 The morphology of the forebrain vesicle in vertebrates. Jour.
Comp. Neur., vol. 19, pp. 457-539.
1911 The telencephalon of selachians. Jour. Comp. Neur., vol. 21,
pp. 1-118.
1913 Nervus terminalis in reptilesandmammals. Jour. Comp. Neur.,
vol. 23, pp. 97-120.
1914 The nervus terminalis in man and mammals. Anat. Rec., vol. 8.
KeEIBEL AND Matt 1912 Human embryology, vol. 2. Philadelphia.
KO6O.uikEeR, A. 1896 Handbuch der Gewebelehre des Menschen, Bd. 2, S. 835.
Kuntz, ALBERT 1911 The evolution of the sympathetic nervous system in ver-
tebrates. Jour. Comp. Neur., vol. 21, pp. 215-236.
1913 a The development of the cranial sympathetic ganglia in the pig.
Jour. Comp. Neur., vol. 23, pp. 71-96.
1913 b On the innervation of the digestive tube. Jour. Comp. Neur.,
‘vol. 23, pp. 173-192.
1914 Further studies on the developmert of the cranial sympathetic
ganglia. Jour. Comp. Neur., vol. 24, p. 235.
Lanpacrg, F. L. 1910 The origin of the sensory components of the cranial
ganglia. Anat. Rec., vol. 4, pp. 71-79.
1916 The cerebral ganglia and early nerves of Squalus acanthias.
Jour. Comp. Neur., vol. 27, p. 19.
Laneuey, J. N. 1900 The sympathetic and other related systems of nerves.
Schifer’s Text-book of Physiology, vol. 2, pp. 616-696.
1903 The autonomic nervous system. Brain, vol. 26, pp. 1-26.
Locy, W. A. 1899 New facts regarding the development of the olfactory nerve.
Anat. Anz., Bd. 16, S. 273-290.
1903 A new cranial nerve in selachians. Mark Anniv. Volume.
1905 a On a newly recognized nerve connected with the forebrain of
selachians. Anat. Anz., Bd. 26, 8. 33-63 and 111-123.
1905 b A footnote to the ancestral history of the vertebrate brain.
Science, N. S., vol. 27, pp. 180-183.
NERVUS TERMINALIS: MAMMALS 67
Matong, E. T. 1913 Nucleus cardiacus nervi vagi and the three different types
of nerve cells which innervate the three different types of muscle.
Am. Jour. Anat., vol. 15, pp. 121-127.
McCorrer, R. E. 1912 The connection of the vomeronasal nerves with the
accessory olfactory bulb in the opossum and other mammals. Anat.
Rece., vol. 6, p. 299.
1913 The nervus terminalis in the adult dog and cat. Jour. Comp.
Neur., vol. 23, p. 145.
1915 A note on the course and distribution of the nervus terminalis
inman. Anat. Rec., vol. 9, p. 243.
1917 The vomeronasal apparatus in turtle and frog. Anat. Rec., vol.
12.
McKissen, P. S. 1911 The nervus terminalis in urodele Amphibia. Jour.
Comp. Neur., vol. 21, pp. 261-309.
1914 Ganglion cells of the nervus terminalis in the dogfish (Mustelus
canis). Jour. Comp. Neur., vol. 24, pp. 437.
Micuaitow, 8S. 1908 Die Nerven des Endocardiums. Anat. Anz., Bd. 32.
1910 Uber die sensiblen Nervenendapparate der zentralen sympath-
ischen Ganglien der Siugethiere. Jour. f..Psych. u. Neur., Bd. 16, 8.
269.
1911 Der Bau der zentralen sympathischen Ganglien. Internat.
Monatsschr. f. Anat. u. Phys., Bd. 28, S. 26.
Miuumr, L. R., unp Dani, W. 1910 Die Beteiligung des sympathischen Ner-
vensystems an der Kopfinnervation. Deutsches Archiv f. klin. Med.,
Bd. 99, S. 48-107. :
Norris, H. W. 1913 The cranial nerves of Siren lacertina. Jour. Morph., vol.
24, p. 254.
Pinxus, Fevrx 1895 Die Hirnnerven des Protopterus annectens. Morph. Arb.
(G. Schwalbe), Bd. 4, 8. 275.
1905 Uber den zwischen Olfactorius und Opticusursprung des Vor-
derhirn (Zwischenhirn) verlassenden Hirnnerven der Dipnoer und
Selachier. Arch. Physiol. Jahrb., sup. Heft. 2, 8. 447.
Prentiss, C. W. 1915 Text-book of embryology. Philadelphia.
Ranson, 8. W. 1912 The structure of the spinal ganglia and of the spinal
nerves. Jour. Comp. Neur., vol. 22, pp. 159-175.
1915 The vagus nerve of the snapping turtle (Chelydra serpentina).
Jour. Comp. Neur., vol. 25, p. 301.
Ranson, 8. W., anp Binutinastey, P.R. 1918 The superior cervical ganglion and
the cervical portion of the sympathetic trunk. Jour. Comp. Neur.,
vol. 29, no. 4.
Reap, Erriz A. 1908 A contribution to the knowledge of the olfactory appara-
tus in the dog, cat and man. Am. Jour. Anat., vol. 7, pp. 17-47.
Ruspascuin, W. 1903 Uber die Beziehungen des Nervus trigeminus zur Riech-
schleimhaut. Anat. Anz., Bd. 22, 8. 407.
SHEtpon,-R. E. 1908 The participation of medullated fibers in the innervation
of the olfactory mucous membrane of fishes. Science, N. 8., vol. 27,
no. 702, p. 915.
68 OLOF LARSELL
SHEeLpon, R. E. 1909 The nervus terminalis in the carp. Jour. Comp. Neur.
vol. 19, p. 191.
1912 The olfactory tracts and centers in teleosts. Jour. Comp. Neur.,
vol. 22, pp. 177-389.
SHELDON, R. E., AND Brookover, CuHas. 1909 The nervus terminalis in tele-
osts. Anat. Rec., vol. 3, p. 257.
SHERRINGTON, C. 8S. 1911 The integrative action of the nervous system. New
Haven.
Sewertzorr, A. N. 1902 Zur Entwickelungsgeschichte des Ceratodus forsteri.
Anat. Anz., Bd. 21, p. 606.
Smirnow, A. 1895 Uber die sensiblen Nervenendigungen im Herzen bei Am-
phibien und Saugethieren. Anat. Anz., Bd. 10, 8. 737-749.
Smitu, G. Exitior 1895 Jacobson’s organ and the olfactory bulb in Ornitho-
rhynchus. Anat. Anz., Bd. 11, S. 161.
STREETER, G. L. 1912 The development of the nervous system. 4. The sym-
pathetic nervous system. Human Embryology, Keibel and Mall, vol.
2, pp. 144-154.
von Brunn, A. 1892 Die Endigung der Olfactoriusfasern in Jacobson’schen
Organe des Schafes. Arch. f. Mikros. Anat. u. Entwick., Bd. 39, S.
651.
von Lenuoss&éK, M. 1892 Die Nervenurspriinge und Endigungen im Jacob-
son’schen Organ des Kaninchens. Anat. Anz., Bd. 7, 8. 628. ,
1911 Das Ganglion ciliare der Végel. Arch. f. Mikros. Anat. u.
Entwick., Bd. 76, S. 745.
Weser, Ernst 1908 Uber die Selbstindigkeit des Gehirns in der Regulierung
seiner Blutversorgung. Arch. f. Phys. u. Anat., Phys. Abteil., S. 457.
Wiaasmrs, C. J. 1905 Action of adrenalin on the cerebral vessels. Amer. Jour.
Physiol., vol. 14, p. 452.
1907 The innervation of the cerbral vessels as indicated by the action
of drugs. Amer. Jour. Physiol., vol. 20, p. 206.
1908 Some vasomotor changes in the cerebral vessels obtained by
stimulating the carotid plexuses. Amer. Jour. Physiol., vol. 21, p.
454.
Resumido por el autor, Edward Phelps Allis Jr.
Los nervios oftalmicos de los peces gnatostomos.
En Polypterus sale del crdneo, con las fibras del lateral que
van a los nervios oftalmico superficial y bucal lateral, un fascf-
culo de fibras del comiin, formandose un ganglio con estos dos
fasciculos, colocado en posicién dorsal con relacién al ganglio
cutaneo general del trigémino. De este ganglio comun-lateral
parten fibras del comin, las cuales van a parar al ganglio cutdneo
general, que no envia fibra alguna al primero. El oftdlmico
superficial se origina en su totalidad en el ganglio lateral comtin
y contiene fibras del lateral y comtin, pero en pequefo ntimero,
careciendo probablemente de fibras del cutdneo general. El
llamado trigémino-oftaélmico profundo es en realidad un ramo
oftalmico profundo puesto que nace de un ganglio profundo e
independiente, formado por una raiz profunda independiente;
aparentemente contiene solo fibras cutdneas generales. Posee
ramas frontales y nasales que son homdlogas de las mismas
ramas del nervio oftalmico de‘’los vertebrados mds superiores.
Kste ultimo nervio es también por esta causa un nervio profundo
y no una parte del trigémino. En los Dipnoos la distribuci6én
es, en apariencia, las misma que en Polypterus. En los Holé-
steos hay un ramo oftdlmico superficial que contiene fibras del
lateral, comun y cutdneo general y una porcién oftalmica pro-
funda, pero el ramo oftalmico profundo, cuando existe, esta de-
generado. En ciertos Teleésteos (Gasterosteus) existe una dis-
tribucién semejante a la mencionada en los Holdsteos, pero en la
mayor parte de ellos hay un ramo oftdlmico superficial de com-
posicion variable, faltando el ramo oftdlmico y la porcién
oftalmica profunda.
Transiation by Dr. José Nonidez,
Columbia University
AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY
THE BIBLIOGRAPHIC SERVICE, DECEMBER 9
THE OPHTHALMIC NERVES OF THE GNATHOSTOME
FISHES.
EDWARD PHELPS ALLIS, JR.
Menton, France
There has long been, and still is, confusion in the terms em-
ployed to designate the ophthalmic nerves of the gnathostome
fishes. ‘These nerves were formerly considered to be the equiva-
lents of the ramus ophthalmicus trigemini of higher vertebrates
(Stannius, ’49), and as in many fishes, one of them lies super-
ficial to the other they were called the rami superficialis and
profundus trigemini. The term profundus was, however, fre-
quently given to two distinctly different nerves, one of which
runs forward between the superior and inferior divisions of the
nervus oculomotorius and then ventral to the nervus trochlearis,
while the other runs forward dorsal to both those nerves. The
former nerve is alone properly called a profundus and it is typ-
ically found in the Elasmobranchii. A nerve that, in certain
fishes, connects the basal portion of this profundus nerve with
that of the superficialis was called the portio ophthalmici profundi.
Later, it was found that those nerve fibers of the superficialis
nerve that innervate the organs of the supraorbital laterosensory
line all issue from the medulla as an apparent part of the root
of the nervus facialis, and as this was considered to be incontro-
vertible evidence that they belonged to the latter nerve, that part
of the ophthalmicus superficialis that was formed by them was
called the ramus ophthalmicus superficialis facialis. The re-
maining fibers of the superficialis nerve were still considered to
belong to the trigeminus, and were usually called the ramus oph-
thalmicus superficialis trigemini, but as, in the Holostei and
Teleostei, they lie deeper than the lateralis fibers, they were fre-
quently called the ramus ophthalmicus profundus trigemini, and
69
70 EDWARD PHELPS ALLIS, JR.
apparently considered to be the homologue of the similarly
named nerve of the Elasmobranchii. This latter nerve of the
Elasmobranchii was still called the ramus ophthalmicus profun-
dus trigemini, but there was a growing opinion that it belonged
to a cranial segment next anterior to that of the trigeminus.
It was then still later found that communis fibers might also
form part of the ophthalmicus superficialis, and that, like the lat-
eralis fibers of that nerve, they issued from the medulla as an
apparent part of the root of the nervus facialis. Consistency
then evidently demanded that these communis fibers also be
included in the ophthalmicus superficialis facialis, but as that
term had come to mean a purely lateralis nerve, the communis
fibers were either still relegated to the ophthalmicus super-
ficialis trigemini or a new term, truncus supraorbitalis, was
given to the entire ophthalmic nerve, the term ophthalmicus
superficialis facialis still being employed to designate the later-
alis fibers only of the nerve.
The ramus ophthalmicus superficialis of fishes thus came to be
considered to be a nerve formed by the secondary juxtaposition
of fibers derived from two adjacent segmental nerves, the gen-
eral cutaneous fibers of the nerve being derived from the nervus
trigeminus and the lateralis and communis fibers from the ner-
vus facialis. This schema of the nerve still, however, left un-
accounted for those fibers that were known to be derived, in cer-
tain fishes, from the portio ophthalmici profundi, and although
this portio, as an independent nerve, had only been described
in a few fishes, there was no apparent reason for assuming that
it did not also exist in many, if not all, of those fishes in which
the profundus and trigeminus ganglia have completely fused with
each other.
This confusion of terms and complication of conditions led
me, in my work on the mail-cheeked fishes (Allis, ’09) to readopt
the term, ophthalmicus superficialis trigemini, first given to this
nerve, and to call its lateralis and communis fibers the lateralis
and communis trigemini. A name of some sort had to be given
to the nerve, and this one seemed to me to be the “‘single name
already current” that Herrick (’09) has later suggested should be
OPHTHALMIC NERVES OF GNATHOSTOME FISHES 71
selected for each so-called composite nerve, and it certainly had
valid claim to priority over the one that he had himself earlier
employed, namely, truncus supraorbitalis. Furthermore, it was
at that time, and still is, my opinion that it has by no means
as yet been definitely established that the lateralis and com-
munis fibers found in the ophthalmicus superficialis do not be-
long definitely to the trigeminus, their centers of origin having
simply fused with, or become contiguous to, those of similar
fibers that belong to the nervus facialis. The conditions in
Polypterus, described immediately below, certainly favor this
conclusion, but they at the same time further complicate the
choice of a proper name for the nerve.
In Polypterus there is a profundus ganglion which, in a 75-mm.
specimen examined in serial transverse sections, is extracranial
in position and lies wholly anterior to, and independent of, the
trigeminus ganglion. The root of this ganglion traverses a fora-
men that lies anterior to the foramina for the roots of the nervus
trigeminus, enters the medulla slightly anterior to the general
cutaneous root of the latter nerve, and is, so far as can be deter-
mined from my somewhat imperfect and unsatisfactory sections,
composed exclusively of general cutaneous fibers. From this
profundus ganglion a typical ramus ophthalmicus profundus
arises, and also either a single nerve, which immediately bifur-
cates, or two independent nerves, these latter one or two nerves
forming the portio ophthalmici profundi shown by van Wijhe
(82) in his figure of this fish. The branches of this ramus and
portio all join the ramus ophthalmicus superficialis, and, accom-
panying it and its branches, but in no way fusing with them, are
distributed mainly to tissues on the dorsal surface of the anterior
portion of the head. The ramus ophthalmicus superficialis arises
from a ganglion formed on two bundles of fibers, one of which
contains all the lateralis fibers that go to the rami ophthalmicus
and buccalis lateralis, while the other is an intracranial branch
of the communis root of the nervus facialis. These two roots
of this ganglion, which thus quite certainly contain no general
cutancous fibers, issue together from the cavum cerebrale cranii,
and the ganglion formed on them lies dorsal to the ganglion
72 EDWARD PHELPS ALLIS, JR.
formed on the general cutaneous root of the trigeminus. Com-
munis fibers are sent from this lateralis-communis ganglion to
the general cutaneous one, but no fibers can be traced from the
latter to the former ganglion. The ramus ophthalmicus super-
ficialis, which arises wholly from the lateralis communis ganglion
and contains both lateralis and communis fibers, thus certainly
contains but few, and probably no general cutaneous ones.
The ophthalmicus profundus must then supply most, and prob-
ably all, of the general cutaneous fibers that go to that part of
the dorsal surface of the head that is supplied, in most teleosts,
exclusively by branches of the so-called ophthalmicus super- —
ficialis trigemini, and the latter nerve is wholly wanting in
Polypterus.
The ophthalmic nerves of Polypterus thus are: a nerve that I
should call the ramus ophthalmicus superficialis trigemini, but
which, according to currently accepted views, would be called the
ramus ophthalmicus superficialis facialis; and a ramus ophthal-
micus profundi, which has ramuli frontalis and nasalis that are,
respectively, the portio ophthalmici profundi and ramus ophthal-
micus profundus trigemini of current descriptions. If the oph-
thalmicus superficialis were to abort, as the fibers that form it
always do in all land vertebrates, there would remain a nerve
that would quite unquestionably be the homologue of the oph-
thalmic nerve of Hoffmann’s (’86) descriptions of embryos of rep-
tiles, for the radix longa of my 75-mm. Polypterus, which arises
from the profundus ganglion, is the ramus ciliaris of Hoffmann’s
descriptions of reptiles, which latter nerve fuses, for a certain
distance, with the ramus nasalis to form the ramus nasociliaris.
The opinion long ago expressed by His (’87, p. 398), that the
ramus nasalis, or nasociliaris, of mammals corresponds to the
so-called ramus ophthalmicus profundus trigemini of selachians
is thus confirmed, as is also my conclusion, in my work on Mus-
telus (Allis, ’01, p. 299), that the ramus ophthalmicus profun-
dus of Potypterus and the portio ophthalmici profundi of gan-
oids are, respectively, the homologues of the nasal and frontal
branches of the ophthalmic nerves of higher vertebrates. There
would then be no so-called ramus ophthalmicus superficialis tri-
OPHTHALMIC NERVES OF GNATHOSTOME FISHES 73
gemini in these latter vertebrates, which is in accord with Pinkus’s
(94) conclusion that the presence of this nerve in the Amphibia
is very doubtful, and with Norris’s statement (’13, p. 292) that,
in Siren lacertina: ‘“‘It is questionable whether any general cu-
taneous fiber should be considered as a constituent part of the
dorsal, or supraorbital division [of the truncus supraorbitalis] of
Siren.”’ The ramus ophthalmicus profundi of Polypterus thus
quite certainly being the homologue of the so-called ophthalmicus
trigemini of man, the introduction of a proper and uniform ter-
minology becomes a somewhat radical proceeding, for it evi-
dently requires a renaming of the nerve in man.
The conditions in those fishes, other than Polypterus, in which
either a ramus ophthalmicus profundus trigemini, or a portio
ophthalmici profundi, has been described, may now be consid-
ered, and I have, furthermore, examined the conditions in Gas-
terosteus, Cottus, and Clinocottus, in. which fishes there is an
anterior portion of the ascending process of the parasphenoid
that occupies the position of the pedicel of the alisphenoid of
Amia. ‘The reason for examining these latter fishes was the con-
viction that, if there were a portio ophthalmici profundi, it
would lie anterior to the above-mentioned anterior process of
the parasphenoid, for that is the relation that the radix profundi
" of these fishes has to that process; and, conversely, if no branch
of the profundus, sent to the superficialis, were found in that posi-
tion, it would be, in my opinion, conclusive evidence that the
portio ophthalmici profundi was wanting in these fishes, and
hence presumptive evidence that it was also wanting in those
of the Teleostei in which the process is not found.
In Protopterus, Pinkus (’94) describes a ramus ophthalmicus
profundus trigemini, but no portio ophthalmici profundi. The
first three branches of the ophthalmicus profundus are small, and,
running forward dorsal to the nervi oculomotorius and troch-
learis, become associated with a ramus ophthalmicus superficialis
facialis. The ophthalmicus profundus then separates into two
nearly equal portions, one of which is shown, in the figure given,
running forward dorsal to both divisions of the nervus oculomo-
torius, and the other ventral to them, the dorsal branch then
74 EDWARD PHELPS ALLIS, JR.
apparently passing ventral to the nervus trochlearis and joining
the first three small branches of the nerve. The relations of this
profundus nerve to the oculomotorius thus differ radically from
those in Polypterus, but it seems probable that its first three
branches correspond to the frontal branch of the nerve of the
latter fish, and the two larger ones to the nasal branch. A small
nerve is said to arise from that part of the trigeminus ganglion
from which the ophthalmicus profundus has its origin, and to
immediately join the ophthalmicus superficialis facialis, and
Pinkus doubtfully calls it the ramus ophthalmicus superficialis
trigemini. Comparison with Polypterus would, however, indi-
cate that it is a persisting remnant of the communis component
only of the ramus ophthalmicus superficialis of the latter fish.
In Ceratodus there is a bar of cartilage that represents the
pedicel of the alisphenoid (Allis, 714), and the ophthalmicus pro-
fundus of Greil’s (13) descriptions of embryos of this fish issues
from the cranium anterior to that bar. The nerve is then shown,
in one of Greil’s figures (l.c., fig. 8, pl. 55), separating into two
branches, one of which is evidently a portio ophthalmici profundi
and the other a typical so-called ophthalmicus profundus tri-
gemini. The ophthalmicus superficialis of this figure, called by
Greil the ophthalmicus superficialis trigemini in another figure
(fig. 4) on the same plate, receives no branch from what is ap-
parently the general cutaneous’ ganglion of the trigeminus, the
profundus thus supplying, as in Polypterus, all the general cu-
taneous fibers sent to the dorsal surface of the anterior portion
of the head. |
In Amia I have fully described the nerves here concerned,
without, however, definitely determining their components (Allis,
97). It is, however, probable that the ophthalmicus superficialis
contains lateralis, communis and general cutaneous fibers, and
it receives an important portio ophthalmici profundi from an
independent profundus ganglion, the root of the latter ganglion
issuing from the cranium anterior to the pedicel of the alisphenoid.
A delicate nerve that arises from the anterior end of the profun-
dus ganglion was considered by me to be a greatly degenerated
ramus ophthalmicus profundus.
OPHTHALMIC NERVES OF GNATHOSTOME FISHES iD
In Lepidosteus, van Wijhe (’82) describes both a ramus oph-
thalmicus profundus trigemini and a ramus ophthalmicus super-
ficialis trigemini. In a 75-mm. specimen of this fish I find the
so-called ophthalmicus superficialis trigemini composed of later-
alis, communis and general cutaneous fibers, the lateralis fibers
forming a bundle which lies somewhat above the communis and
general cutaneous ones. A radix profundi arises from the me-
dulla anterior, and close to the general cutaneous root of the tri-
geminus, and issues from the cranial cavity by an independent
foramen. A profundus ganglion forms on this root, and from it
a radix longa, a ramus ciliaris longa, and a portio ophthalmici
profundi arise. There is no perceptible trace of a ramus oph-
thalmicus profundus. The portio ophthalmici profundi runs
upward and joins and fuses with the communis and general cu-
taneous components of the ophthalmicus superficialis, as shown
but not index-lettered in Luther’s figure of this fish (13, fig. 1), but
as there is no pedicel to the alisphenoid of this fish (Allis, 09) the
relations of the nerve to that element of the skull are, as in Polyp-
terus, undefined. The conditions in this embryo thus show that
the so-called nervus ophthalmicus profundus of Landacre’s (712)
descriptions of a 10-mm. embryo of this fish is probably a portio
ophthalmici profundi and not a ramus profundus; and this nerve
accordingly cannot be, as Landacre concludes in a later work
(16, p. 27), a nerve comparable to the ramus ophthalmicus pro-
fundus of the Selachii.
In Acipenser: van Wijhe (’82) describes a so-called ophthalmi-
cus profundus trigemini, but as this nerve is said to run forward
dorsal to all the muscles of the eyeball, and is shown in his figure
lying dorsal to the nervus trochlearis, it must be either a portio
ophthalmici profundi, a ramus ophthalmicus superficialis tri-
gemini, or both those nerves combined. Which one it is cannot
be told either from his descriptions or from those of Gorono-
witsch (’83), for the profundus ganglion is completely fused with
the trigeminus ganglion.
In a 40-mm. specimen of Gasterosteus acus, in which fish there
is an anterior portion of the ascending process of the parasphenoid
that replaces the pedicel of the alisphenoid (Allis, in press), the
76 EDWARD PHELPS ALLIS, JR.
ophthalmic and maxillomandibular branches of the trigeminus
issue from the cranium posterior to that process, but two small -
branches that arise from the intracranial portion of the trigemi-
nus ganglion run forward in the cranial cavity and issue from it
anterior to the process. One of these two branches is the radix
longa. The other separates into two parts, one of which is the
ramus ciliaris longa and the other a portio ophthalmici profundi.
In a 37-mm. specimen of Cottus aspera, and in a 40-mm. speci-
men of Clinocottus, in both of which fishes there is also a process
of the parasphenoid that replaces the pedicel of the alisphenoid
(Allis, ’09), there is a radix profundi which issues from the cra-
nium anterior to that process and separates into a radix longa
and ramus ciliaris longa, but there is neither portio ophthalmici
profundi nor ramus ophthalmicus profundus. In each of these
three fishes the profundus ganglion is so completely fused with
the trigeminus ganglion that, while its general outlines can be
readily determined, it cannot be told whether or not there is
any exchange of fibers between the two ganglia. The conditions
nevertheless show that where there is both a portio ophthalmici
profundi and a process of the parasphenoid that corresponds to
the pedicel of the alisphenoid, the portio issues from the cranium
antericr to that process. Where there is neither pedicel of the
alisphenoid nor corresponding process of the parasphenoid, as
in most of the Teleostei, it might be assumed, as already stated,
that the portio still existed, but the conditions in the three fishes
above described, and those in Scomber (Allis, ’03), where there
is an independent profundus ganglion, but neither portio pro-
fundi nor ramus ophthalmicus profundus, tend to show that
both these nerves have wholly disappeared in most of the
Teleostei.
In a 22-mm. embryo of Squalus acanthias, Landacre (’16) finds
both a so-called ramus ophthalmicus profundus and a general
cutaneous nerve which he considers to be the ramus ophthalmi-
cus superficialis trigemini, but the manner of origin of this
latter nerve from the trigeminus ganglion is not incompatible
with its being a portio ophthalmici profundi. There is, how-
ever, an anterior prolongation of the profundus ganglion that
OPHTHALMIC NERVES OF GNATHOSTOME FISHES 77
strongly suggests this portio. Of this process Landacre says
that it “is evidently the remains of the structure which Neal
identifies as a persistent connection of the ganglion with the ec-
toderm, and which Scammon identifies as the utrochlea process,
i.e., the remains of the connection of this ganglion with the
neural crest.”’
The conditions in the several fishes above considered thus show
that there are two distinctly different nerves that may supply
the general cutaneous fibers that are distributed to the dorsal
surface of the anterior portion of the head. One of these nerves
is the so-called ramus ophthalmicus superficialis trigemini, the
other what I have called the ramus ophthalmicus profundi, the
frontal branch of the latter nerve being the So-called portio
ophthalmici profundi. The trigeminus one of these two nerves
always issues from the cranium posterior to the pedicel of the
alisphenoid, or posterior to a corresponding process of the para-
sphenoid, while the profundus always issues anterior to that
pedicel or process, and I consider these peripheral relations to
these structural elements to be as definite and positive evidence
of the segments to which the nerves belong as are the facts of
development and the central origins of the nerves.
The relative importance of these two ophthalmic nerves varies
greatly in different fishes, as does also the relative importance
of the frontal and nasal branches of the ramus ophthalmicus pro-
fundi, and it would seem as if the frontal branch alone ‘of the
latter nerve might be the serial homologue of the entire ophthal-
micus trigemini. In the Selachii, where there is apparently
both a ramus ophthalmicus profundi and. a ramus ophthalmicus
trigemini, the portio ophthalmici profundi is wholly wanting,
unless it be represented in some part of the so-called ophthalmicus
trigemini. In the Holostei and certain of the Teleostei there
is an ophthalmicus trigemini and a portio profundi, but the
ramus ophthalmicus profundi, if*present at all, is a small and
degenerate nerve (Amia). In certain others of the Teleostei
(Scomber) there is an ophthalmicus trigemini, but neither ramus
ophthalmicus profundi nor portio profundi. In Polypterus, and
probably also in Ceratodus, there is a ramus ophthalmicus pro-
78 EDWARD PHELPS ALLIS, JR.
fundi, with ramuli nasalis and frontalis, but no evident trace of
an ophthalmicus trigemini, excepting as it is represented in the
so-called ophthalmicus superficialis facialis; and this, with the
further absence of the ophthalmicus superficialis facialis, is
apparently the condition found in all higher vertebrates.
Herrick (’09) says that’ the facialis nerve of primitive verte-
brates was a branchiomeric nerve, supplying a gill-bearing seg-
ment and containing at least four components. It is not said
that one of these components was a general cutaneous one, but
that seems understood. In the mandibular arch similar condi-
tions must certainly have existed, and probably also in a pre-
mandibular arch supplied by the nervus profundus. In any
event, it is certain that the general cutaneous tissues of the region
innervated in Polypterus, Ceratodus, and the Selachii by the
nervus profundus are innervated, in the Holostei and Teleostei,
either by both the portio profundi and the ophthalmicus tri-
gemini, or by the latter nerve alone. There must then have
been, phylogenetically, either an actual change of innervation of
these tissues in one of these two groups of fishes or the tissues
innervated, in one of the groups, by one segmental nerve, must
have degenerated, with the related nerve, in the other group,
and there have been replaced by tissues primarily innervated
by the nerve of an adjacent segment. The presence, in Amia,
of a degenerate ramus ophthalmicus profundi would seem to
exclude the possibility of assuming that the remaining fibers of
the latter nerve have simply, in the Holostei and Teleostei, ac-
quired a different and more favorable course than that followed
by them in the Selachii. But however this may have been, it is
evident that, of the several fishes above considered, Polypterus,
the Dipneusti, and the Elasmobranchii alone present actual
conditions of these nerves that could have led to those found in
higher vertebrates, for that a nerve once so degenerated as the
ramus ophthalmicus profundi of the Holostei and Teleostei would
have been redeveloped and perpetuated seems improbable.
Palais de Carnolés, Menton, France
August 26, 1918.
OPHTHALMIC NERVES OF GNATHOSTOME FISHES 79
LITERATURE CITED
Auuts, E. P., Jr. 1897 The cranial muscles, and cranial and first spinal nerves
in Amia calva. Jour. Morph., vol. 12, Boston.
1901 The lateral sensory canals, the eye-muscles, and the peripheral
distribution of certain of the cranial nerves of Mustelus laevis. Quart.
Journ. Microse. Sci., vol. 45, London.
1903. The skull, and the cranial and first spinal muscles and nerves in
Secomber scomber. Jour. Morph., vol. 18.
1909 The cranial anatomy of the mail-cheeked fishes. Zoologica,
Bd. 22 (Hft. 57), Stuttgart.
1914 The pituitary fossa and trigemino-facialis chamber in Ceratodus
forsteri. Anat. Anz., Bd. 46, Jena. i
(In press), The myodome and the trigemino-facialis chamber of fishes,
and the corresponding cavities in higher vertebrates.
Goronowitscu, N. 1888 Das Gehirn und die Cranialnerven von Acipenser
ruthenus. Ein Beitrag zur Morphologie des Wirbelthierkopfes.
Morph. Jahrbuch, Bd. 13, Hfts, 3/4, Leipzig.
Greit, A. 1913 Entwickelungsgeschichte des Kopfes und des Blutgefasssys-
‘temes von Ceratodus forsteri. Zweiter Teil: Die epigenetischen Er-
werbungen wiihrend der Stadien 39-48. Jenaische Denkschriften, Bd.
4, Jena.
Herrick, C.J. 1909 The criteria of homology in the peripheral nervous system.
Jour. Comp. Neur., vol. 19, Philadelphia.
His, W. 1887 Die morphologische Betrachtung der Kopfnerven. Archiv f-
Anat. u. Physiol., Anatom. Abtheil., Leipzig.
Horrman, C. K. 1886 Weitere Untersuchungen zur Entwicklungsgeschichte
der Reptilien. Morph. Jahrb., Bd. 11, Leipzig.
Lanpacre, F. L. 1912 The epibranchial placodes of Lepidosteus osseus and
their relation to the cerebral ganglia. Jour. Comp. Neur., vol. 22, no.
ile
1916 The cerebral ganglia and early nerves of Squalus acanthias.
Jour. Comp. Neur., vol. 27, Philadelphia.
Luruer, A. 1913' Uber die vom N. Trigeminus versorgte Muskulatur der
Ganoiden und Dipneusten. Acta Soc. Scientiarum Fennicae, Tome
41, No. 9, Helsingfors.
Norris, H. W. 1913 The cranial nerves of Siren lacertina. Jour. Morph., vol.
24, no. 2, Philadelphia.
Pinxus, F. 1894 Die Hirnnerven des Protopterus annecten. Morphol. Ar-
beiten, Bd. 4, Jena.
Srannius, H. 1849 Das peripherische Nervensystem der Fische. Rostock.
Wie, J. W. van 1882 Uber das Visceralskelet und die Nerven des Kopfes der
Ganoiden und von Ceratodus. Niederl. Archiv fir Zoologie, Bd. 5,
H. 3, Leiden.
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. |
Resumido por el autor, Darmon Artelle Rhinehart.
El nervio facial del ratén albino.
La porcion motriz del nervio facial del rat6én de distribuye
sobre la musculatura facial e hioidea. El nervio intermedio se
compone de fibras aferentes y eferentes, y una parte de las
Ultimas forma una raiz separada. El ganglio geniculado contiene
los cuerpos celulares de las fibras aferentes, perteneciendo estas
células al tipo bipolar. El nervio petroso superficial mayor lleva
fibras aferentes y eferentes al ganglio esfenopalatino, que no recibe
fibra aleuna del trigémino; sus fibras terminan en la glindula de
Stenson, glandula lagrimal, glandula del tabique y vasos san-
guineos de la nariz y en el paladar. El autor no ha podido
determinar exactamente la inervacién de las glindulas palatinas
y botones gustativos. La cuerda del timpano envia fibras a los
ganglios de las glindulas submaxilar y sublingual y termina en
la lengua; lleva las fibras gustatorias de los dos tercios anteriores
de dicho 6rgano y probablemente desempena otras funciones.
El autor hallé cerca de mil células ganglionares, de relaciones y
funcién desconocida, en el trayecto de los nervios de una de las
mitades de la lengua. Las fibras cutdneas procedentes del
ganglio geniculado terminan en la piel del meato auditivo ex-
terno, piel de la oreja, y es posible también que inerven parte
de la membrana del timpano; estas fibras forman el ramo
cutdneo-facial, rama independiente de’ facial. Las fibras del
ramo auricular del vago forman una parte de este nervio y
fibras del cutdneo-facial se distribuyen por el ramo auricular
del vago. El nervio petrosomenor superficial falta en el rat6n.
Translation by Dr. José Nonidez,
Columbia University.
AUTHOR’S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 2
THE NERVUS FACIALIS OF THE ALBINO MOUSE
D. A. RHINEHART
Anatomical Laboratory of the Medical Department of the University of Arkansas
FOURTEEN FIGURES
CONTENTS
MURR ERSE RMS e's. so ccs cine Svs eee acim eel Se ewe sce ceases babe ek 81
ee Ie ep TPICE IOUS S).545 8 5 koe et EE eae Oe ER «hE iks GAs EME Silay 82
Perea n Gee RReRNer WS, LACIAlIS..... 0. sisi coseoremomhereniel Mirtle xe.’
er \ Ce:
MTenTym
Fig. 3 Sagittal section through the nervus facialis and the ganglion geniculi.
60 » lateral to figure 2. X 166.
N.Int.e.). Followed centrally from this position, it passes pos-
teriorly and medially for a short distance across the surface of
the pons. At the ventral edge of the spinal root of the trigemi-
nal nerve it enters the pons and passes dorsally and medially,
parallel with, but more lateral and posterior to the emerging
motor part of the facial nerve. At the level of the ascending
part of the facial nerve the compact bundle becomes broken up
and the fibers scattered, so that they cannot be followed farther.
92 D. A. RHINEHART
Peripherally, the efferent root of the nervus intermedius
passes around the anterior border of the motor part of the facial
nerve, under the rest of the nervus intermedius, through the
medial border of the geniculate ganglion and into the great super-
ficial petrosal nerve. It is certain that only a few, if any, of its
fibers are connected with the cells of the geniculate ganglion.
This part of the nervus intermedius is present in all my series
and undoubtedly contributes efferent fibers to the great super-
ficial petrosal nerve.
In the material used in this work it was impossible to tell what
proportion of the fibers of the nervus intermedius were medul-
lated and what non-medullated. All of the axis-cylinders are
small in diameter and the medullated sheaths must be very
thin. Weigner (’05) has given an excellent description of the
fibers of the nervus intermedius of the ground-squirrel. In
teased preparations stained with osmic acid he found that most
of them are very fine with very delicate myelin sheaths (2 u
in diameter), that a few are larger (5 uw), and that the nerve con-
tains nonmedullated fibers. The fibers of the nervus inter-
medius of the mouse are probably similar in size and character.
Weigner describes numerous anastomoses between the nervus
intermedius, the vestibular nerve, and the facial nerve, which
are characterized by the presence of ganglion cells, scattered
and in groups, along the anastomosing bundles. These ganglion
cells, he states, are similar to those of the geniculate ganglion
and furnish additional centers of origin for nervus intermedius
fibers. In the mouse, no ganglion cells are present central to the
dorsal part of the ganglion and no true anastomoses were observed.
That the afferent fibers of the nervus intermedius terminate in
the anterior extremity of the nucleus of the fasciculus solitarius
or a group of cells lying anterior to it seems well established.
All of the available papers dealing with this point in lower ver-
tebrates state this to be true. Van Gehuchten (’00) found this
to be the termination of these fibers in rabbits. The efferent
fibers to the submaxillary and sublingual glands arise from a
nucleus in the formatio reticularis at the level of the motor
facial nucleus. This nucleus was called the nucleus salivatorius
NERVUS FACIALIS OF ALBINO MOUSE 93
by Kohnstam (’02) and has been described by Yagita and Ha-
yama(’09). Herrick (’16) calls it the nucleus salivatorius superior.
The efferent fibers of the great superficial petrosal nerve arise
from other cells in the same region (Yagita, ’14).
In the mouse, three distinct bundles of nerve fibers continue
the nervus intermedius peripheral to the ganglion. From the
anterior angle of the ganglion the great superficial petrosal nerve
emerges, while from its lateral angle two bundles of fibers pass
peripherally in a common sheath with the motor fibers of the
facial nerve. Before the genu is reached these bundles lie along
the antero-ventral side of the facial nerve, the smaller of the two
lying ventral to the larger (figs. 9 and 10, N.Ch.Ty., R.Cut.N.F.).
In ‘sagittal series medial to the genu these bundles are always
more or less separated by delicate connective tissue septa from
each other and from the motor fibers of the facial nerve. Beyond
the genu they lie along the lateral side of the nerve (fig. 10). In
this position the septum is not as distinct as it is nearer the gan-
glion. They are well separated in some series, while in others an
indentation along their medial side is all that indicates a division
into two parts. The ventral and smaller of these two bundles
becomes the chorda tympani, the dorsal and larger contains
fibers which are distributed to the skin of-the auricle and form a
nerve which I have called the ramus cutaneus facialis.
With the exception of the efferent fibers of the great super-
ficial petrosal nerve, it was impossible in this material to tell
the exact relation of the fibers of the nervus intermedius to the
cells of the geniculate ganglion. That some of the fibers are
connected with the cells of the ganglion and others pass through
without interruption seems well established. |
Cutting the chorda tympani in the middle ear results in a
degeneration of about four-fifths of the cells of the geniculate
ganglion (Amabilino, ’98; DeGaetani, ’06). Nissl degeneration
of nerve cells in the brain stem after cutting the fibers to the
submaxillary and sublingual glands is proof that the chorda
tympani contains fibers which do not have their fibers in the
ganglion (Kohnstam, ’02; Yagita and Hayama, ’09).
94 D. A. RHINEHART
There is also considerable evidence for the presence of both
afferent and efferent fibers in the great superficial petrosal nerve.
By staining the geniculate ganglion of the mouse by the Golgi
method von Lenhossék (’94) found fibers which pass directly
from the nervus intermedius into the great superficial petrosal
nerve. From this he erroneously concluded that the great
superficial petrosal is a motor nerve for the supply of the levator
veli palatini and levator uvulae muscles. After cutting the
great superficial petrosal nerve in dogs, Yagita (’14) found that
about one-twelfth of the cells of the geniculate ganglion show
typical Nissl degeneration, and that there were degenerated
cells in the formatio reticularis of the same side of the pons
extending from the middle to the upper third of the facial nucleus.
Weigner (’05) mentions a bundle which passes directly from
the facial nerve to the great superficial petrosal through the
human geniculate ganglion. In the ground-squirrel he describes
a bundle between the nervus intermedius and the great super-
ficial petrosal nerve which has no connection with the ganglion
cells. He could not determine, he states, whether the fibers of
this bundle are processes of the scattered cells in the nervus inter-
medius or of those in the great superficial petrosal nerve.
It is safe to conclude, therefore, that both the chorda tympani
and the great superficial petrosal nerve contain afferent fibers
whose cell bodies are located in the geniculate ganglion, and effer-
ent fibers which pass through the ganglion without connection
with its cells.
BRANCHES OF THE NERVUS FACIALIS
1. Nervus petrosus superficialis major
It has been shown above that the nervus petrosus superficialis
major arises within the cranial cavity from the anterior pointed
extremity of the ganglion geniculi, and that it is composed of
fibers whose cell bodies are located in the ganglion, and others
which form a separate efferent root of the nervus intermedius.
From its origin this nerve passes anteriorly for a short distance |
along the lateral side of the ventral surface of the ganglion semi-
NERVUS FACIALIS OF ALBINO MOUSE 95
lunare. After a short course it bends at almost a right angle and
passes medially, ventral to the nervus trigeminus, and ventral to
the internal carotid artery and the sympathetic plexus which
surrounds it. At a point just medial to the artery the nerve
bends anteriorly and is joined along its medial side by the nervus
petrosus profundus, the two uniting to form the nervus canalis
pterygoidei.
A. Nervus petrosus profundus. The nervus petrosus profun-
dus, in the mouse, is formed by two small bundles of fibers from
the internal carotid plexus. It does not have a separate course
for the fibers join the great superficial petrosal nerve immediately
after leaving the plexus and while still in relation to the artery.
The fibers composing this nerve can readily be followed to the
interior of the superior cervical sympathetic ganglion. Along
the nerve of the pterygoid canal they occupy, at first, a position
on its medial side Farther anteriorly, however, they become
intermingled to such an extent with those from the great super-
ficial petrosal nerve that the fibers from the two sources cannot
be separately identified.
B. Nervus canalis pterygoidet. In mice, the internal carotid
artery enters the cranium and the nerve of the pterygoid canal
leaves it through a slit-like fissure between the tympanic and
periotic bones (eustachian aperture). From this fissure the nerve
passes anteriorly and slightly medially along the ventral surface
of the body of the sphenoid bone. One of the pharyngeal muscles
(probably the salpingo-pharyngeus), the pharyngeal opening of
the auditory tube and a mass of glands lie ventral to it. At the
anterior border of the opening of the auditory tube the muscle
and glands disappear and the nerve lies between the bone and
the mucous membrane of the nasal cavity. At this place the
pterygoid process extends ventrally to the soft palate, the nerve
lying medial to it where it fuses with the body of the bone above.
In this position the nerve extends anteriorly for some distance
finally passing laterally through a foramen into the most posterior
part of the orbit.
The nerve of the pterygoid canal is, at first, flattened dorso-
ventrally becoming circular in outline more anteriorly with the
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, No. 1
’
96 D. A. RHINEHART
fibers more loosely arranged. All of its fibers are very fine.
There are no ganglion cells either along the great superficial or
the deep petrosal nerves and only a few along the nerve of the
pterygoid canal. With the exception of one elongated microscopic
ganglion about the middle of its course, these are single and
widely separated.
In one series a few fibers were given off from the nerve of the
pterygoid canal to the mass of glands lying dorsal to the audi-
tory tube. Similar fibers could not be found in any other series,
nor were there other branches from this nerve.
In addition to the sympathetic fibers to the sphenopalatine
ganglion by way of the nervus petrosus profundus and the nervus
canalis pterygoidei, others from the internal carotid plexus
reach the ganglion by way of the nervus abducens. ‘These fibers
join the nervus abducens as two bundles slightly anterior to
the point of union of the two petrosal nerves.’ They leave the
nervus abducens as four small bundles, two of which join the
ophthalmic nerve and the other two the sphenopalatine ganglion
just posterior to its middle (figs. 4 and 7, bundles 6 and c).
Koch (’16) mentions the presence of sympathetic nonmedul-
lated fibers from the cavernous plexus in the nervus abducens of
the dog. These leave the nerve more anteriorly to pursue an in-
dependent course. Piersol (’13) states that branches of the
sphenopalatine ganglion have been described joining the nervus
abducens.
C. Ganglion sphenopalatinum. It will be shown later that
none of the fibers of the sphenopalatine nerves end in the spheno-
palatine ganglion. This ganglion belongs, therefore, more to
the facial than to the trigeminal nerve. Most of the fibers from
it, however, are distributed as constituents of the palatine
nerves. For this reason, it has been found necessary to study
the ganglion itself and the nerves connected with it in any way.
The sphenopalatine ganglion in the mouse is an elongated
mass of ganglion cells lying between the medial side of the max-
illary nerve and the medial wall of the orbit. It begins as a
small accumulation of cells along the ventral border of the nerve
of the pterygoid canal immediately after that nerve enters the
NERVUS FACIALIS OF ALBINO MOUSE 97
orbit. Anteriorly, the ganglion gradually increases in size, the
cells completely surrounding the nerve. This continues until
the anterior limit is reached where it ends in a blunt extremity.
The posterior half of the ganglion is flattened laterally and is,
in cross-section, an elongated oval with the long axis extending
Fig. 4 Transverse section through the most posterior part of the orbit show-
ing the nervus oculomotorius, nervus trochlearis, nervus ophthalmicus, nervus
maxillaris, nervus abducens, the posterior part of the ganglion sphenopalatinum
and the origin of the nervus palatinus posterior. By referring to figures 7 and
8 the connections of the nerve bundles marked with small letters and arabic
numerals will be seen. Figures 4, 5, and 6 are from the same series. X 90.
vertically (fig. 5). Near its middle the ganglion changes its
shape and becomes flattened dorsoventrally (fig. 6). At this
point a ventral bend in the ganglion together with its change in
shape imperfectly separates it into a posterior and an anterior
part.
98 D. A. RHINEHART
Medially the sphenopalatine ganglion is separated by the
bone from the nasal cavity. Laterally it is related to the max-
illary nerve throughout its entire extent, the sphenopalatine
nerves extending ventrally along its lateral side at about its
middle.
The sphenopalatine ganglion of mammals is described as receiv-
ing the sphenopalatine nerves from the maxillary nerve and as
giving off branches which are distributed to the mucous mem-
brane of the nose, mouth and pharynx, and branches to the orbit.
All of the nerves which are usually mentioned in connection with
the ganglion, with the exception of the pharyngeal branch, are
represented in the mouse by similar nerves.
a. Nervi palatinus posterior et medius. The posterior and
middle palatine nerves arise from the lateral side of the maxillary
nerve widely separated and distinct from both the sphenopalatine
ganglion and the sphenopalatine nerves. They will be described,
therefore, before the sphenopalatine nerves.
The posterior palatine nerve is represented by one or two small
nerves coming from the lateral side of the maxillary nerve at the
level of the middle of the pesterior part .of the sphenopalatine
ganglion (fig. 4, N.Pal.P.). It is separated by the maxillary
nerve from the sphenopalatine ganglion, and arises from a part
of the maxillary nerve which gives off branches for the supply of
the teeth and gums of the upper jaw.
The posterior palatine nerve receives two small bundles of fine
fibers from the ventral border of the sphenopalatine ganglion
(fig. 4, d,e). These pass laterally, ventral to the maxillary nerve,
and join the posterior palatine immediately before it enters the *
canal in the bone through which it passes to the palate. They
are so small that in some series it is difficult to follow them even
by using the high power of the microscope. Unsuccessful at-
tempts were made to count the fibers they contain. I believe it
safe to conclude, however, that they do not exceed thirty-five or
forty in number.
The middle palatine nerve is represented by a single smaller
branch which takes origin from the ventral side of the maxillary
nerve. This nerve passes inferiorly where it, too, enters a canal
NERVUS FACIALIS OF ALBINO MOUSE 99
in the bone and extends to the mucous membrane of the palate.
In some series it was possible to follow a minute bundle from the
sphenopalatine ganglion: into it; in other series this could not be
done. When present this bundle is even smaller than the ones
joining the posterior palatine.
The termination of these nerves will be discussed later.
Fig.5 Transverse section through the nervus maxillaris, the ganglion spheno-
palatinum, and the nervisphenopalatini. 810. anterior to figure 4. X 90.
b. Nervi sphenopalatini. The sphenopalatine nerves take
origin from the ventromedial part of the maxillary nerve at
about the middle of the sphenopalatine ganglion (fig. 5, bundles
a, b, c, d, e, g, 7, and part of f.). The bundles of fibers present
in these nerves vary in different series. In the one represented
in the graph, figure 7, and from which figures 4, 5, and 6 were
drawn, there were six distinct bundles at their point of origin.
100 D. A. RHINEHART
Almost immediately their arrangement becomes so complicated
that it cannot be easily described. ‘This arrangement is, how-
ever, accurately represented in the graph (fig. 7, N.Sp.). A
greater part of five of the six bundles assist in forming the an-
terior palatine nerve and its posterior inferior nasal branch, and
the greater part of the other bundle assists in forming the naso-
palatine nerve.
From these six bundles there are only two small subdivisions
that enter the sphenopalatine ganglion, joining it well toward its
anterior end and close to the point of origin from the ganglion
of the fibers which enter the nasopalatine nerve. In transverse
series these bundles can be followed through the ganglion into
the nasopalatine nerve, which is probably their fate in the other
series.
ce. Nervus palatinus anterior. The anterior palatine nerve is
the largest of the nerves in this region. The bundles from the
maxillary nerve which form it pass medially and anteriorly across
the ventral border of the sphenopalatine ganglion. While in
this position they are joined by four bundles whose fibers come
from the interior of the sphenopalatine ganglion. The fibers
from the ganglion are smaller than those from the maxillary
nerve and represent about one-third of the fibers of the anterior
palatine nerve.
From the ventral border of the sphenopalatine ganglion the
anterior palatine nerve passes ventrally through a canal to reach
the mucous membrane of the palate. In this canal one bundle
separates from the nerve, and as soon as the mucosa is reached,
takes a posterior direction. The remainder of the nerve passes
anteriorly just lateral to the middle of the hard palate.
The posterior inferior nasal branch of the anterior palatine
nerve is made up of two small bundles of fibers from the spheno-
palatine nerves (fig. 6, a, 6) and four smaller bundles of fine fiber
from the sphenopalatine ganglion (fig. 7). This nerve extends
anteriorly along the lateral wall of the nasal cavity, the smaller
fibers soon leaving it to enter a mass of glands in the lateral wall
of the nose (gland of Stenson). The fibers of larger size end
in the mucous membrane of the lateral wall of the nasal cavity.
NERVUS FACIALIS OF ALBINO MOUSE 101
d. Nervus nasopalatinus. One bundle of the sphenopalatine
nerves passes obliquely forward and medially across the inferior
surface of the anterior part of the sphenopalatine ganglion,
finally reaching the medial border of the anterior end of the gan-
glion (fig. 5, bundlea). In this position it receives a large contri-
bution of fine fibers from the ganglion and becomes the nasopala-
J
s, 3
* $8 20% o90 20. 3
S9ggsee Zoos Reese Oe
GarL2039 26 6®.
%% 8
01 Oe
Cae ®
Fig. 6 Transverse section through the nervus maxillaris and the anterior
extremity of the ganglion sphenopalatinum showing the origin of the nervus naso-
palatinus and two of the orbital branches of the ganglion. 600 » anterior to
figure 5. X 90.
tine nerve (fig. 6). The fibers from the ganglion form a little
more than half of this nerve.
From its origin the nasopalatine nerve extends medially along
the posterior wall of the inferior meatus of the nose to reach the
posterior and ventral part of the septum. From this point its
course is anterior along the ventral part of the septum. Through-
102 D. A. RHINEHART
out its extent branches are given off, some of coarse fibers to the
mucous membrane, and others composed of finer fibers from the
sphenopalatine ganglion which pass into the glands of the sep-
tum. The septal glands are numerous in the region posterior to
the vomeronasal organ. Finally, reduced in size and composed
only of fibers of larger size, the nasopalatine nerve passes through
the incisive foramen to reach the roof of the mouth.
e. Other branches. In addition to the fibers from the spheno- —
palatine ganglion described above, there are several other small
nerves either arising or ending within it. One small nerve ex-
tends from the ganglion posteriorly along the medial side of the
semilunar ganglion (fig. 7). Several other small bundles extend
from the sphenopalatine ganglion dorsally and laterally to join
the ophthalmic nerve. (These are not shown in the graph, fig.
7.) The origin, the direction or the endings of these nerves
could not be determined.
Several small nerves take origin from the anterior part of the
sphenopalatine ganglion and pass dorsally along the medial wall
of the orbit. Some of these join the nasociliary nerve, and
through it finally terminate among the gland ducts of the an-
terior and inferior part of the lateral wall of the nasal cavity.
‘Others accompany a blood-vessel into the medial part of the
lacrimal gland, while one other joins a small branch of the maxil-
lary nerve which terminates in the nasolacrimal duct.
Other branches from the anterior end of the ganglion join the
arteries in the neighborhood and accompany them in their
peripheral courses. .
f. The nerve supply of the palate. The nerve supply of the
palate was investigated to determine, if possible, the relationship
between the nerves of the taste-buds, the nerves to the palatal
glands, and the facial nerve. Although no positive conclusions
were reached, certain features are of interest and will be recorded
here.
From the ventral ‘end of its bony canal the anterior palatine
nerve extends anteriorly, in company with.a medium-sized ar-
tery, as far forward as the nasopalatine canal. Throughout its
entire extent it sends branches medially and laterally to supply
NERVUS FACIALIS OF ALBINO MOUSE 103
one-half of the palate. Anterior to the nasopalatine canal the
mucous membrane of the medial part of the palate is supplied by
the terminal branches of the nasopalatine nerve, that of the
lateral part by a branch of the maxillary nerve.
Branches from these nerves are abundant under the transverse
ridges of the palate and particularly so around the nasopalatine
canal. In the tunica propria the nerve fibers form rather wide
meshed plexuses from which individual fibers can be followed
into the epithelium. No encapsulated nerve endings were
found. In this areaof the palate there are no taste-buds or
glands. From the anterior palatine nerve many fine fibers can
be followed along the walls of the blood-vessels, presumably to
end in their muscle fibers, although this termination could not be
positively identified.
The graph (fig. 8) shows the distribution of the middle and
posterior palatine nerves and the posteriorly directed branch of
the anterior palatine nerve to the posterior part of the hard pal-
ate and the soft palate. The transverse line X shows the place
of union of the hard and soft palate. The graph includes one-
half of the palate, the right margin representing the median line
and the left margin the lateral edge. In this connection it
must be remembered that the anterior part of the palate shown
in the upper part of the graph is wider than the posterior part,
thus accounting for the apparent preponderance of nerves in the
upper part of the graph. The small circles show the approximate
locations of the taste-buds, and the broken line represents the
limits of the palatal glands.
In the series represented in the graph, there were twenty-five
taste-buds, twelve in the posterior part of the hard palate, and
thirteen in the soft palate. In other series the number is ap-
proximately the same and the arrangement is similar. They
are. most numerous along the posterior part of the hard palate
and gradually decrease in number posteriorly.
The nerves of this region can be followed from their emergence
from the bony foramina to their termination. Those supplying
the epithelium can be traced to the basement membrane and
some of the fibers were seen to pass among the epithelial cells.
104 D. A. RHINEHART
Fig. 7 A graph, made according to the method devised by Prof. A. G. Pohl-
man, showing the connections of the nervus petrosus superficialis major and the
ganglion sphenopalatinum. A graph of this sort is made by selecting that part
of a series which is to be used, and marking out-on one edge of a piece of milli-
meter plotting paper a set of stair steps for each row of sections on the slides,
representing each section by a single step. Nerves or other structures are repre-
sented in the graph by broken or solid lines, dots, etc., and as these structures con-
tinue through the series the lines or dots are continued on the paper, passing
through a millimeter for each section. By continuing this and representing
changes in position, branchings, anastomoses, etc., by changes in the lines, al-
most any structure which passes for some distance through a series can be graph-
ically shown.
The graph shown as figure 7 represents certain nerves and ganglia and their
connections in slides 16 to 24 of a tramsverse series of one half of a mouse head.
The numbers on the left of the graph are the numbers of the slides and the rows
on each slide. Each section is represented by a single step in the stairs. As an
illustration, the ganglion geniculi is present in section 1, row 1, slide 16, and is
shown at the lower right part of the graph. The ganglion ends and the nervus
petrosus superficialis major begins in the last section in that row. This nerve
continues through the second into the third row where it bends medially, this
being shown by a bend in the line representing the nerve. In section 8, row 1,
slide 17, and in section 2 of the second row the nervus petrosus profundus joins the
nervus petrosus superficialis major to form the nervus canalis pterygoidei. The
nervus canalis pterygoidei continues through the series until row 3, slide 21, is
reached where the ganglion sphenopalatinum begins, the ganglion being repre-
sented by an unshaded area continuing the course of the nerve.
Bundles of sympathetic fibers and branches of the ganglion sphenopalatinum
are represented by broken lines, cranial nerves and their branches by solid lines,
and ganglia by unshaded areas along or within the nerves. Figures 4, 5, and 6
were drawn from sections of thé same series used in making the graph, the loca-
tion of these sections is indicated by the broken transverse lines. The small
letters along the lines indicate bundles of nerves which are shown and similarly
labeled in the figures.
Fig. 8 A graph showing the distribution of the a ee and middle palatine
nerves and a part of the anterior palatine nerve to one half of the soft palate
and the posterior part of the hard palate, made from the same slides which were
used in making the graph shown as figure 7. The small circles indicate the ap-
proximate locations of the taste-buds and the broken line the limits of the pala-
tal glands. The arabic numerals indicate branches of the nerves which are shown
under the mucosa of the palate in figures 4 and 5. The broken transverse line
marked x indicates the place of junction of the hard and the soft palate. For
further explanation see the text. ’
105
ALBINO MOUSE
FACIALIS OF
NERVUS
1
Sy
Nt
K iD
NY
is
aN
a dy .
=O
S Lae
N
SS \ NX
eS \
NS Ne
\ \
Is =
W ~
\\
. i
:
Wout i
¥y i f
1 im 1
Fw AT
=! Hatt |e aes
EEO: | 4 reer
= liner! a
= e
Pa le )
| Pa
X ries
eels
Lm
io)
N.Max .R Inf.Na.
23
106 D. A. RHINEHART
On these nerves no encapsulated nerve endings were found. The
nerves of the glands are not so well demonstrated. Bundles of
fibers can be traced until they break up into smaller bundles or |
isolated fibers among the gland alveoli. The nerve supply of
the taste-buds, with one exception, could be identified as coming
from one or the other of the palatine nerves.
In the series represented in the graph the posterior palatine
nerve is represented by two relatively large nerve bundles.
These emerge through separate canals in the bone. One of
them is directed anteriorly and supplies a small part of the lateral
side of the hard palate, the other passes posteriorly to supply a
limited area of the hard palate and one entire half of the soft
palate. The middle palatine nerve is distributed to the lateral
part of the hard palate anterior and lateral to the place of emer-
gence from its canal. Two bundles of the anterior palatine nerve
are shown, one passing medially and the other posteriorly, to
supply an area of the hard palate medial to that supplied by the
other nerves. Practically the entire mass of the palatal glands
and all but a few of the taste-buds are located in the area sup-
plied by the posterior. palatine nerve.
In the nerves connected with the sphenopalatine ganglion
there are two varieties of fibers, those of large size from the
maxillary nerve, and those of small size from the sphenopalatine
ganglion. In the nasopalatine nerve and in the posterior inferior
nasal branch of the anterior palatine the fine fibers are grouped
and leavethe nerves as separatebranches. Thesebranches enter
glands and are assumed to contain glandular fibers although they
are not stained within the glands and could not be followed to
their final termination.
In the posterior branch of the anterior palatine and in the
middle and posterior palatine nerves the fine fibers are inter-
mingled with those of larger size and terminate in company with
the larger fibers. Except certain of the fine fibers which enter
the walls of the blood-vessels the endings of the finer fibers could
not be determined. In the area of the palate supplied by the
posterior palatine nerve the fine fibers presumably supply the
palatal glands, the muscle fibers in the walls of the blood-vessels,
NERVUS FACIALIS OF ALBINO MOUSE 107
and the taste-buds, if these structures receive their innervation
from the facial nerve. From what has been said above, however,
concerning the minute bundles of fine fibers from the sphenopala-
tine ganglion to the posterior palatine nerve, it does not seem
possible that they are numerically sufficient for the supply of all
of these structures. The only other possible source for their
nerve supply is from the sympathetic or the trigeminal nerve.
The evidences from comparative anatomy do not materially
assist in clearing up this problem. In petromyzonts, Johnston
(08) found that the maxillary nerve supplied the roof of the
mouth. The palate of bony fish (Herrick, ’99, ’00, ’01) is sup-
plied by the ramus palatinus VII. In the amphibians Coghill
(01, 702) and Norris (’08, 713) describe anastomoses between the
ramus palatinus VII and branches from the fifth nerve, the re-
sulting nerves being distributed to the roof of the mouth, the
teeth and the nasal capsule. In none of these forms is there a
sphenopalatine ganglion present, although Johnston (’08) de-
scribes ganglion cells in the roof of the mouth in cyclostomes, and
Norris (08), in Amphiuma means, mentions ganglion cells at cer-
tain points on the anastomoses between the ramus palatinus VII
and ramus ophthalmicus profundus V. Norris, in Siren lacertina
(13) also describes branches from the ramus es an VII to
the vessels of the roof of the mouth.
In this connection it is interesting to note that Herrick (16,
p. 248) says, ‘‘Unlike the visceral. sensory system, however, its
(referring to the gustatory apparatus) peripheral fibers have no
connection with the sympathetic nervous system and the reac-
tions may be vividly conscious.” If this statement be literally
true, then the taste-fibers of the palate must come from the
trigeminal nerve. This is hardly probable, for Herrick (’01) has
shown that in the siluroid fishes, where the taste-buds are very
numerous, none of them are directly supplied by the trigeminal
nerve. If this statement means that taste-fibers have no con-
nection with sympathetic ganglion cells, then it is possible for
the taste-buds of the palate to be supplied by fibers from the
facial nerve which pass through the sphenopalatine ganglion
without interruption.
108 D. A. RHINEHART
In conclusion it can be said that, in the mouse, the epithelium
of the palate is supplied by fibers from the maxillary nerve, and
that the muscle fibers in the walls of the blood-vessels are sup-
plied by fibers from the sphenopalatine ganglion. It does not
seem possible for the taste-buds and glands of the palate to be
supplied by fibers from the sphenopalatine ganglion. The evi-
dences that they are supplied by the trigeminal nerve is equally
inconclusive. In the absence of other direct anatomical obser-_
vations on this problem in mammals it must, for the time being,
remain an unsettled question.
Until recently the nerve supply of the m. tensor veli palatini
and the m. levator uvulae was believed to come from the facial
nerve. In the mouse the branch from the mandibular nerve to
the internal pterygoid muscle passes ventrally around the sta-
pedial artery and, after supplying the internal pterygoid, sends
one branch into the levator veli palatini and another which
passes ventral to the otic ganglion and terminates in the tensor
tympani muscle. The tensor veli palatini and the levator uvulae
muscles are supplied by a branch from the accessory portion of
the spinal accessory nerve which sends fibers into these muscles
and terminates in the muscles of the pharnyx.
2. Nervus stapedius
The first branch of the facial nerve peripheral to the genicu ate
ganglion is the nervus stapedius which arises in the dorsal wall
of the tympanic cavity. In the mouse this is undoubtedly a
motor nerve. The fibers composing it come from the medial
side of the facial stem (figs. 10, 11, and 12, N.Stap.). They
unite into a small bundle at the junction of the medial and dorsal
borders of the facial nerve and pass into the stapedius muscle,
to terminate after the manner of motor nerves elsewhere. There
are no ganglion cells along its course, nor does it contain fine
fibers.
Weigner (’05), in the ground-squirrel (Ziesel), describes a
ganglion at the place of origin of the nervus stapedius and states
that it contains many fine fibers similar to those of the nervus
NERVUS FACIALIS OF ALBINO MOUSE 109
intermedius. He also reports that cutting the facial nerve where
it passes across the cochlea did not result in a degeneration of
all of the nervus stapedius although there is a degeneration of
the entire facial nerve distal to the cut. He offers no explanation
for this, other than the implied one that the nervus stapedius is
composed mostly of fibers arising in the microscopic ganglion
located at its place of origin from the facial nerve.
sifieg fife My sf BN
yaa se bON Vest.
Fig. 9 Sagittal section through the nervus facialisat the lateral edge of the
ganglion geniculi and medial to the genu externum showing the two bundles of
fine fibers which pass from the ganglion peripherally in the nervus facialis. Same
series as figures 1, 2, and 3. 225 u lateral to figure 3. x 166.
Fig. 10 Transverse section through the nervus facialis just posterior to the
genu externum and through the origin of the nervus stapedius. The two bundles
of fine fibers are shown along the lateral side of the nerve. X 166.
Remembering that the mouse and the ground-squirrel both
belong to the order rodentia, an order of mammals in which
there is little variation in anatomical structure, I am unable to
account for the differences in this nerve in the two animals.
110 D. A. RHINEHART
3. Nervus chorda tympani
The chorda tympani arises from the ventral side of the facial
nerve along the posterior part of the upper wall of the tympanic
cavity between the origin of the nervus stapedius and the anas-
tomosis with the ramus auricularis vagi. It is formed by the
more ventral and smaller of the two bundles of fine fibers which
come from the interior of the geniculate ganglion and pass pe-
ripherally as a part of the facial stem (figs. 9, 10, 11, 12, and 138,
N.Ch.Ty.). It has been shown above that it contains fibers
which are processes of cells in the geniculate ganglion and others
which pass through the ganglion without interruption.
In the mouse, soon after its origin, the chorda bends medially
and anteriorly and passes through a fissure-like opening into the
tympanic cavity. Here it extends anteriorly in a shallow groove
along the upper part of the lateral wall, and then along a groove
in a spicule of bone which projects into the cavity. At the apex
of this spicule the nerve crosses a small gap to reach the medial
surface of the head of the malleus, across which it passes lying
ventral to the attachment of the tendon of the tensor tympani
muscle. In the anterior part of the cavity it passes along the
medial side of the anterior process of the malleus, accompanying
that process into the fissure between the tympanic and periotic
bones (petrotympanic fissure) through which it extends into the
infratemporal fossa. In the infratemporal fossa the chorda tym-
pani passes medially and ventrally posterior to the emerging
mandibular nerve. After a course of varying extent it joins the
posterior aspect of the lingual nerve. .
Just before joining the lingual there is usually an anastomosis
with a bundle of fibers from another source. This is, in a ma-
jority of my series, a small branch arising from the mandibular
nerve. In one series a branch from the auriculotemporal nerve
joins it; in another a branch from the lingual joins it, the two
then uniting with the lingual; in one series there is no anastomo-
sis of any sort. All of the branches which anastomose with the
chorda tympani are composed of fibers of larger size than those
in the chorda. When followed centrally they become lost in
the trunk of the mandibular nerve.
NERVUS FACIALIS OF ALBINO MOUSE is
In none of the series is there a connection of any sort with the
otic ganglion.
Weigner (05) mentions an almost constant anastomosis be-
tween the lingual and mandibular nerves in man, the chorda
joining the lingual just distal to it. He also describes a com-
munication from the great superficial petrosal nerve to the fibers
in the facial which form the chorda, and states that there are
many scattered ganglion cells along the chorda tympani both
before and after its origin from the facial. In the mouse no such
communication can be identified. With the exception of one
series, in which there is a small ganglion at its place of origin from
the facial nerve, no ganglion cells are present in the chorda.
A. The nerve supply of the tongue and the salivary glands. In
one transverse series the nerves in the tongue were carefully fol-
lowed. The results of this study will be briefly presented.
The hypoglossal nerve is composed of fibers of uniform size re-
sembling closely those in the motor portion of the facial nerve.
Immediately after entering the tongue it divides into two branches
which run anteriorly, one close to the septum and the other
farther laterally. From these, small branches are given off in
all directions, many of which were followed to their termination
in motor nerve endings in the muscle fibers of the tongue.
The glossopharyngeal nerve is composed of fibers of medium
and small size resembling those of the lingual nerve. It enters
the tongue between the hyoid bone and the thyroid cartilage
and passes anteriorly for a short distance along the side of the
base of the tongue. In this part of its course there is, in all my
series, a ganglion almost as large as the otic and having the his-
tological characteristics of a sympathetic ganglion. Just ante-
rior to this ganglion the nerve breaks up into five bundles which
enter the tongue and spread out in a fan-shaped manner.
The most medial of these passes medially and posteriorly and
supplies the mucous membrane of the tongue as far as the larynx.
The next passes medially and anteriorly and supplies the dorsum
of the tongue in the region of the single circumvallate papilla.
This branch sends a number of fibers into the base of the papilla,
which, together with a similar contribution from the opposite
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 1
En? D. A. RHINEHART
side, supply this structure. These fibers form a plexus exter-
nal to the valley and within the core of the papilla. From these
plexuses fibers supply the taste-buds and: the epithelium of the
papilla and the surrounding mucous membrane. In two series
a microscopic ganglion is present within the papilla.
The most lateral of the five branches of the glossopharyngeal
nerve extends anteriorly along the lateral side of the tongue and
terminates in the foliate papillae. The other two bundles are
distributed to the tongue over an area extending as far forward
as a line connecting the anterior limits of the foliate and cir-
cumvallate papillae.
The lingual nerve, because of the fibers in it from the facial
nerve, was carefully studied. From one transverse series a
graph of the course, branches, and distribution of this nerve
within the tongue was made, and all the branches, as far as pos-
sible, followed to their termination.
From its union with the chorda tympani the lingual nerve
passes medially, ventrally, and anteriorly to reach the tongue.
It lies between the internal and external pterygoid muscles, then
between the internal pterygoid and the mandible, and finally
between the mylohyoid and the side of the tongue. It curves
around the ducts of the submaxillary and sublingual glands, turns
anteriorly lateral to the genioglossus, and inclines dorsally into
the substance of the tongue.
A few branches are given off from the lingual nerve before it
enters the tongue. At the ventral edge of the internal pterygoid
and under the mylohyoid muscle a number of small branches are
given off which pass posteriorly to the submaxillary and sublin-
gual glands. At the ventral edge of the internal pterygoid a
larger branch arises which passes medially to supply an area of
the mucous membrane of the cheek opposite the folliate papil-
lae. This branch sends fibers into the epithelium, into a mass of
glands, and into a number of taste-buds (eight in the series
studied). Along the side of the tongue, in relation to the gland
ducts, one relatively large and a number of smaller branches are
given off which pass forward in company with the ducts and
supply the mucous membrane of the floor of the mouth andthe
gum behind the lower incisor teeth.
NERVUS FACIALIS OF ALBINO MOUSE 113
In the tongue the lingual nerve passes forward midway be- -
tween its lateral edge and the median septum. One branch is
given off immediately .after the nerve enters the tongue. This
passes dorsally and posteriorly to supply the area in front of
that supplied by the glossopharyngeal nerve. The remaining
branches arise irregularly and supply all of the tongue in front
of this area. Some of the branches extend laterally, some to the
ventral surface, but. the larger and most numerous branches pass
to the dorsal surface. In the ventral part of the tongue a few
small branches join those of the hypoglossal nerve, and a few
others end in an undetermined manner among the muscle fibers.
The majority of the branches, however, could be followed to the
mucous membrane.
In the series studied the taste-buds are located in both waiis
of the valley around the circumvallate papilla, in the foliate
papillae, in a small area of the cheek opposite the foliate papillae,
and irregularly scattered over the dorsal surface of the tongue
in the area supplied by the lingual nerve. There are no scat-
tered taste-buds in the area supplied by the glossopharyngeal
nerve nor along the sides or the ventral surface of the tongue.
In the area supplied by the lingual nerve there are thirty-six
taste-buds in one-half of the tongue. These are widely separated
posteriorly and become more numerous as the tip of the tongue
is approached. Each of them is placed near the surface of the
epithelium on the top of a flat tunica propria papilla.
One or two bundles of nerve fibers from the lingual nerve were
followed into the bases of the papillae on which the taste-buds
are located. In the papillae these fibers break up into a number
of branches and form an intricate complex of very fine fibers.
From these plexuses some fibers pass into the taste-buds. Other
fibers pass into the surrounding mucous membrane, so that the
taste-bud and a considerable area of the mucous membrane is
supplied from the plexus in each papilla.
The most striking feature of the nerves within the tongue is
the large number of ganglia and ganglion cells along them. The
majority of these cells are in clumps or microscopic ganglia along
the nerve bundles. Along the glossopharyngeal nerve there are
114 D. A. RHINEHART
six of these ganglia, along the lingual nerve they are far more
numerous, being forty-one in number. Along the hypoglossal
nerve there are scattered smaller ganglia and isolated cells. Fib-
ers from the lingual, the glossopharyngeal and even the hypo-
glossal nerve can be followed into these ganglia to end in an unde-
termined manner among the ganglion cells.
A count of the cells in these ganglia along the lingual and hy-
poglossal nerves was made. Only those in the ganglia in which
there is a definite nucleolus were counted, so that the figures
obtained are less rather than more than the actual number
present. In this one series, in one-half of the tongue, there were
1079 cells, 957 along the lingual and 122 along the hypoglossal
nerve.
While the presence of ganglion cells in the tongue has long
been known, I have failed to find a reference as to their number
or their significance. They are usually dismissed with the state-
ment that they are sympathetic cells. I am of the opinion that
much will be added to the knowledge of the nerve supply of the
tongue when the central connections, the endings of the fibers,
and the functions of these ganglion cells have been worked out.
In this work nothing has been found other than in support of
the generally accepted view of the nerve supply of the tongue.
That the taste-buds are supplied by the chorda tympani has
long been known. The course of the taste-fibers into the brain
was, however, for a time, an unsettled question. From clinical
cases evidences have been deduced in support of every possible
pathway for these fibers into the brain stem. Many of these
are included in the review of the literature in the articles by
Cushing (’03), Weigner (’05), and Sheldon (’09).
The carefully conducted experiments, of Cushing (’03) have
done much to prove that the taste-fibers enter the brain over
the nervus intermedius. In the mouse the anatomical evidence
supports this view. No connection, such as was found by Weig-
ner (’05) between the chorda tympani and the great superficial
petrosal nerve is present, and the entire absence in all my series
of a communication from the facial nerve to the tympanic plexus
excludes even the possibility that the taste-fibers reach the
brain except through the nervus intermedius.
NERVUS FACIALIS OF ALBINO MOUSE 115
Cushing (’04) has presented very convincing clinical eviderce
that the chorda tympani supplies the tongue with certain forms
of common sensation: After trigeminal neurectomy he found
that sensations of pain and temperature and tactile sensations
are absent in the area supplied by the lingual nerve. There re-
mained, however, the ability of the patient to appreciate the
presence, the general location, and the movement of a piece of
cloth or a cotton swab across this area.
The nerve supply of the submaxillary and sublingual glands in
the mouse corresponds closely to that given by Langley (’90) and
Huber (’01) for the dog and cat. In the mouse the sublingual
or the retrolingual gland is located lateral and ventral to the sub-
maxillary and is pure mucous in type while the submaxillary is
pure serous.
The nerves to these glands leave the lingual as a number of
branches (five to eight) arising deep to the mylohyoid muscle
ard rather widely separated. These nerves come into relation
with the ducts and pass with them into the glands. Along these
rerves there are several small ganglia. In the series most care-
fully studied there were five of these, two sending their fibers
along the submaxillary duct, one along the sublingual duct, while
the other two send fibers along the ducts into both glands.
Within the submaxillary gland there are two large ganglia and
' several smaller ones. The ganglion cells in these ganglia are so
‘ numerous that if each is supplied with a basket-work from the
chorda tympani as described by Huber, each chorda fiber must
divide and terminate in relation to several cells. 33.
After the cutaneous branch of the facial and the auricular
branch of the vagus have separated from the facial stem, each
nerve contains fibers from the other. Because the greater pro-
portion of the fibers of one nerve come from the facial, it is con-
sidered as a branch of that nerve, the other for the same reason.
being considered as a branch of the vagus.
118 D. A. RHINEHART
0
N.to TymMemb.
BL ONCh-Ty.
N.to Dia,
ies N.to Sty-Hy,
13
Fig. 13 Outline drawings of every other section through the nervus facialis
in a horizontal series, beginning above and passing ventrally through the anas-
tomosis with the ramus auricularis vagi, to show the mixing of the cutaneous facial
fibers with those from the vagus. The motor facial fibers are unshaded, the fibers
from the ganglion geniculi are black, and the vagus fiber cross-hatched. X 100.
NERVUS FACIALIS OF ALBINO MOUSE 119
In different series there are some differences in the arrange-
ment of these nerves. In one series the ramus auricularis vagi
is composed of a single large bundle, in another the ramus cu-
taneus facialis is composed, after its origin, of two bundles,
and there are apparently some differences in the number of
fibers which interchange. However, the essential arrangement
is constant and as described above.
Immediately after its origin the ramus cutaneus facialis passes
from the medial to the lateral side of the posterior auricular
nerve. From here it extends anteriorly and laterally above the
external auditory meatus until the point of the attachment of
the cartilage of the auricle to the side of the cranium is reached.
At this point it enters the auricle, extending upward and anteri-
orly along the medial side of that part of the auricle which bounds
the pouch-like concha laterally (figs. 14A and 14B). It gives
off branches in this region which supply the skin and hair follicles.
Other branches pass dorsally beyond the concha and supply the
skin of the posterior third of the lateral surface of the auricle.
The ramus auricularis vagi sends a small branch to supply a
part of the external auditory meatus and the tympanic membrane,
this branch containing a few fibers from the facial (fig. 13D).
Otherwise the nerve supply of the tympanic membrane is the
same as that given by Wilson (’07, 10-11). After the origin
of the branch to the tympanic membrane the auricular branch
of the vagus passes along the cranial side of the concha. It
supplies this area of skin and its terminal branches pass dorsally
beyond the concha to supply the anterior two-thirds of the
lateral surface of the auricle (fig. 14).
These two nerves have been followed with great care from their
origin to their place of ending. Each, in that part of its course
where it is related to the auricular cartilage, passes between the
cartilage and the skin, a location where there are no muscle
fibers. Branches of each have been followed to their final end-
ing and have been found to end in a plexiform manner immedi-
ately under the epithelium or around the hair follicles. It is
safe to conclude that they are both common sensory in function
for the supply of the skin and hair of the auricle.
120 D. A. RHINEHART
The anatomical and clinical evidence for the presence of gen-
eral cutaneous fibers in the nervus facialis is meager in amount
and inconclusive. In petromyzonts, Johnston (’08) describes gen-
eral cutaneous fibers arising from the geniculate ganglion and
passing to the ventrolateral surface of the head below and behind
Cervical-coe
Vogus “#”
Facial x%*
- 14
Fig. 14 A A diagram of the connections and distribution of the nervus facialis
of the albino mouse, based largely on a flat reconstruction made by projection of
sagittal sections on to a median plane.
14B_ A diagram of a coronal section through the auricle to show the areas of
skin suvvlied bv the vagus, facial and cervical nerves.
the orbit. In bony fish, Herrick (’99, ’00, ’01) found a general
cutaneous component in the facial nerve for the region of the
operculum, the fibers, however, being derived from the Gasserian
ganglion. In the amphibia, Norris (’13) found a general cutane-
ous component in Siren, and Herrick (’14) described fibers from
the geniculate ganglion in Amblystoma larva which enter the
‘NERVUS FACIALIS OF ALBINO MOUSE 13t
spinal V tract, this tract being considered as a tract whose
nucleus is generally cutaneous in function.
Van Gehuchten (’98) found that a few cells of the geniculate
ganglion degenerate after cutting the facial nerve at its exit from
the facial canal. In an abstract of an article by DeGaetani (’06)
sensory fibers in the ramus temporofacialis are, mentioned.
Weigner (’05) describes fibers of small caliber, presumably in-
termedius fibers, in the facial nerve of man distal to the origin
of the chorda tympani.
The chief clinical evidence for the presence of such fibers in
the facial nerve has been presented by Hunt (15). By a method
which he calls the herpetic method and which is based on the fact
that herpetic vesicles on the skin are caused by pathological proc-
esses involving the ganglion from which the cutaneous fibers
arise, he attributes the supply of certain parts of the auricle,
including the concha, to the facial nerve. One of the probable
pathways to the skin for these fibers which he mentions is through
the auricular branch of the vagus. In the different cases reported
the areas showing the herpetic eruption vary to some extent. If
homologous conditions are found in man to those here reported
in the mouse this variation can probably be accounted for by the
mixing of the vagus and facial fibers at the point of the anastomosis.
In the mouse there is no evidence that either the inner or
middle ear receive fibers from the facial nerve. A communicating
branch from the facial nerve to the tympanic plexus is not
present in any of my series.
5. Nervus auricularis posterior and the nerves to the stylohyoid and
digastric muscles
The posterior auricular nerve arises from the lateral side of
the facial nerve just distal to the origin of the ramus cutaneus
(figs. 11 and 12,R.Aur.P.). It passes dorsally behind the auricle
and is distributed to the muscles posterior to and along the
cranial side of the auricle. It is composed of fibers from the
motor part of the facial and is a purely motor nerve. A similar
area of skin of the auricle is supplied by sensory fibers from the
second and third cervical nerves.
122 D. A. RHINEHART
From the most posterior part of the convexity of the facial
nerve, and just distal to the anastomosis with the ramus auric-
cularis vagi there arise from the motor part of the facial nerve
two small motor nerves which are distributed to the posterior
belly of the digastric and the stylohyoid muscles (figs. 11, 12, and
13G, N. to Dia., and N. to Sty-hy.).
SUMMARY AND CONCLUSIONS
The facial nerve of the mouse corresponds very closely to that
of other mammals and man, and more closely resembles the glosso-
pharyngeal than any of the other cranial nerves. It consists of
two parts: one part made up of those fibers which arise inthe
motor nucleus and form the motor part of the nerve, the other
part being formed by the nervus intermedius and its peripheral
continuation.
The motor part of the facial nerve supplies the stapedius, the
stylohyoid, the posterior belly of the digastric, the auricular
muscles, and the superficial muscles of the face including those of
the vibrissae (special visceral efferent component; Herrick, ’16, p.
146). ;
The nervus intermedius is composed of afferent fibers having
their cell bodies in the ganglion geniculi, and efferent fibers
which have no connection with the cells of the ganglion. The
ganglion itself is of the cerebrospinal type of ganglia and is
composed of unipolar ganglion cells.
The nervus intermedius has three branches, the great super-
ficial petrosal nerve, the chorda tympani, and the ramus cutaneus
facialis. The first two contain both afferent and efferent fibers
the third is composed entirely of afferent fibers. The efferent
fibers which enter the great superficial petrosal nerve form a
separate efferent root of the nervus intermedius.
The great superficial petrosal nerve enters the sphenopalatine
ganglion. Sympathetic fibers from the superior cervical sympa-
thetic ganglion reach the sphenopalatine ganglion by way of the
deep petrosal nerve and the nerve of the pterygoid canal, and
through the nervus abducens. Because the termination of these
fibers in the sphenopalatine ganglion has not been determined,
the function of each set of fibers is uncertain.
NERVUS FACIALIS OF ALBINO MOUSE 123
Fibers from the sphenopalatine ganglion terminate in the
gland of Stenson, in the medial part, at least, of the lacrimal
gland, in the glands of the nasal septum, and along the blood-
vessels of the nose and in the palate (general visceral efferent
‘component). If fibers of the great superficial petrosal nerve
supply the taste-buds of the palate, which, in the mouse is by no
means certain, they probably pass through the sphenopalatine
ganglion without connection with the ganglion cells. Thenerve
supply of the glands of the palate could not be determined.
The afferent fibers in the chorda tympani carry taste impulses
from the anterior part of the tongue (special visceral afferent
component). The efferent fibers of the chorda tympani termi-
nate in the ganglia in connection with the submaxillary and sub-
lingual salivary glands, each fiber ending in relation with more
than one ganglion cell. There is also some clinical evidence that
the chorda tympani contains afferent fibers which carry impulses
of certain kinds of common sensation from the tongue.
The afferent fibers of the ramus cutaneus facialis terminate in
the skin of the external auditory meatus, in the skin of a part of
the auricule and possible a part of the tympanic membrane (gen-
eral somatic afferent component). Fibers from the ramus auric-
ularis vagi are distributed as a part of this nerve, and cutaneous
fibers from the facial nerve are distributed through the ramus
auricularis vagi.
There are several unsolved problems in connection with the
anatomy and function of the mammalian facial nerve which have
been suggested by this work. Among these may be mentioned:
1) The presence and the termination of cutaneous fibers in the
facial nerve of larger mammals and man. 2) If such be found
present, the central connections of the cutaneous fibers. 3) The
termination of the afferent and efferent facial fibers and the
sympathetic fibers of the nervus canalis pterygoidei in the
sphenopalatine ganglion and the function of each set of fibers
4) The nerves within the tongue, especially the termination of
the fibers and the significance of the large numbers of ganglia
and ganglion cells along the nerves.
It is hoped that in the near future the time and the material
will be available for the working out of some of these problems.
124 D. A. RHINEHART -
BIBLIOGRAPHY
AMABILINO, B. 1898 Sui rapporti del ganglio geniculato con la corda del tim-
pano e col faciale. Il Pisani, T. 19. Abstract in Jahresbr. tiber d.
Fortschr. d. Neur., 1898, p. 219.
Cocuity, G. E. 1901 The rami of the fifth nerve in amphibia. Jour. Comp.
Neur., vol. 11, pp. 48-60.
1902 The cranial nerves of Amblystoma tigrinum. Jour. Comp.
Neur., vol. 12, pp. 205-289.
CusHinG, Harvey 1903 The taste-buds and their independence of the nervus
trigeminus. Johns Hopkins Hosp. Bull., vel. 14. pp. 71-78.
1904 The sensory distribution of the fifth cranial nerve. Johns Hop-
kins Hosp. Bull., vol. 15, p. 213.
De Gaerant, L. 1906 Del Nervo intermediario do Tienes e della corda del
timpano. Le Nevraxe, T. 8. Abstract in Zentralbl. f. nor. u. path.
Anat., Bd. 4, 8. 184, 1907.
Herrick, C. Jupson 1899 The cranial and first spinal nerves of Menidia; a
contribution upon the nerve components of the bony fishes. Jour.
Comp. Neur., vol. 9, pp. 153-455.
1900 A contribution upon the cranial nerves of the codfish. Jour.
Comp. Neur., vol. 10, pp. 265-316.
1901 The seantal nerves and cutaneous sense organs of the North
American siluroid fishes. Jour. Comp.-Neur., vol. 11, pp. 177-249.
1914 The medulla oblongata of larval Amblystoma. Jour. Comp.
Neur., vol. 24, p. 361.
1916 An introduction to neyrology. Philadelphia, Pa.
Houpeir, G. Cart 1901 Observations on the innervation of the sublingual and
submaxillary glands. Jour. Exp. Med., vol. 1, p. 281.
Huser, G. Cart, and Guitp, Stacy R. 1913a Observations on the peripheral
distribution of the nervus terminalis in mammalia. Anat. Ree., vol.
(Ds 258"
1913 b Observations on the histogenesis of protoplasmic processes
and of collaterals, terminating in end bulbs, of the neurones of the
peripheral sensory ganglia. Anat. Rec., vol. 7, p. 331.
Hunt, J. Ramsty 1915 The sensory field of the facial nerve; a further contri-
bution to the symptomatology of the geniculate ganglion. Brain, vol.
38, p. 418.
Jounston, J. B. 1908 Additional notes on the cranial nerves of petromyzonts.
Jour. Comp. Neur., vol 18, p. 369.
Kocu, 8. L. 1916 The structure of the third, fourth, fifth, sixth, ninth, eleventh,
and twelfth cranial nerves. Jour. Comp. Neur., vol. 26, pp. 541-552.
KounstamMM, Oscar 1902 Der Nucleus salivatorius chorda tympani. Anat.
Anz., Bd. 21, S. 362-363.
Lancuiey, J.N. 1890 On the physiology of the salivary secretion. Jour. Phys.,
Vole ap eli2a:
Lenuossték, M. von 1894 Das Ganglion geniculi nervi facialis und seine Ver-
bindungen. Beitrige zur Histologie des Nervensystems und der
Sinnesorgane. Wiesbaden.
NERVUS FACIALIS OF ALBINO MOUSE 25
Norris, H. W. 1908 The cranial nerves of Amphiuma means. Jour. Comp.
Neur., vol. 18, pp. 527-568.
1913 The cranial nerves of Siren lacertina. Jour. Morph., vol. 24, p.
285.
Prersot, Geo. A. 1913 Human anatomy. Philadelphia, Pa.
Prenzo, R. 1893 Uber das Ganglion geniculi und die mit demselben zusammen-
hiingenden Nerven. Anat. Anz., vol. 8, p. 738.
Ranson, 8. W. 1912 The structure of the spinal ganglia and the spinal nerves.
Jour. Comp. Neur., vol. 22, pp. 159-177.
Rerzius, G. 1880 Untersuchungen iiber die Nervenzellen der cerebrospinal
Ganglien und den iibrigen peripherischen Kopfganglien mit besonderer
Riicksicht auf die Zellenausliufer. Arch. f: Anat. u. Phys., Anat.
Abt., S. 369.
SHEevLpon, R. E. 1909 The phylogeny of the facial nerve and the chorda tym-
pani. Anat. Rec., vol. 3, p. 593.
Van GruucHTEN, A. 1898 Recherches sur l’origine réelle des nervs craniens.
Jour. de Neur., T. 9.
1900 Recherches sur la terminateon centrale des nerfs sensibles peri-
pheriques. 1, Le nerf intermédiaire de Wrisberg. Le Nevraxe, T. 1.
1906 Anatomie du systéme nerveux de l’homme. 4th edition, Lou-
vain.
Vincent, 8S. B. 1913 The tactile hair of the white rat. Jour. Comp. Neur., vol
23, pp. 1-39.
Weiener, K. 1905 Uber den Verlauf des Nervus intermedius. Anat. Hefte,
Bd. 29, S. 97-163.
Witson, J. G. 1907 The nerves and nerve endings in the membrana tympani.
Jour. Comp. Neur., vol. 17, p. 459.
1910-11 The nerves and nerve endings in the membrana tympani of
man. Am. Jour. Anat., vol. 11, p. 101.
Yaaita, K. 1910 Experimentelle Untersuchungen iiber den Ursprung des Ner-
vus facialis. Anat. Anz., Bd. 37, S. 195.
1914 Einige Experimente an dem Nervus petrosus superficialis major
sur Bestimmung des Ursprungsgebietes des Nerven. Folia Neurobio-
logica, Bd. 8, S. 361-382.
Yaarra, K., anv Hayama, 8S. 1909 UberdasSpeichelsekretionzentrum. Neurol.
Centralbl., Bd. 28, S, 738-753.
Resumido por el autor, Kiyoyasu Marui.
Sobre la fina estructura de la sinapsis de la célula de Mauthner,
con especial menci6n de la ‘‘red de Golgi’”’ de Bethe,
los ‘‘piés nerviosos terminales”’ y la ‘‘red
terminal nerviosa pericelular”’
de Held.
A pesar de numerosas investigaciones la estructura de Ja
sinapsis de una célula nerviosa no esté completamente aclarada
y todavia se necesita una investigacién cuidadosa y exacta,
basada en material adecuado. El autor ha estudiado la fina
estructura de las sinapsis de la célula de Mauthner de los peces
éseos, empleando diferentes métodos, e intenta explicar la natu-
raleza de la ”’ red de Golgi’’ de Bethe, su relacién con las estruc-
turas nerviosas pericelulares, y la estructura de los ‘‘pies ner-
viosos terminales.”’? ‘También intenta dar una explicaci6n acerea
de la existencia de la ‘‘red nerviosa pericelular” de Held y la
teoria de los contactos o continuidad de las: neurofibrillas. Los
resultados de estas investigaciones pueden resumirse del siguiente
modo: 1) La red de Golgi es de naturaleza gliar. 2) La parte
no medulada de las fibras nerviosas est’ envuelta por una vaina
de tejido gliar cuya fina estructura se desconoce. 3) La red
nerviosa pericelular terminal no es realmente una red sino una
imagen producida por el tefido simultaneo de la red de Golgi. 4)
La teoria de los contactos es una imposibilidad histolégica; la
continuidad de las fibrillas intra y extracelulares pudo compro-
barse claramente. 5). Los piés terminales no son- 6rganos
especificos de contacto, sino puntos del trayecto de las fibras
axOnicas en los cuales tiene lugar una disolucién de dichas fibras.
6). No ha podido observar el autor estructura reticular en los
piés terminales ni en la célula de Mauthner.
Translation by Dr. José Nonidez,
Columbia University
AUTHOR’S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 9
ON THE FINER:* STRUCTURE OF THE SYNAPSE OF
THE MAUTHNER CELL WITH ESPECIAL CONSID-
ERATION OF THE ‘GOLGI-NET’ OF BETHE, NERVOUS
TERMINAL FEET AND THE ‘NERVOUS PERICELLU-
LAR TERMINAL NET’ OF HELD
KIYOYASU MARUI
Sendai, Japan
Neurological Laboratory of the Henry Phipps Psychiatric Clinic, Johns Hopkins
Hospital, Baltimore, Md.
FIFTEEN FIGURES
INTRODUCTORY NOTE
The problem of the synapse and the transmission of stimulus
from one nerve cell to another has been a topic of study for more
than a century; we can, however, say in the retrospect that it
was really impossible to come to any safe result on this subject
before the discovery of Golgi’s method. Although recent stud-
ies by many investigators by means of different methods (Apathy,
Bethe, Held, Cajal, Bielschowsky) have brought forth many a
valuable contribution, we are not yet fully enlightened in many
respects. On the basis of their exploration by means of the
Golgi method, Golgi and his pupils came to the hypothesis that
the gray matter of the central nervous system contains a con-
tinuous mesh-work, which consists of the telodendria of the axone
- and of the initial collaterals of the cells of the first type and of
the collaterals from tracts. Forel (9), Cajal (12, 13), Kélliker
(9), Retzius (26), and others denied the existence of this net-
work on the strength of their observations. They declared that
the Golgi net is a false net-work, which in fact is a kind of felt-
work made of the dense processes of neighboring nerve cells.
127
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 1
128 KIYOYASU MARUI
The embryological study of the nervous system supported this
idea, inasmuch as it was believed that the growth of the axis-
cylinder from the cell-body of an embryonic ganglion cell could
be pursued for some distance. Waldeyer (9) summed up all
the facts obtained by means of the Golgi technic in 1891 and
formulated the ‘neurone theory,’ in which he took for granted
that the nervous system is composed of numerous nerve ele-
ments, which are independent anatomically as well as geneti-
cally. This theory implied the so-called ‘contact theory; ac-
cording to the latter, the telodendrion of the neurone terminates
with free endings and the conduction of a stimulus from one
neurone to another takes place by means of contact between
the nerve endings and the nerve cells.
Apathy (9), who demonstrated by his own method the neuro-
fibrils in clear and sharp pictures, assumed that the neurofibrils
at certain points of the central nervous system cross the border
of the nerve cell and through abundant splitting and anastomoses
with the neurofibrils from the adjacent nerve cells form a real
net-work (‘neuropil’). Bethe (7), who by means of his molyb-
den method stained the neurofibrils in the nervous system of the
vertebrate, found on the surface of the ganglion cell as well as
its dendrites a fine net-work of irregular configuration, which he
called ‘Golgi’s net’ in honor of the discoverer. He claimed an
analogy of this net structure with the ‘neuropil’ of Apathy in the
invertebrate, and interpreted it as the connecting link between
the ganglion cells and the nerve fibers, which come afar from
other cells. He claimed to have found that on the one hand
the nerve fibers go over to this pericellular net-work with their
telodendria and on the other hand the intracellular neurofibrils
enter this same net structure. These theories of Apathy and
Bethe cannot therefore be harmonized with the ‘neurone doc-
trine’ ‘n which the nerve fibers are considered as non-anastomos-
ing. It is to be added here that a pericellular net-work con-
tinuous with the nerve fibers was described also by Auerbach
(2, 3), Semi Meyer (23) (24), and Held (17, 18). But the first
two stood on the standpoint of neurone and contact theory, while
_ FINER STRUCTURE OF SYNAPSE 129
the last on the basis of his investigations accepted his doctrine
of double interneuronal continuity. According to Held, the
pericellular ends of axis-cylinders are characterized by the loos-
ening of their axospongium and the densely embedded ‘neuro-
somes’ and the so constituted axis-cylinder-ends (‘Achsencylin-
derendfliche’) are connected with the protoplasm of the nerve
cell, which is covered by them, so closely, that there exists no
contact but a continuity between the two. Moreover, Held
assumed that the different axis-cylinders, which enter together
in the nervous cover of a ganglion cell, do not lie isolated side
by side, but are combined into a continuous net-work, which he
called the pericellular nervous terminal net (perizellulire ner-
vose Terminalnetze’). He believed that also by means of Golgi’s
method he could obtain this ‘pericellular terminal net,’ which
covers widely the cell-body as well as its processes and seemed to
receive numerous axis-cylinders to its beams. When Held (’97;
17) demonstrated for the first time this silver-impregnated mesh,
he identified it at first with the Golgi net, which Golgi produced
with the same technic and interpreted as a neurokeratin cover
of the nerve cell. Later on (17) he changed his opinion and
then considered the Golgi net as not identical with his nervous
net. He asserted that the ‘Golgi net,’ which was demonstrated
by Golgi, Semi Meyer, and then by Bethe; was probably not of
a nervous nature and denied the direct connection of this net-
work with the nervous elements. On the contrary, he was of the
opinion that two kinds of net-work—the Golgi net and the peri-
cellular nervous terminal net—occupy the surface of the nerve
cell, and that alternately. He transferred his ‘neurosome-con-
glomerations’ into the mesh of the Golgi’s net and claimed to
have found that these conglomerations-are connected into a net-
work by minute anastomosing bridges, which extend between
them, over or beneath the beams of the Golgi net. He also fig-
ured the direct connection of the axis-cylinders with these neuro-
some-conglomerations. As regards the formation described by
Auerbach, Held assumed that he observed the same structure.
Ramon y Cajal (’03, 12) summed up the results of his inves-
130 * KIYOYASU MARUI
tigations by means of his own technic and still adhered to the
contact theory. According to the Spanish author, Apathy’s ele-
mentary grating (‘Elementargitter’) does not exist; the ‘neuropil’
of the latter is not a three dimensional net-work, but is to be
regarded as a plexus composed of delicate processes of ganglion
cells. We should emphasize here that Retzius (26) could not
confirm the ‘Elementargitter’ even in Apathy’s own prepara-
tions. In opposition to Bethe’s hypothesis of the independence
and the individuality of the neurofibrils, Cajal claimed to have
observed a neurofibril reticulum in the nerve cells, which is said
to be especially clear at the surface of the cell and around the
nucleus. The Golgi net did not appear in his new preparations;
moreover, his study by the Bethe’s technic showed him a direct
conjunction of the Golgi net with the ‘Fillnetz’ of Bethe, which
the latter interpreted as a coagulation product. On the strength
of these facts Cajal came to the conclusion that the Golgi net is
nothing but. an artificial product formed by the precipitate from
the lymph in Obersteiner’s pericellular space. He denied also the
transition of the axone-fibrils into the intracellular fibrils; accord-
ing to him, the axones end on the surface of nerve cell in the
form of knobs, and there exists always a thin layer of protoplasm,
which is free from neurofibrils at the edge of the cell between the
knobs and the intracellular neurofibrils. He considered this fact
as a new argument of the contact theory. Held’s (19, 20) ob-
servations by Cajal’s method, however, brought to light two
different things, namely, the reticular fibrillous structure of the
knobs themselves and the continuity between the axone-fibrils
and the fibril net-work of the ganglion cell. Holmgren (22)
confirmed these findings of Held. Cajal (13) also figured later
the reticular structure of the knobs.
By means of his method, Bielschowsky (8, 9, 10) reached a
conclusion similar to that of Bethe; in the greater majority of
cell types he could confirm Bethe’s description of the isolated
course of neurofibrils. In the other types, however, he demon-
strated clearly a net-structure of the fibrils, which Bethe also ad-
mitted in some types of the nerve cells. Bielschowsky could not
|
|
|
ie
FINER STRUCTURE OF SYNAPSE 131
acknowledge Cajal’s supposition of a constant existence of a net-
structure, and he attributed this to a fault of Cajal’s method.
With regard to the central endings of axones he formed an idea
that the contact concept must come from the imperfect staining
of the structure referred to. Like Held, he recognized the recticu-
lar structure of the knobs and also believed he could demonstrate
neurofibrils entering the cell so as to constitute a continuity of
the neurones. In addition to this fibril continuity Wolff empha-
sized recently the existence of the plasmatic continuity between
nerve fibers and nerve cells. Bielschowsky and Wolff (11) ob-
served also a pericellular net-work, which resembles the ‘Golgi
net.’ They considered that it is formed by the mutuai com-
munication of the fibrillous and plasmatic substances of the nerve
fibers and stated their opinion that it is probably a structure
identical with the Golgi net.
According to the accurate investigation by Held (18, 19), the
Golgi net is directly continuous with the ‘Fillnetz’ and they are
both of glious nature. Economo (15), Paladino (25), and others
confirmed this. As I mentioned above, Held (18) supposed the
existence of two kinds of pericellular net-work. Economo (16)
agreed with him. But this problem is not at all altered because
Heidenhain (16) described recently that he would consider the
Golgi net an artefact rather than a glious element. With regard
to the pericellular terminal net, he admitted so far only that
occasionally there exist sling-like loops within the arborization of
one and the same branch of nerve fiber.
Despite many investigations, the theory of contact still opposes
the doctrine of the neurofibril continuity. Hedenhain (16) said
that one can undoubtedly deny the latter, declaring that the figures
by Held (17, 19, 20), Holmgren (22), Bielschowsky (9), and Wolff
(27) prove nothing at all and that with others (Cajal (12, 13),
Dogiel, (14), Retzius (27), Lenhossek (cited in 16) ) he sought in
vain the continuity of nerve fibers and the cells in question.
In 1915 Bartelmez (4) studied the Mauthner cell of teleosts,
which had been investigated by Mayser' and Becarri (5), and
1 Zeitschrift f. wissenschaft. Zoologie, Bd. 36, 1882.
12 KIYOYASU MARUI
he further analyzed the elements of the peculiar synapse of this
giant cell. Lately I had the opportunity to do some experimental
pathological study on this wonderful cell and its synapse, the
results of which will appear soon. Careful and thorough investi-
gations led me, however, to many interesting conclusions in re-
gard to the structure of the synapse and the mode of conjunction
of nerve cells; I will therefore describe these results in the present
paper with special considerations of the condition in higher ver-
tebrates. I beg to express here my gratitude to Dr. Adolf
Meyer for his constant help and frequent advice in this work.
MATERIAL AND METHODS OF STUDY
The present work is based upon the investigation of serial sections of
brains of adult Ameiurus nebulosus and Carassius auratus. The fish
were decapitated, bled, and the brains were dissected out carefully but
quickly, and placed promptly in different fixatives; the methods of
preparation used are as follows:
At first I mention the toluidin-blue preparation, in which brains
were fixed in 95 per cent alcohol for twelve to twenty-four hours. Par-
affin sections 8 to 10 uw thick were stained in 1 per cent warm aqueous
solution of toluidin blue, differentiated in alcohol and sometimes coun-
terstained with eosin. ‘
Other brains were fixed in formol-Zenker fluid twelve to twenty-four
hours, cut at 8 yw in paraffin, stained in a saturated solution of thionin
in a 1 per cent aqueous solution of carbolic acid and counterstained
with eosin after differentiation in alcohol.
For Cajal preparations brains were placed twelve to twenty-four
hours in alkaline alcohol (96 per cent alcohol with 1 per cent ammonium
hydroxid). The fixed brains were then rinsed with 95 per cent alcohol,
placed in distilled water till they sank to the bottom of the container
and then they were placed in the silver-bath of 37° to 40°C. for ten
to fourteen days, according to the size of the pieces. As silver-bath I
used a 1 per cent silver-nitrate solution for the first seven to ten days,
with one change, and then a 2 per cent solution with one change, in
which the pieces were kept three to four days. They were then rinsed
with distilled water and developed twelve to twenty-four hours in 1
per cent pyrogallic acid solution in 5 per cent formalin, again rinsed
with water, dehydrated and cut in paraffin 5 to 10 w thick. This al-
ways gave good results; in my experience the Mauthner cells are very
apt to show shrinkage space after fixing in acetic alcohol for a short
time (Bartelmez) (4), which is unfavorable for the study of the synapse.
I next applied the Levaditi method for spirochaeta pallida to the
fish brain. The brains were fixed in 10 per cent formol twelve to
FINER STRUCTURE OF SYNAPSE 133
twenty-four hours, rinsed with water, placed in 95 per cent alcohol for
twenty-four hours, and then placed in distilled water, till they sank to
the bottom and they were then put into the silver-bath for five to seven
days (in a 1 per cent silver solution for three to four days and in a 2
per cent solution for two to three days). For the reduction of the
impregnated silver a 2 per cent pyrogallic acid solution in 5 per cent
formol was used. ‘The sections were cut at 5 to 10 yu. This preparation
offered excellent pictures of the synapses and led to many interesting
findings, which will be described later.
The Bielschowsky method of silver reduction also yielded nice pic-
tures of the intracellular neurofibrils and thesynapse. The brains were
fixed in 10 per cent formol for twenty-four hours, rinsed with water,
dehydrated and cut in paraffin at 5 to 10 uw. The paraffin was removed
with xylol and alcohol and the sections were thoroughly washed with
distilled water and placed in a 2 per cent silver solution for two to
three days. The sections were treated on the slide, following exactly
the directions of Bielschowsky. Some of the series were counter-
stained with eosin.
For Heidenhain preparations, brains were fixed in Zenker’s fluid or
in formol-Zenkr fluid. Some brains were put into the latter, after
being fixed first in 10 per cent formol for twelve to twenty-four hours.
The sections of 5 to 8 » were stained in the iron-hematoxylin of Hei-
denhain. This method gave not only a clear picture of the cell-body
and the synapse, but also a blue stain of the myelin sheaths that was
sometimes very advantageous.
In addition to these preparations I prepared also a few series with
Held’s neuroglia. stain, Weigert’s stain, and Mallory’s stain. In all I
studied ninety-seven series of normal fish brains in the different methods,
classified as follows:
(1). 5 series of Ameiurus ee Fixed in 95 per cent alcohol, and stained
5 series of Carassius auratus { with toluidin-blue and sometimes eosin.
(2) 5 series of Ameiurus nebulosus | Fixed in formol-Zenker fluid, stained with
5 series of Carassius auratus | thionin-eosin.
(3) 5 series of Ameiurus nebulosus
11 series of Carassius auratus
J
\ Cajal preparation.
(4) 14 series of Ameiurus pees
i : Levaditi’ :
15 series of Carassius auratus eae eethod
(5) 10 series of Ameiurus nebulosus
11 series of Carassius auratus
(6) 5 series of Ameiurus nebulosus
4 series of Carassius auratus
(7) 2 series of Carassius auratus Held’s neuroglia stain.
(8) 2 series of Ameiurus nebulosus Mallory’s stain.
(9) 2series of Ameiurus nebulosus Weigert’s stain of myelin sheath.
Bielschowsky’s stain.
Heidenhain preparation.
134 KIYOYASU MARUI
NEUROFIBRIL STRUCTURE OF THE MAUTHNER CELL
It is not the purpose of the present work to study the posi-
tion, general relations, and size of this wonderful cell. My main
purpose lies, on the contrary, in the further investigation of the
finest structure of this giant cell. I merely refer here to the
works of Mayser, Beccari (5), Herrick,? Bartelmez (4), and
others in regard to those points. As far as the internal morphol-
ogy of the Mauthner cell is concerned, it would not do to ignore
the condition of the intracellular neurofibrils in the study of
synapse, since, as I mentioned above, the relation between extra-
and intracellular neurofibrils is not clearly decided yet. Though
Bartelmez and others described the neurofibril structure of this
cell, I will add the results of my study here, as there are many
points not sufficiently clear.
The Cajal, Levaditi, and Bielschowsky preparations were
chiefly used in my study of the neurofibril structure of the
Mauthner cell. By means of the first two methods the neuro-
fibrils were exceedingly clear in many cases, but in other cases
they proved very uncertain in,their results. The cell was now
and then stained merely a diffuse yellow or brown, without any
neurofibrils being differentiated in it, and in other cases the
ground substance of the cell was also tinged more or less inten-
sive yellow or brown, so that the neurofibril structure did not
present itself clearly enough. In comparison with them, the
Bielschowsky technic offered always excellent results; I shall
therefore speak mainly of the Bielschowsky preparations, besides
giving consideration to the best preparations of Cajal and Le-
vaditi. Owing to the wealth of neurofibrils in the Mauthner
cell, the sections must be thin in order that the course of the in-
dividual neurofibril could be followed distinctly.
The results of my investigation on the neurofibril structure in
Mauthner’s cell harmonize in main features with those of Bethe
(6, 7), Bielschowsky (8, 9, 10), and Economo (15), in so far as
the neurofibrils do not form a real net-work in the cell. On the
2 Journal of Comparative Neurology, vol. 24, 1914, p. 343.
FINER STRUCTURE OF SYNAPSE LS)
strength of his study by means of his method, Bethe came to the
conclusion that the neurofibrils pass generally through the cell-
body without any net-formation. With regard to this, how-
ever, the cells of the Ammon’s horn and the Purkinje’s cell re-
mained uncertain to him. As far as the cells of the spinal gan-
glia and of the lobus electricus of Torpedo are concerned, he was
quite sure that there exists a net-work in them. According to
Bielschowsky, who worked with his method, the neurofibril
structure corresponds in essential features to that of Bethe.
Eeconomo also demonstrated neurofibrils which do not anasto-
mose with the other fibrils. Like Bethe, he found also fibrils
which run isolated from one branch of one dendrite to another and
are not continuous with those in the cell-body.
As figure 1 shows, the neurofibrils run straight or winding,
but smoothly through the cell-body and, generally speaking, par-
allel to the axis of Mauthner’s cell, without any anastomoses for
a long distance. Some of the neurobrils can even be followed
through the whole length of Mauthner’s cell, without any bifur-
cation of neurofibrils. The latter occurs, after all, only now and
then, and this strongly suggests that the neurofibrils do not
form a real mesh-work in the cell. On this point I certainly
agree with Bethe (6), who argued that bifurcation would appear
oftener, if there were really a net-work in the cells. Hconomo
(15) suggested that there probably occurred merely a sticking
together of neurofibrils, which lie side by side in the cells. In the
neighborhood of the nucleus the course of the fibrils is slightly
more irregular, but I could not observe any anastomosis between
the neurofibrils. At the starting point of the cell processes the
fibrils radiate forth into the cell-body and at this point certainly
nothing of a net structure or even of a branching of single fibrils
could be noticed, which, with Economo (15), we would here ex-
pect in a real net-structure. From the processes and from the
interior of the cell the neurofibrils stream into the axone hillock,
and they could be followed pretty far into the latter. All these
findings speak no doubt against the existence of a neurofibril
net-work in Mauthner’s cell.
136 KIYOYASU MARUI
Raméon y Cajal (12) emphasized on the basis of his study by his
own technic a reticular structure of the neurofibrils in the gan-
glion cells. He distinguished two kinds of fibrils—the primary
and the secondary neurofibrils; described the existence of a super-
ficial and a deep neurofibril net-work, and he stated that the
parallel fibrils of the dendrites enter into correspondence with
both the net-works. He declared, moreover, that the fibrils of
the axis-cylinder originate from both these reticula. Retzius (26),
v. Lenhossék, and Held (19, 20) produced by means of Cajal’s
method similar results as Cajal, while others (Marinscu, v. Ge-
huchten, and others cited in 15) maintained a certain reserve
on this point, at least for some cell types. Bielschowsky (8),
who studied the motor cells of the anterior horn by Cajal’s
method, pointed out the disadvantages of the latter, through
which a false picture of a reticulum might be brought out. Wolff
(28) expressed an opinion concerning the spatial relation be-
tween the honeycomb of the protoplasm and the neurofibrils
and declared that the neurofibrils always lie in the wall of.the
former. According to Economo, the neurofibrils through simul-
taneous impregnation of this’wall of the honeycomb pattern
may look as if they were connected by cross-beams so as to form
a net-work. —
As far as the anastomotic connections between the neurofibrils
are concerned, Wolff (28) did not want to speak of a real net-
work; on the contrary, he assumed theoretically a plexus forma-
‘tion of the neurofibrils. Heidenhain also, on the basis of the
theoretical considerations, assumed a metaplexus in the cell. In
the present work, however, I will not enter further into theoreti-
cal considerations, but restrict myself merely to the microscop-
ically visible things. I came to the conclusion that the neuro-
fibrils in Mauthner’s cell form no net-work at all. Above all I
wish to emphasize here that I could not observe any neurofibril
reticulum in this cell even in my best preparations with Cajal’s
and Levaditi’s methods (fig. 2).
FINER STRUCTURE OF SYNAPSE 17
THE ‘GOLGI NET’ OF BETHE IN THE SYNAPSE OF MAUTHNER’S
CELL :
Careful investigation of the synapse of this giant cell by means
of different methods revealed a most characteristic net-work,
which covers the cell-body as well as its processes like a basket
and also fills the ‘axone cap’ (figs. 3 and 4). There is no doubt
that this net-work is identified with the structure, which was
described for the first time by Golgi (’93) and later called the
Golgi net by Bethe.
This net-work was constantly demonstrated in Mauthner’s
cell and was most beautiful in the Levaditi preparations; as far
as I know this technic has never been applied before for the study
of this structure. In Ameiurus brain the net-work appeared in
a brown or dark color, whereas in Carassius it was stained in a
brown to yellow color. As another advantage of this technic
it must be emphasized that the nerve fibers in the synapse were
also demonstrated as clean cut: dark brown fibers simultaneously
‘with the Golgi net, especially clear in Carassius. In the latter
the Golgi net substance appeared mostly yellow and the con-
trast with the stain of the nerve fibers was so striking that it
threw very good light on the problem of the mutual relation be-
tween both structures, which had been discussed by different
authors and still had remained undecided. In both Ameiurus
and Carassius the ‘Fiillnetz’ of Bethe was also stained yellow in
this method.
In the Heidenhain preparations the same structures were also
demonstrated, especially clearly in formol-fixed material; but
here the Golgi net could only be observed around the cell and in
the ‘axone cap.’ On the surface of the Mauthner cell I could
hardly see this structure, owing to the fact that the cell-body
and the Golgi net appeared in too closely similar color (fig. 5).
In the Heidenhain preparations (fig. 7) and the thionin-eosin
preparations of formol-Zenker material one could hardly recog-
nize the Golgi net-work, unless he has impressed the picture
upon him in the above-mentioned preparations. In the Cajal
preparations the Golgi net and the ‘Fiillnetz’ were not observable.
138 KIYOYASU MARUI
I will first describe my findings in the Levaditi preparations.
On the dendrites and on the part of cell-body which is not covy-
ered by the ‘axone cap’ the Golgi net is a single layer; that is to
say, one layer of a net surrounds the cell and the dendrites.
Each nodal point of this net-work is connected with the cell sur-
face by means of a beam, which is attached to the cell surface
perpendicularly, with a slightly extended basis (fig. 4): The
‘axone cap’ appears in a triangular shape in Ameiurus sections
and is filled with more or less numerous layers of the Golgi net;
the nodal points of the net layer, which lies closest to the cell
surface, are connected with the latter by a Golgi net beam each.
The meshes of the net-work are irregular and of variable size;
but, generally speaking, the mesh is smaller the nearer we come
to the cell. In Carassius the ‘axone cap’ has a round shape and
here it shows a dense conglomeration of the Golgi net substance.
I believe I can compare this heap of the Golgi net substance in
the ‘axone cap’ with that conglomeration of the Golgi net structure
which Bethe (7) found on the Purkinje cell of the cerebellum
and in other parts of the central nervous system. Most meshes
are five or six cornered, but there are many four-cornered and
some seven- and eight-cornered ones. . At the nodal point of the
mesh-work three beams join together as a rule; the nodal points,
at which four mesh beams unite together, arerarer. The nodal
points of the mesh-work in both the ‘axone cap’ and on the cell
surface are a little thickened, and in my preparations of Caras-
sius brain some of them contrast clearly as deeply brown stained
round points with the yellow or light brown impregnated Golgi
net beams. These points might well be interpreted as the cross-
sections of the nerve fibers in. the synapse, as is to be described
more precisely below. Unlike Bethe (7), who denied the direct
connection between the Golgi net and the ‘Fillnetz,’ I could find
the direct continuity of both the net-works, as has already been
claimed by Held (18, 21), Economo (15), and others. The ‘Full-
netz’ pervades the whole gray and white matter and fills the space
between the myelin sheaths, surrounding the latter with its net-
work. At the nodal points and also in the beams of this mesh-
_—,
FINER STRUCTURE OF SYNAPSE 139
work we find in this preparation brown- or black-stained points of
different caliber, which are certainly the cross-sections of nerve
fibers. The ‘Fiillnetz’ is less sharply marked, looks lighter than
the Golgi net; the mesh itself is larger and more irregular and
the mesh beams are thicker than those of the Golgi net-work.
Contrary to the statement of Bethe (7), I could observe very
distinctly that the Golgi net beams are connected with glia
nuclei by means of somewhat extended bases, and also with the
walls of capillaries.
In the Heidenhain preparations of formol material I could
confirm similar relations of the Golgi net and the ‘Fillnetz,’ as
described above (fig. 5). The nodal points of the net-work also
were found a little thickened here and there. The only thing
which is different is that the nervous elements here are stained
in a color similar to that of the Golgi net, and the distinction
between the nervous elements and the Golgi net becomes natu-
rally difficult.
In both the above-mentioned preparations I observed indis-
putably the direct transition of neurites into the Golgi net beams,
as is shown very clearly in figures 5, 6, and 11, although I do
not mean by that at all that the Golgi net is of nervous nature.
To this question of the nature of the Golgi net and its relation
to the nerve fibers I will return later.
As already remarked, Held (18) demonstrated the net-like
formation on the different kinds of the central ganglion cells,
after Golgi had. described it for the first time. Held at first
identified his net-like formation with Golgi’s net-work and char-
acterized it as a dense net-work with coarse nodal points, formed
by the fusion of arborized nerve fibers in the gray matter. He
therefore described it as the ‘pericellular nervous terminal net.’
Besides Held, the net-like framework of the ganglion cell was
also described by Semi Meyer (23, 24), Auerbach (2), and Bethe
(7). Auerbach described the terminal nervous net around the
ganglion cell on ground of his own method. Bethe, who demon-
strated his Golgi net with the help of his molybden technic, con-
cluded in harmony with Held, Semi Meyer, and Auerbach that
140 KIYOYASU MARUI
it is probably of nervous nature, because he believed to have
found in the first place that the ends of the neurites are directly
connected with the Golgi net and in the second place that the
neurofibrils of the ganglion cells are combined with the nodal
points of the Golgi net-work. But Held (18) raised the ques-
tion whether the different authors have really seen the same
structure. He said: ‘‘There is no doubt that certain net-struc-
tures of the Golgi preparations and the methylene-blue picture
of Semi Meyer and Bethe’s net-work are identical. What char-
acterizes them is above all a certain monotony of the net figure
and the stiffness of the net beams and further the nodal points
of the mesh are generally not considerably thicker than the
beams themselves.’”? He then declared that Semi Meyer had —
seen something different, because Meyer’s figures offered no proof
of the nervous nature of his net-work, as it was stained isolated
and without any relation to the nerve fibers. As regards Bethe’s
opinion, Held assumed that neither Golgi nor Bethe had fur-
nished the necessary evidence. According to him, the statement
of Golgi? that the neurofibrils of the gray matter unite with the
net-work around the nerve cells proves nothing, because in Gol-
gi’s method the impregnation can jump from the branching nerve
fibers to the thoroughly disconnected net structure and create a
false connection between them. On account of the same reason,
Held (18) denied the conclusiveness of his own Golgi prepara-
tions, in which the close relation between the nerve endings and
the net-work was distinctly visible (fig. 8, table 13, 1902). Nor
did the figures and the statement of Bethe about the transition
of neurites into Golgi net beams seem trustworthy to him, al-
though he was not able to give any special reasons for this. As
far as the findings of Auerbach are concerned, Held thought
that Auerbach had observed a similar structure and had inter-
preted it essentially alike. On the basis of these considerations,
Held (18) came to the conclusion that on the surface of certain
ganglion cells of vertebrates there exist two different kinds of net-
8 Jahresbericht v. Merkel u. Bonnet, 1894. Cited in Held’s 1902 (18).
FINER STRUCTURE OF SYNAPSE 141
work—the Golgi net and the nervous pericellular terminal net—
which are quite distinct in their relation to other elements of
gray substance and in their functions.
Held furnished as an argument for this hypothesis the fact
that the nodal points of each net-work are distinct from each
other. He figured round or star-shaped formations in the mesh
of the Golgi net, connected among each other by delicate threads,
which lay above or beneath the beams of Golgi net. He further
claimed, pointing to the figures and statement of Bethe (7), that
some round or oval specks, which are visible here and there as
the contents of some meshes, might well be interpreted as the
residue of discoloration of those star-shaped formations, which
remained stained in his own preparations. Held (17) stated
formerly that the surface of the ganglion cells are covered by
the variably densely scattered and variably large protoplasm
pieces, which have granular structure and are connected among
each other-with fine threads so as to form a net-work. At that
time he spoke of them as the neurosome conglomerations on ac-
count of their thick granulations, and he identified them with the
nodal points of his nervous terminal net, which he demonstrated
by means of the Golgi method. Later he (18) identified these
neurosome conglomerations with the contents of the Golgi net
meshes referred to. Bethe (7) tried to argue that the neurosome
conglomerations of Held must be regarded as the broken prod-
ucts of his Golgi net. Against this opinion, however, Held
(18) furnished the proof that his neurosome conglomerations
could be pretty clearly demonstrated in sections, which were
prepared for Bethe’s molybden method II by means of other
stains, whereas in the alternating sections the intact Golgi net
could be demonstrated by Bethe’s stain.
In the following I wish to describe my results in Mauthner’s
cell in comparison with the findings of other authors in other
cells. My observations led me to many interesting conclusions,
which could not be brought into conformity with the findings
of other authors without herein wanting to generalize those re-
sults in the fish brain immediately. In the first place, attention.
142 KIYOYASU MARUI
must be called to the fact that in my Levaditi preparations and
Heidenhain preparations of formalin material the meshes of the
Golgi net were free from any formation or granule, which could
be homologized with the neurosome conglomeration of Held.
As the figures (6 and 11), produced from a Levaditi preparation
of Carassius, show, the cell-body is covered by a net-work, whose
nodal points are considerably thicker than the beams and appear
as deeply brown-stained spheroidal points. This net-work
passes directly into the net structure of the ‘axone cap,’ which
is itself continuous with the ‘Fiillnetz’ of Bethe. There can be
no doubt that this net-work is the Golgi net. The meshes of
this net-work are free from any further material, as the figures
clearly indicate. In the axone cap, too, we cannot find any
content in the mesh of the net-work. Above all, stress must be
laid in these figures (6 and 11) upon the fact, that numerous nerve
fibers, which have lost their myelin sheaths at the border of the
‘axone cap,’ run toward the surface of the cell, always keeping
their place along the beams of the net-work. Moreover, we find
the nodal points of the net-work here and there showing round
brown and thick points, which could well be interpreted as the
cross-sections of the nerve fibers, running along the beams of the
net-work. Now and then we find a cross-section of a nerve
fiber, which looks as though it lay isolated in the mesh of the
Golgi net, but through careful observation with continual rota-
tion of micrometer screw we realize that it is connected with the
neighboring nodal points of the net-work by means of delicate
yellow threads. On the surface of the Mauthner cell we also,
find occasionally in the mesh of the Golgi net a certain granule,
which might perhaps be homologized with the neurosome con-
glomeration of Held. This granule, which is seemingly an iso-
lated round particle, but in reality a star-shaped structure with
radiating beams, does not lie in the mesh of the net-work, but is »
found in a level different from that of the neighboring net beams,
and with careful observation it is, at least at times, found to be
connected with the Golgi net by means of the radiating beams.
In the Heidenhain preparations of formol material a similar
a
FINER STRUCTURE OF SYNAPSE 143
condition of the synapse was in evidence and the close relation
between the nerve fibers and the Golgi net-work was very clear,
as is shown in figure 5. In the Heidenhain preparations of
formol-Zenker material the synapse appears denser than in the
previous preparations in sections of the same thickness, but one
can after some practice easily find the similarity of the net-work
in the cell circumference and in the ‘axone cap.’ In figure 7,
produced from this preparation, we observe here and there in
the meshes of the Golgi net around.the Mauthner cell a star-
shaped granule which is found oftener here than in the previous
- preparations. It is obvious that these granules correspond to
Held’s neurosome conglomerations, which he located in the mesh
of the Golgi net and interpreted as the nodal points of his ‘peri-
cellular terminal net.’
As was noted in previous preparations, these granules do not
lie exactly in the mesh of the Golgi net, so far as my observations
go. We cannot observe both these granules and the surrounding
Golgi net beams with equal distinctness in one and the same
focus of the microscope. By careful observation with slight
rotation of the micrometer screw we could ascertain here as well
that these granules are connected by means of their radiating
threads with neighboring nodal points of the Golgi net. Al
though I noted also, as Held (18) did, that now and then the
radiating threads do not pass over the neighboring nodal points,
but reach the adjacent granules, passing either over or beneath
the net beams, ‘this should not be wondered at in the three-
dimensional net-work. On the ground of these considerations,
I am inclined to assume that these granules correspond to nodal
points of the Golgi net, which lie in different levels and give the
false idea that they are within the mesh of the Golgi net. Even
in the thionin-eosin preparations a similar structure was recog-
nizable, although it is much more difficult to perceive than in the
previous preparations.
Held (18) figured the mutual relation between the neurosome
conglomerations; but it is not indicated clearly enough in his
figures (7 a, b, 11, 12, 1902). His Golgi figures (3, 4) showed the
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL, 30, NO. 1
144 KIYOYASU MARUI
situation more clearly; but his interpretation seems to be wrong
in so far as he wanted to explain these figures as his ‘nervous
pericellular terminal net,’ which according to his opinion les
alternating with the Golgi net on the surface of the nerve cell.
I shall come back to the argument on this point later. As men-
tioned, Held (18) denied Bethe’s opinion concerning the direct
transition of the neurites into the Golgi net; also Economo (15)
figured such a picture (table 3, fig. 25), but he remained in agree-
ment with Held. I have already remarked that my prepara-
tions (figs. 5, 6, and 11) showed indisputably the close relation
between the Golgi net and the neurites; I could even demonstrate
pretty often delicate deep brown impregnated nerve fibers in the
beams of the Golgi net, as shown in figures 3, 6, and 11, which
arrive at the surface of the celi. Bethe (7) enumerated five points
in his paper which according to him speak in favor of the con-
nection of neurites with the Golgi net and in favor of the nery-
ous nature of the Golgi net. With regard to his second argu-
ment, that the Golgi net is always dense and fine in places of the
central nervous system where numerous axis-cylinders ramify, I
would mention that in the axone cap of Mauthner’s cell, where
abundant axone-fibers form a dense plexus, the Golgi net struc-
ture offers a very thick conglomeration. As far as Bethe’s third
argument goes, concerning the probability of demonstrating the
neurofibrils in Golgi net beams, I can explicitly assert that I
could without doubt find the nerve fibers in the latter, as I
stated already. I believe I have settled this point, on which
Bethe (7) was not sure himself. At the same time I must em-
phasize that I could not confirm Bethe’s statement, that the
neurofibrils form a net-work in the net beams; as I shall discuss
in the following chapter, I do not find any net-work of the nerve
fibers in the synapse of Mauthner’s cell. Although these argu-
ments of Bethe, as Held (18) declared with good reasons, give
merely indirect proofs for the nervous nature of the Golgi net,
I think they indicate at least that the neurites have something
to do with the Golgi net. I do not mean by that, however, at
all, that the Golgi net is a nervous structure. Now one might
FINER STRUCTURE OF SYNAPSE 145
raise objection against me that the nerve fibers, which I have
demonstrated in the beams of the Golgi net, might be neuroglia
fibers. Concerning this I should raise the following points: in
the first place, I could not, like Bartelmez (4), find any neuroglia
fiber by means of specific neuroglia methods, and in the second
place, I was able to follow the nerve fibers far back to the part
where they have myelin sheaths.
As quoted above, the connection of nerve fibers with the Golgi
net was stated by Golgi; also Held himself figured such points
from his Golgi preparations (fig. 8, table 13, 1902). But he
immediately denied the demonstrative power of these findings
on account of the drawback of the Golgi technic mentioned above.
I wonder why Held was so severe in his criticism concerning
Golgi’s method, when the picture was unfavorable to his opinion,
whereas he avails himself of it to its full extent, if it is convenient
for him. Held (17) formerly demonstrated (figs. 5 to 8, 1897)
his ‘pericellular nervous terminal net’ around a certain kind of
nerve cell. Later, however, he altered his interpretation of these
figures so that only his figures 7 and 8 can hold their ground even
in his extended critical consideration. If one, however, com-
pares Held’s figures 5 and 6 with 7 and 8, one can hardly see why
in the figure 6 and in the light part of the figure 5, which showed
the impregnation of the Golgi net, he should later have seen
evidence of a nervous net-work. To me all these figures, espe-
‘cially 6 and 7, look very much alike; in the light part of the figure
5 the nodal points of the net-work are not thickened, but to my
mind one should not be surprised at this in a method which is
so inconstant in its results. Besides, it must be emphasized here
that Held (17) did not rely on satisfactory evidence, when he set
up his ‘pericellular nervous terminal net.’ He considered the
pericellular net-work as the nervous-one merely for the reason
that he found the connection between the net-work and the ar-
borization of the nerve fibers and the thickening of the nodal
points of the former. The relation between the Golgi net and
the nerve fiber arborizations has been settled, I believe, through
my present investigation. Moreover, despite the characteriza-
146 KIYOYASU MARUI
tion of the Golgi net by Held, I have found a thickening of the
nodal points of the Golgi net, although I would not deny the
fact that the nodal points sometimes, especially in my Levaditi’s
preparations of Ameiurus, were not particularly thickened,
when the impregnation of the nervous elements was not suf-
ficient. According to my opinion, however, this mark of dis-
tinction is not of considerable value and the variation must be
regarded as the consequence of variable impregnation and
methods of demonstration. It would be necessary to add here
that in my Heidenhain preparations the nodal points appeared
always thickened. On ground of these considerations, all the
figures of Held seem to be pictures of the Golgi net, which is
connected with nerve fibers and possesses the thickening on its
nodal points; the latter might well be regarded as the cross-
section of the nerve fiber or the so-called terminal foot of Held,
which I will describe and discuss in the following chapter. The
figures of Held’s publication (02; 3, 4) might come under the
same point of view. Anybody who would compare those figures
of Held with mine (figs. 6 and 11) would note immediately that
they demonstrate thoroughly, similar situations. I therefore
came to the conclusion that on the surface of the Mauthner cell
there exists a single net-work—the Golgi net—and that the
endings of nerve fibers do not lie in the mesh of the latter, as it
was supposed by Held, but the nervous elements of the synapse
are related to the Golgi net and reach the cell surface at the °
same points where the Golgi net beams are attached to the cell
surface. Held seemed to have been misled in his hypothesis on
account of his one-sided interpretation of Golgi and neurosome
preparations and his belief in the glious nature of the Golgi net.
As far as the nature of the Golgi net is concerned, the inter-
pretation of Golgi,t Cajal (12), and Heidenhain (16), and Wolff
(27), who wanted to see in it a neurokeratin structure, an arti-
fact by coagulation, or then the superficial part of the honey-
comb structure of the nerve cell, must immediately be denied.
4 Cited in Plasma u. Zelle, Heidenhain, Bd. 1, a, p. 911.
FINER STRUCTURE OF SYNAPSE 147
Also the hypothesis of Bethe (7) who considered it as the nervous
structure, I should deny positively with Held (18, 21), Economo
(15), and others. Held discovered in both the white and gray
matter of the brain a glia’ reticulum which provides myelin
fibers with neuroglia sheaths and he identified this with the
‘Fiillnetz’ of Bethe; he also demonstrated that the beams of
the glia reticulum pass into those of the Golgi net, increasing
thereby the density of its substance. Afterwards he added his
further observation that the neuroglia cells accompanying the
ganglion.cells form the Golgi net of the latter. Economo found
the direct connection between the Golgi net beams and the gla
nucleus by pediform appendages. All these findings I could es-
tablish in the synapse of Mauthner’s cell; there is no doubt now
that the Golgi net is of glious nature; it must be added here that
also Schiefferdecker,? Paladino (26), Besta,® and Alzheimer
admitted its glious nature.
What, then, is the relation between the nervous elements and
the Golgi net structure? As I remarked already, the myelin fibers
are enclosed in glia sheaths of, net-like structure, which is con-
tinuous with the diffuse glia reticulum. As far as I know, the
relation of the unmedullated part of the nerve fibers to the glia
tissue is not known very clearly. Held (19) said that, only oc-
casionally and rather unevenly by means of Bethe’s molybden
technic, he obtained pictures in which the unmedullated fibers
are enveloped in a delicate Golgi net. He added also that these
sheaths of unmedullated nerve fibers are connected with the
diffuse delicate glia reticulum by means of projecting spikes.
This description of Held comes pretty near to my own findings
in this work (nerve fibers in the beams of the Golgi net, cross-
sections of nerve fibers at the nodal points of the net-work),
which indicate that the unmedullated part of the nerve fiber is
also enclosed in glia sheaths. I am very sorry that I could not
establish the finer structure of these glia sheaths in the present
5 Heidenhain, Plasma u. Zelle, Bd. 1, a, p. 911.
6 Riv. di. patol. nerv. e ment., T. 16, p. 604.
7 Nissl’s Arbeiten, Bd. 3. 3, p. 412-421.
148 KIYOYASU MARUI
investigation; but on the basis of my findings I should be able
to declare, I believe, that they accompany the neurites to their
ends, connecting each other by means of delicate beams and
thus forming the Golgi net in the axone cap as well as on the
cell surface. Also the minute cap dendrites described by Bartel-
mez (4) are enclosed in the yellow net beams of the Golgi net.
Bielschowsky (18) also demonstrated the Golgi net by his
method and he could even observe in harmony with Bethe (7) |
that nerve fibers stream on many points into the beams of the
Golgi net. Further, he and Wolff (11) expressed themselves about
the relation between Bethe’s net-work and the ‘terminal nervous
net,’ which they demonstrated as follows:
The mesh-formation in our picture is not as regular as that which
Bethe’s technic yields. Also the net beams appear in ours more deli-
cate than those in Bethe’s. Still we consider it is probable that both
methods demonstrate in this the identical structure. The difference
might depend upon the fact that in Bethe’s technic the plasmatic
component of the net-work and in ours, the fibrillous component, be-
comes more manifest in the preparations. As Bethe (7) insisted, the
beams of the net-work are not homogeneous but delicate fibrils are
recognizable in them, which are continuous with the intracellular
neurofibrils. :
Now, if one compares this statement with my description, it
becomes highly probable that the so-called “‘plasmatic compo-
nent’ of the net-work of Bielschowsky and Wolff is to be inter-
preted as the Golgi net substance. As will be described in the fol-
lowing chapter, in some of my Bielschowsky preparations the
sharp and intensely impregnated endingsof neurites in the synapse
of Mauthner’s cell were seen connected with each other by pale
stained beams forming a net-work. In the eosin counterstain
of this preparation the sharp and dark stained components of
the net-work were covered by more or less thick red sheaths and
they were interconnected by red-tinged bridges so as to form a
net-work (fig. 15). It was extremely interesting to find that
the red-stained component of the net-work was directly contin-
uous with the glia reticulum and with the glia nucleus. This
picture is to some extent equaled by my findings in the Levaditi
FINER STRUCTURE OF SYNAPSE 149
preparations. Bielschowsky and Wolff (11), Held (20), and
others took for granted that the transition of the axone-fibers
into the nerve cell consists of axone-plasma and fibrils. I agree
with this opinion. But I believe that the component of the
net-work, which appeared red in eosin counterstain, would be
the Golgi net substance, to some extent at least. The statement
of Bielschowsky and Wolff (11) that the simultaneous existence
of a glious pericellular net-work is compatible with their previous
findings, can no longer be maintained, on ground of my above-
described considerations.
THE NERVOUS TERMINAL FEET AND THE NERVOUS TERMINAL NET
OF HELD AND THE NEUROFIBRIL CONTINUITY
The connection of neurones by means of special structures
(nervous terminal feet) was first discovered by Held (17, 18),
confirmed by Auerbach (2), and popularized by Ramon y Cajal
(12, 18). Through the further investigations of many authors
by means of Cajal’s and Bielschowsky’s methods many valuable
contributions were added to this interesting and important prob-
lem. The question of the nervous pericellular terminal net and
the neurofibril continuity also has long been the subject of dis-
pute among the histologists. We are not as yet enlightened
thoroughly on these questions.
I shall now go over these questions on the basis of my inves-
tigation on the Mauthner cell. In my Cajal preparations the
nervous elements of the synapse were demonstrated almost
exclusively as clean-cut deep brown fibers; the glia nuclei were
the only things impregnated distinctly besides the nerve fibers.
Figure 8 is reproduced from a preparation of Carassius; the lat-
eral dendrite of Mauthner’s cell is enveloped here in a sheaf of
unmedullated nerve fibers. Each nerve fiber is provided with a
thickening, which is to be homologized with ‘bontones de Auer-
bach’ of Cajal and lies more or less close to the surface of the
dendrite. Besides this, some fibers have one or multiple thick-
enings on their way to the cell and sometimes present the pic-
ture of a string of pearls. There is no doubt, that these are iden-
150 KIYOYASU MARUI
tical with the ‘bontones de trajecto’ of Cajal. Both these ‘bon-
tones’ are, generally speaking, spindle-shaped or spheroidal and
variously large. As far as my observations reached, there is no
particular difference in structure between these two kinds of
‘bontones.’ They are now and then impregnated massively or
as if punched; most of them show, however, the splitting up of
the axone fiber into several delicate neurofibrils in them, as
figure 8 x, indicates very distinctly. Some of them appear in
the .preparations in cross-section and then they show the cross-
sections of these delicate fibrils in them (fig. 8, xx). I must
emphasize here that I did not find any net structure in the ‘bon-
tones,’ as is described by many authors; the one, as we find in
figure 8 at xxx looks as though it had net structure, but I believe
it is by no means a real net-work; on the contrary, it shows
merely the splitting up of the axone fiber into multiple delicate
fibrils. Besides these large ‘bontones’ we find many minute
rings, lying close to the surface of the dendrite, and these rings
are continuous with very delicate fibers. The latter come from
the other fibers as their ramifications or from the end of the
large ‘bontones’ (fig. 8, xaxx}. It is quite obvious that these
rings are similar to those structures which were described and
figured by Cajal (13). Now, the nerve fibers, which possess the
above-mentioned qualifications, pass sometimes very near the
surface of .the dendrite, and thus remind us, especially in the
tangential sections, of fibers coming into contact with the cell
surface of the dendrite by means of their ‘bontones de trajecto,’
as was claimed by Cajal (13). But on careful observation, es-
pecially on examination of profile pictures, we can easily con-
vince ourselves that there is no contact between them; at least
I can state definitely that there are many of these ‘bontones de
trajecto,’ which lie quite remote and free from the surface of the
cell (fig. 8).
On the ventral dendrite of Mauthner’s cell I found a similar
condition in the synapse. In the ‘axone cap’ we see abundant
unmedullated nerve fibers, which, as far as my investigation went,
1orm a piexus of nerve fibers. Beccari (5) used the term ‘canes-
FINER STRUCTURE OF SYNAPSE Lol
tro,’ to express the condition of nerve fibers in the synapse; but
he did not give any statement about the behavior of the fibers
with reference to each other. We find nerve fibers of multi-
farious directions and ramifications of the fibers are very often
observed; but the anastomosis formation between them and
even the net formation of the nerve fibers could not at all be ob-
served in the ‘axone cap.’ The nerve fibers of the axone cap
showed also the ‘bontones’ and the minute rings. The whole
condition of the synapse is, generally speaking, similar to that
of the lateral dendrite.
Cajal (13) stated and figured that the endings of nerve fibers
come in contact with the cell surface not only by his ‘bontones
de Auerbach’ and the minute rings, but also by means of the
‘bontones de trajecto.’ His figure 9 indicates that the latter lies
quite close to the cell membrane of the ganglion cell. Economo
(15) claimed to have observed that the axis-cylinder enters into
connection with the nerve cell by multiple terminal feet, run-
ring on the surface of the ganglion cell. Heidenhain (16) de-
clared further that most of the nerve cells carry a very thick pelt
of end-knobs. As already described, there was no evidence in
my preparations of the connection between the cell surface and
the ‘bontones de trajecto.’ At least I can state that I have
found many of these ‘bontones’ quite free and remote from the
cell surface. The figures of Cajal (13) and Economo (15) do not
prove that they are connected with the latter. I have the idea
that they present merely an optical illusion, caused by the nerve
fibers with those ‘bontones’ passing quite near the nerve cell.
Cajal (12) at first characterized his ‘bontones’ as spheroidal
or elongated structures, impregnated solidly or at the most
punched. He regarded them as the foundation of his contact
theory, as he believed to have found that the nerve fibers end
on the cell surface with those structures. Dogiel (14) also de-
scribed ring- or net-shaped terminal structures and remained a
partisan of the contact theory. According to Held’s (19, 20)
observation with Cajal’s method, the neurofibrils of his terminal
feet form a net-work and communicate directly with the neuro-
1a KIYOYASU MARUI
fibril net-work of the nerve cell in two different ways. In one
group of the terminal feet the net-work is connected with the intra-
cellular neurofibrils by means of a feebly stained neurofibril net-
work, embedded in a homogeneous substance, and in the other
group delicate neurofibrils enter the cell body from the terminal
feet in a radial direction, also surrounded by homogeneous ma-
terial. Holmgren (22) distinguished also two types of terminal
feet; his first type was of annular shape and his second type
showed a tiny net-work, which is connected directly with the
neurofibril net. of the nerve cell. Later Cajal (13) also figured
and described net-shaped terminal and transitional knobs and
minute ring-shaped structures, which he believed to come into
contact with the cell surface. Heidenhain (11) declared even
that these are nothing but the net corpuscles, which appear in
the periphera' nerve endings. In my Cajal preparations, how-
ever, the net-like structure of these nerve endings, as remarked,
did never come to my observation. Some of the terminal feet
looked merely granular in my preparations, owing to a fine silver
precipitate; occasionally I observed in them a false net figure
caused by the irregular distribution of the latter. But the pic-
tures which are shown in figure 8 exclude beyond doubt the net
figure of the ‘bontones.’ I wonder if the terminal feet possess .
really the net structure, as it was claimed by Held (19, 20),
Holmgren (22), Cajal (13), and others. I suppose that this net
figure comes from artificial admixtures of the impregnation of
the honeycomb structure of the neuroplasm, which occurs now
and then in Cajal’s technic. To this question I shall return
later. |
As far as my observations go, the terminal feet must not be
regarded as the definite ends of the axone fibers, as was assumed
by Cajal and others. In my Cajal preparations I observed very
often that single or multiple delicate neurofibrils come from the
end of the terminal feet and advance toward the cell surface,
embedded in a homogeneous substance and entering the cell
body (fig. 12). I am very sorry to confess, however, that I
could not demonstrate definitely the neurofibril continuity in
FINER STRUCTURE OF SYNAPSE 153
these preparations, as the intracellular neurofibrils were often
not impregnated quite distinctly in them. So my finding corre-—
sponds to that type of the terminal feet of Held, in which the
fibrils were followed in radial direction from the terminal feet
into the cell body; the connection by means of a fibril net-work
did not come to my observation.
Figure 9 demonstrates the condition of the synapse in the
Bielschowsky preparation; on the surface of the axone cap, we
see a number of nerve fibers going into the region of the ‘axone
cap;’ Weigert and Heidenhain preparations show that they lose
their myelin sheaths just at the border of the ‘axone cap.’ Some
of these nerve fibers have each a large club-like expansion here
and before the loss of their myelin sheaths. According to my
experience, this phenomenon is more noticeable in Ameiurus
than in Carassius; in regard to the significance of this finding I
am not able to say anything definite as yet. On their way to
the surface of the cell the unmedullated fibers have single or
multiple spheroidal or spindle-shaped swellings; sometimes the
fibers look like a string of pearls showing many of these swel-
lings. There is no doubt that these enlargements are identical
to the ‘bontones de Auerbach’ and the ‘bontones de trajecto’ of —
Cajal. In my preparations these nodes came out partially dif-
fusely black and partially punched as Bielschowsky (18) described
them. Besides these I could find also ‘bontones’ with splitting
of the nerve fibers into numerous delicate fibrils (fig. 13), just
like those which were described above in my Cajal preparations.
The net-work, as was described by Wolff (27) from his Bielschow-
sky preparations, I could not demonstrate at allin them. In my
Bielschowsky preparations I could not find any place, where the
contact between the ‘bontones de traecto’ and the cell surface
takes place, as was described and figured by Cajal. It is quite
obvious in my figure 9 that at least there are many of those
‘bontones’ which lie quite remote and free from the cell surface.
From the peripheral end of the terminal feet single or multiple
delicate fibrils come out and proceed toward the cell surface,
surrounded by the homogeneous substance, and enter the cell
154 KIYOYASU MARUI
body communicating directly with the intracellular neurofibrils.
Figure 10 demonstrates clearly that the neurofibrils from the
terminal feet enter the cell body. In this figure it is also shown
that the intra- and extra-cellular neurofibrils communicate with
each other; one might raise the objection against me, that the
fibers which enter the cell-body might be the ‘cap dendrites’ of
Bartelmez (4). On this point I can give the following arguments:
first, some of the fibers could be traced back to the part, where
they are to be enveloped in myelin sheaths, and, second, it is
connected with the cell by means of the terminal foot (fig. 10).
Terminal ramification of nerve fibers occurs very often within the
‘axone cap;’ between the ends of individual nerve fibers neither
anastomosis formation nor even a simple net formation of nerve
fibers came to my observation in the ‘axone cap’ oron the cell
surface. Through careful examination I could always isolate the
sharply and intensely dark stained nerve fibers from each other.
In the previous chapter I already stated that through the simul-
taneous impregnation of the Golgi net substance a false picture
of a nervous net-work becomes observable in the synapse. Ac-
cording to my experience, the,Golgi net substance is to a certain
extent antagonistic in its staining reaction to the nerve elements
in the Bielschowsky method; when the nervous elements are
brought out sharply and dark enough the Golgi net substance
remains unstained or is stained quite feebly, whereas the latter
is impregnated more or less intensely when the former is stained
more feebly. I also stated above that by means of the eosin
counterstain the Golgi net substance is demonstrable in red color
and is connected directly with the glia reticulum and glia nucleus
on the one hand and the nervous elements on the other hand.
Figure 15 shows that the nervous elements are covered with a
more or less thick layer of the red-stained substance; also the
terminal feet are surrounded by the same substance. This pic-
ture corresponds to that of the Levaditi preparation, which was
described in the previous chapter.
After this description of my results in both the Cajal and
Bielschowsky preparations I go over the discussion of the ques-
FINER STRUCTURE OF SYNAPSE 155
tions referred to. As remarked, my findings in the Cajal and
Bielschowsky preparations are in opposition to those of Held
(19, 20), Cajal (13), Holmgren (22), Wolff (27), and others, in so
far as I demonstrated in the terminal feet merely the splitting
of the nerve fiber into fine fibrils instead of a net structure. Now
it is remarkable that some of the above-mentioned authors (Ca-
jal, Held, and others) assume also a net structure of the intra-
cellular neurofibril, while others (Wolff, 27) deny the net struc-
ture in the cell body, at least in many kinds of nerve cells. I
could not find any real net structure in either the Mauthner cell
or in the terminal feet. It appears extremely interesting to me
that in Wolff’s (27) figures Biitschli’s honeycomb structure was
stained in the cell body as well as in the dendrites. Economo
(15) supposed that the net figure of the terminal feet might de-
pend upon the simultaneous impregnation of the Biitschli struc-
ture. Also Auerbach (2), who was at first a partisan of the
contact theory, declared that the neurofibrils do not form a reti-
culum in the terminal feet,. but that one, two, or three delicate
fibrils go radially into the cell body, embedded in the ground
substance of the terminal feet.
Ramon y Cajal (12, 13), Dogiel (14), Retzius (27), Heiden-
hain (16),and others despite the repeated argumentation of the
antagonists Held (19, 10), Holmgren (22), Bielschowsky (9), An-
toni (17), Auerbach (3), and others) remained firm on the stand-
point of the contact theory. According to Held, Bielschowsky,
and others, however, the theory of the contact must result from
the imperfect impregnation of the structure in question; they
observed, as remarked, that the fibrils of the terminal feet go
into the cell body and enter into relation with the cell fibrils. I
could also confirm the neurofibril continuity in Mauthner’s cell
in the Bielschowsky preparations. The ‘bontones de Auerbach’
of Cajal evidently must not be regarded as the contact organs,
in which the nerve fibers come to their ends, but are rather tobe
interpreted as the stations in the course of nerve fibers, where
the modification of the substance takes place, which was claimed
by Bielschowsky (8). With the supposition of the latter, how- .
156 KIYOYASU MARUI
ever, who took for granted that here the dissolution of the nerve
fibers occurs as a consequence of the loss of the perifibrillar
cement substance, I cannot agree. On the contrary, I assume
that the dissolution of the fibers is to be regarded as the result
of the accumulation of the perifibrillar neuroplasm.
Bartelmez (4) described in the ‘axone cap’ of Mauthner’s
cell two kinds of endings—‘free endings’ and ‘knob endings,’
which latter are in contact with the cell surface; besides these he
mentioned on the lateral dendrite ‘club endings.’ As far as my
investigation went, the ‘free endings’ of Bartelmez are very
hard to accept; though I found in my preparations nerve fibers
which do not reach to the cell surface, it is always probable that
they appeared in section. Nor can I find any essential difference
between the endings in the ‘axone cap’ andin the lateral dendrite,
as already described. Moreover, he did not state the structure
of these endings in detail; he figured as ‘pericellular net’ (figs.
12, 13) a minute solid or ring-shaped structure, which is perhaps
identical with the ring-shaped ending apparatus described by
Cajal. Above all I must emphasize that I found the neurofi-
bril continuity on the cell surface as well as on the dendrites;
the ‘plasma membrane,’ which Bartelmez described on the sur-
face of his club endings, must not be regarded as the last end
of the nerve fibers.
Held (18) assumed a ‘pericellular nervous terminal net,’ which
exists on the cell surface alternating with the Golgi net, as re-
marked before. In the previous chapter I showed that there is
only one net-work on the cell surface formed by both the Golgi
net and the nervous elements. Now, in my Cajal preparations,
which do not show the Golgi net substance, I was not able to
find any net-work in the synapse. I showed that Held’s figures
of his neurosome and Golgi preparations do not argue for his
hypothesis. It appears extremely interesting to me that in his
Cajal preparations the deeply stained terminal feet are not at
all (19) connected by bridges or at the most connected among
each other by feebly (20) stained bridges. Moreover, it must
be remembered here that in the neurosome preparations of Held
FINER STRUCTURE OF SYNAPSE LST
the ‘terminal feet’—the nodal points of his terminal net—were
chiefly granular, whereas the beams between them appeared free
from granules. Auerbach (2, 3) also assumed the existence of
the nervous terminal net, whose nodal points coincide with his
‘terminal knobs;’ but his figures prove nothing. Besides, it must
be emphasized that his net-work possesses a three-dimensional
character not only on the cell surface, but also in certain other
parts of the central nervous system—the substantia gelatinosa
and the molecular layer of the cerebellum. The existence of a
nervous net-work of this sort, which would penetrate the gray
matter of the central nervous system, was denied by many au-
thors (Cajal, (13) Kdlliker (cited in 9), Retzius (26), etc.).
Even Held (20) admitted that by means of Cajal’s technic he
could observe merely the indication of the extension of the ‘peri-
cellular nervous terminal net.’ The figures of Holmgren (22)
also prove nothing.
Concerning the pericellular net-work, which is found now and
then in Bielschowsky preparations, I expressed my idea before,
that there might be a false picture caused by the simultaneous
impregnation of the Golgi net. This consideration will perhaps
be strengthened by the description and the figures of Wolff (27).
He figured on the surface of the cells as well as between the
latter a net structure, which shows a striking resemblance to
Bethe’s net-work and is mainly stained more feebly than the
terminal feet, which latter are directly continuous with the former.
In the text he stated that the axone fibrils are not naked, but
are enveloped in a mantle of bubble-like structure, which goes
over into the honeycomb structure of the cell border. On the
ground of his findings and Gegenbauer’s intercellular bridge
theory, Wolff declared that Bethe’s pericellular net-work is
nothing but the impregnated honeycomb structures of the cell
border, the neuroplasmatic anastomoses and the perifibrillary
mantles. That Bethe’s net-work is not, however, to be regarded
as the honeycomb wall, but as a glious structure, I have already
stated.:
158 KIYOYASU MARUI
SUMMARY
1. In the ‘axone cap’ and on the surface of Mauthner’s cell a
Golgi net structure was very distinctly demonstrated by means
of Levaditi’s method; in Heidenhain and other preparations a
similar net-work was brought out. The Golgi net is of glious
nature and is in close relation to the nervous elements in the
synapse. According to the results of this study, the unmedul-
lated parts of the nerve fibers are also enveloped in a sheath of
glious tissue; the finer structure of this glia sheath is as yet un-
known. |
2. The hypothesis of Held concerning the existence of two
kinds of net-work on the cell surface—a Golgi net and a ‘peri-
cellular nervous terminal net’—is denied; there exists, as far as
my observations go, only one net structure, which is formed by
both the nervous and the glious tissues. The so-called pericel-
lular nervous terminal net is not to be regarded as a real nervous
net-work, but to be considered as a picture produced by the
simultaneous stain of the Golgi net-work.
3. The contact theory is a histological impossibility. The
terminal feet cannot be regarded as the specific contact organs,
but as the points in the course of axone fibers, where the dissolu-
tion of fibers takes place. The continuity of the intra- and
extracellular neurofibrils is very clearly demonstrated.
4. The intracellular neurofibrils do not form any reticulum in
Mauthner’s cell; nor did the net structure which was described
by many authors in other animals, ever come to my observation
in the nervous terminal feet. The nerve fibers showed merely a
splitting into numerous delicate fibrils in them.
10
11
13
14
18
19
FINER STRUCTURE OF SYNAPSE 159
LITERATURE CITED
Antoni, N. 1908 Die Frage von einer neurofibrilliren Kontinuitit im
Zentralnervensystem der Wirbeltiere. Folia neuro-biologica, Bd. 2
(Sammelreferat).
AUERBACH, L. 1899 Das terminale Nervennetz in seinen Beziehungen zu
Ganglienzellen der Zentralorgane. Monatschift f. Psychiatrie u.
Neurologie, Bd. 6.
1904 Extra sowie intrazellulire Netze nervéser Natur in den Zentral-
organen von Wirbeltieren. Anatomischer Anzeiger.
Bartetmez, G. W. 1915 Mauthner’s cell and the nucleus motorius teg-
menti. Jour. Comp. Neur., vol. 25.
Beccart 1907 Ricerche sulle cellule e fibre del Mauthner e sulle loro con-
nesioni in pesci ed anfibii. Arch. Ital. Anat. e. Embr., vol. 6.
Betue, A. 1898 Uber die Primitivfibrillen in den Ganglienzellen vom
Menschen und anderen Wirbeltieren. Morphologische Arbeiten, Bd.
8.
1900 Uber die Neurofibrillen in den Ganglienzellen von Wirbeltieren
und ihre Beziehungen zuden Golginetzen. Arch. f. mikroskop. Ana-
tomie, Bd. 55.
Birtscuowsky, M. 1904 Silberimpriignation der Neurofibrillen.. Journ.
f. Psycholog. u. Neurolog., Bd. 3.
1905 Die histologische Seite der Neuronenlehre. Journ. f. Psycholog.
u. Neurolog., Bd. 5.
1908 Uber die fibrillare Struktur der Ganglienzellen. Journ. f. Psy-
cholog. u. Neurolog., Bd. 10.
BretscHowsky, M., anv Wourr, M. 1904-1905 Zur Histologie der Klein-
hirnrinde. Journ. f. Psycholog. u. Neurolog., Bd. 4.
Casat, S. R. 1903 Consideraciones criticas sobre la teoria de A. Bethe
acerca de la estructura y conexiones do las células nerviosas. Tra-
jabos de Laboratorio de investigaciones biologicas de la Universidad
de Madrid, T. 2, cited in Heidenhain, Held, etc.
1908 L’hypothése de Mr. Apathy sur la continuité de cellules nervenses
entre elles. Anatomischer Anzeiger, Bd. 23.
Doatrt, A. 8S. 1904 Uber die Nervenendigungen in den Grandryschen und
Herbstschen Kérperchen im Zusammenhange mit der Frage der Neuro-
nentheorie. Anatomisch. Anzeiger, Bd. 25.
Economo, D. 1906 Beitriige zur normalen Anatomie der Ganglienzelle.
Arch. f. Psychiatrie u Neurologie, Bd. 41.
HEIDENHAIN, M. 1911 Plasma u. Zelle.
Hexp, H. 1897 Beitrige zur Struktur der Nervenzellen u. ihre Fortsitze.
Arch. f. Anat. u. Physiolog., Suppl., Anat. Abt.
1902 Uber den Bau der Grauen u. Weissen Substanz. Archiv f.
Anat. u. Physiol., Anat. Abt.
1904 Zur weiteren Kenntnis der Nervenendfiisse und zur Struktur
der Sehzellen. Abhandlungen d. mat. -phys. Kl. d. Kogl.-sichs.
Ges. d. Wissenschaft, Bd. 30.
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 1
160 KIYOYASU MARUI
20 1905 Zur Kenntnis einer neurofibrilliren Kontinuitét im Centralner-
vensystem der Wirbeltiere. Arch. f. Anat. u. Physiolog., Anat. Abt.
21 1903 Uber den Bau der Neuroglia und iiber die Wand der Lymphge-
fisse in Haut u. Sehleimhaut. Abh. d. math.-phys. cl. d. Kégl-
Sichs Ges. d. Wissen., Bd. 28.
HoutmaGren, F. 1905 Uber die sog. Nervenendfiisse (Held). Jahrbiicher f.
Psych. u. Neurolog., Bd. 26.
23. Meyer, 8. 1896 Uber eine Verbindungsweise-der Neuronen. Nebst Mit-
teilungen iiber die Technik, und die Erfolge der Methode der subkutanen
Methylenblau-injektion Arch. f. mikroskop. Anat., Bd. 47.
24 1897 Uber die Funktion der Protoplasmafortsitze der Nervenzellen.
Berichte d. kégl.-siichs. ges. d. Wissen. math.-phys. Kl., Bd. 47.
25 Pauapino, G. 1913 Continuity in the vertebrate nervous system and the
mutual and intimate connections between neuroglia and nerve cells
and fibers. Review of Neurolog. and Psych., vol. 11.
26 Rerztus, G. 1908 The principles of the minute structures of the nervous
system as revealed by recent investigations. Croonian lecture. Pro-
ceedings of the Royal Society, vol. 80.
27 Wourr, M. 1904-1905 Zur Kenntnis der Heldschen Nervenendfiisse.
Journ. f. Psycholog. u. Neurolog., Bd. 4.
bo
bo
All figures are taken from preparations of Mauthner’s cell.
The photomicrographs were not retouched at all. An achromatic Zeiss ocular
no. 4 and a Zeiss oil-immersion ;'s were used. In figures 1 and 8 the length of
bellows was 100 em. and in all the others it was 60 em. The figures 11, 12, 13, 14,
and 15 were drawn using the Abbe camera lucida with slight rotation of the mi-
crometerscrew. (Apochromatic Zeiss ocular no. 4, oil-immersion ;'s, tube length
20 cm.) In those drawings except 12 and 13 the glia structure is presented in a
gray, and the nervous structure in a dark color. It is tobe added here that figure
7 came from a fatigue preparation.
ae - ~
+> ol ae po ph
_e Par - 4%
Pet ee ek
, Me tee
Fig. 1 Carassius auratus. Bielschowsky preparation.
Fig. 2 Carassius auratus. Levaditi preparation.
161
%
Fig. 3 Ameiurus nebulosus. Levaditi preparation.
Fig. 4 Ameiurus nebulosus. Levaditi preparation.
162
Fig. 5 Ameiurus nebulosus.
Fig. 6 Carassius auratus.
Heidenhain preparation (formol material).
Levaditi preparation.
163
Fig. 7 Ameiurus nebulosus. Heidenhain preparation (formol-Zenker mate-
rial).
Tig. 8 Carassius auratus. Cajal preparation.
164
Fig. 9 Ameiurus nebulosus. Bielschowsky preparation.
Fig. 10 Ameiurus nebulosus. Bielschowsky preparation.
165
12 13
Tig. 11 Carassius auratus. Levaditi preparation.
lig. 12 Carassius auratus. Cajal preparation.
Fig. 13 Ameiurus nebulosus. Bielschowsky preparation.
166
Fig. 14 Carassius auratus. Levaditi preparation.
- Fig. 15 Carassius auratus. Bielschowsky preparation with
eosin counter-stain.
167
hare
‘alll
; bal cits al
Ft
.
AUTHOR’S ABSTRACT OF THIS PAPER
ISSUED BY THE BIBLIOGRAPHIC SERVICE
APPLICATION OF THE MARCHI METHOD TO THE
STUDY OF THE RADIX MESENCEPHALICA
TRIGEMINI IN THE GUINEA-PIG
WILLIAM F. ALLEN
Department of Anatomy of the University of Oregon Medical School, Portland,
Oregon
THIRTY-FIVE FIGURES
CONTENTS
RRC IGEN 92 AE tt a2 cc cha 'e's gible epsea eres aise Fee hee Rares; « ‘: eee 169
Feewioweor tue lmperapOre 1608.
6 A somewhat diagrammatic view of a sagittal section of the spinal cord
(neural tube) near the middle of the body. A connection between Reissner’s
fiber and the fibrils of ependymal cells is simulated. The cell in the lumen of
the tube is helping to form, or to repair (probably the latter) Reissner’s fiber.
60 days. X 750.
7 Sagittal section of spinal cord near the middle of the body, showing
Reissner’s fiber apparently free from the walls of the neural tube. The dorsal
wall is not shown in its whole thickness. 100 days. X 750.
8 Sagittal section in the region of the sinus terminalis. X 325.
9 Posterior end of the section shown in figure 8, at a slightly different focus
and more highly magnified. Several cells are seen to be in process of meta-
morphosing into the posterior end of Reissner’s fiber; others, apparently, are
just entering the lumen of the neural tube. 750.
10. Sagittal section through the sinus terminalis, showing the terminal
plug. There is only one cell in process of metamorphosing into the Reissner
fiber, and this is, apparently, in continuity with the connective tissue lying
beyond the posterior end of the neural tube. X 325.
ABBREVIATIONS
a, anterior fbrl."’, cross-sections of fbrl.’
cl. e’end., ependymal cell 7., internal limiting membrane
cl. R., cells in process of forming Reiss- _zvlr., sheath of fbrl.’
ner’s fiber nl. e’end., ependymal nucleus
coag., coagulum nl., nucleus
co’ms. p., posterior brain commissure’ nl. n., nucleus of nerve cell
d., dorsal obt. trm., terminal plug of neural tube
ex., external limiting membrane p., posterior
fbr. p., fibers of posterior commissure par. n., wall of neural tube
fbr. R., Reissner’s fiber sb-co’ms., sub-commissural organ
fbrl., fibrils from deep (proximal) ends tis. co’nt., connective tissue
of ependymal cells v., ventral
fbrl.’ fibrils from superficial (distal)
ends of ependymal cells
4 hes Poe elie, i Ae? A
REISSNER’S FIBER IN TELEOSTS PLATE 1
HOVEY JORDAN
tis.cont ><
A 8) ae De
i uA PAGER Oey yoo
LI TIIS é
——— fe
Tok. is
~Z tis.cont at
ie”
_forR,
PES CERIRE RA)
AT GLEE:
Resumido por el autor, Robert Sidney Ellis.
Estudio preliminar cuantitativo de las células de Purkinje en el
cerebelo normal, subnormal y senfl del hombre, con
algunas notas sobre la localizacién funcional.
En el presente trabajo se compara el ntiimero de células de
Purkinje en el cerebelo normal, subnormal y senil del hombre.
El autor ha hecho correcciones para las. diferencias de volumen
de los cerebelos estudiados, de modo que los valores consignados
representen el ntimero de células existentes en dreas equivalentes,
tomadas como unidad. Ha contado las células de seis dreas de
cada mitad del cerebelo, determinando de este modo el ntiimero
de células de Purkinje que existen normalmente en areas equiva-
lentes del cerebelo de tamafio normal. Comparando con la norma
asi establecida ha podido comprobar que el nimero de células de
Purkinje es muy deficiente en el cerebelo de los idiotas-e im-
béciles. Del mismo modo, en el cerebelo del viejo se observa
también una disminucién progresiva en el ntimero de dichas
células. Existe cierta correlacién entre estas deficiencias celu-
lares y las deficiencias en la coordinacién motriz.
Translation by Dr. José F. Nonidez.
Columbia University
AUTHOR'S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, JANUARY 6
A PRELIMINARY QUANTITATIVE STUDY OF THE
PURKINJE CELLS IN NORMAL, SUBNORMAL, AND
SENESCENT HUMAN CEREBELLA, WITH SOME
NOTES ON FUNCTIONAL LOCALIZATION
ROBERT 8. ELLIS
The Wistar Institute of Anatomy and Biology and The Training School at
Vineland, New Jersey
TWO FIGURES AND ONE CHART
INTRODUCTION!
In the spring of 1916, while examining the cerebellum of a
‘general paralytic, the writer was first impressed by the fact,
familiar perhaps to most neuropathologists, that in this disease
there is often a disintegration and disappearance of a large
number of the Purkinje cells, leaving, however, the basket of fibers
which normally surrounds them. Over a year later, while ex-
amining the cerebellum of a microcephalic idiot, the same scarcity
of Purkinje cells was observed, with the difference, however,
that the section did not show the same evidence of the cells
having become reduced in number by disintegration; the empty
pericellular baskets were not found as in the case of paresis; it
seemed, rather, that through some defect of development the
normal number had never been present.
This difference presented an interesting problem and it seemed
worth while to make a more careful quantitative study of the
Purkinje cells in different types of cerebella. In order to get
1 The writer is indebted to the staffs of the Vineland (N. J.) Training School
and of St. Vincent’s Hospital (Philadelphia) and to several physicians in other
hospitals for assistance in securing the material used in this study. He is also
indebted to the staff of the Wistar Institute, especially Drs. Greenman, Donald-
son, and Hatai, for their encouragement and assistance in numerous ways during
the course of the investigation: to all of these he wishes to express his
appreciation.
229
230 ROBERT S. ELLIS
a fair basis for comparison, a number of cerebella were studied
and the relative frequencies of cells noted. In some of the cases
the cells appeared to be almost uniformly distributed and with
few large spaces between them; others showed losses similar to
the two cases already mentioned.
Among the cerebella examined was one of a man who had died
at about the age of sixty-five years after a protracted illness, and
this, too, showed a distinct loss of cells. So from this preliminary
set of observations it seemed clear that the number of Purkinje
cells is variable under different conditions.
It is well known that in paresis, in extreme old age, and in
low grades of feeble-mindedness there is ordinarily a considerable
degree of deficiency in motor coérdination. The question con-
sequently arose, how far is it possible to find differences in the
number of cells that will account, partially at least, for the
observed differences in behavior?
The writer’s primary interest at the time of taking up this
investigation lay in the question of the anatomical basis of mental
defect, and it seemed not improbable that a careful study of the
Purkinje cells might throw some light on one of the most evident
deficiencies found in such cases. The human motor mechanism
is much more highly developed than that of lower forms,
especially with reference to speech, hand movement, and the
maintenance of equilibrium while standing or walking. Mental
defectives generally show less motor control along these lines,
and it is desirable that we know as far as possible the neural
basis for such lack of codrdination.
A further reason for making a study of the cerebellum in such
cases is found in the fact that a number of writers, especially
Tredgold (’03) and Bolton (’03, 710) in England, have em-
phasized, perhaps unduly, the importance of the frontal lobe of
the cerebral cortex as the area particularly affected in amentia.
It accordingly seemed worth while to determine whether the
brains of aments show defects in other parts, such as the cere-
bellum, which is not generally associated with intelligent
reactions as such.
QUANTITATIVE STUDY OF THE PURKINJE CELLS 231
In order to determine the nature and extent of the variations
in the Purkinje cells, and especially to determine the differences
between normal and subnormal cerebella, a careful study of
the numerical distribution of these cells has been made, the
results of which it is the purpose of this paper to present.
MATERIAL
For the best results in studying the present problem it would
be desirable to have cerebella in perfectly normal condition to
be used as the standards with which the subnormal ones are
compared. These normal cerebella should be from individuals
in the prime of life, who were known to have had good physique
and good motor control and who had died either from accident
or from some acute disease which did not cause a disintegration
of nervous elements. Pathological material is to be had in
abundance, but to secure any number of approximately normal
cases is next to impossible. In the present study, consequently,
it has been necessary to use as our norms cerebella from the
museum collection of The Wistar Institute. No claim is made
that this material is ideal; it is hoped, however, that sufficient
care and caution have been used, both in the procedure followed
and in the interpretation of results, to prevent serious errors
arising from the material employed.
To obtain norms, the cerebella of three adult whites and of
five adult negroes, all males (thirty to forty-two years of age),
were sectioned and a preliminary study was made to determine
whether or not they had suffered any loss of Purkinje cells. In
nearly every case there was at least some maceration and some
loss of these cells, but four of the number, three negroes and one
white, showed very slight losses, and these were accordingly
used as the material for determining the norm. These are the
cases listed in table 1. The other four cases were rejected
because microscopical examination clearly showed that cells had
disintegrated and dropped out, and it would consequently have
been impossible to determine from them the number of cells
normally present.
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 2
232 ROBERT S. ELLIS
Little or no information of importance regarding the clinical
histories of these cases was available. They were simply cases
that had come to autopsy and their brains had been preserved
as a routine measure. Even so, however, the results from the
four should give us a close approximation to the conditions
present in the average cerebellum as it appears in the hospital
population—a group probably somewhat below the average for
the community at large in the development of the nervous
system.
Nine cerebella from mental defectives, who had died at the
Vineland Training School, were selected from the collection in
The Wistar Institute Museum and a study was made of each
case. Several of these cases have been described by Goddard
(Feeble-mindedness, New York, 1914), and his case numbers,
preceded by ‘G,’ enclosed in parentheses, are given after the
serial numbers which these brains have on the records of The
Wistar Institute. The histories of these defectives, all whites,
and listed in table 2 and table 4, are as follows:
14880, male, age 23 years; alcoholism and insanity in family; three
miscarriages before the birth of this child; he had had convulsions,
marasmus at one year, measles at four years; small in size, head small;
poor motor power, poor gait, poor speech; no will power; right handed;
fair memory.
15144, male, age 34 years; height, 6 ft. 3 in.; brain weight 1619
grams; supposed insanity and feeble-mindedness on father’s side; no
deaths in infancy in family; had whooping-cough, measles, and pneu-
monia; walked at three years; head, bullet-shaped; right side not
developed; dragged feet, could thread needles and tie shoes; talked
little; could not count.
15145, male, age 24 years; brain weight 1491 grams; normal (?)
heredity; first child dead; had measles and whooping-cough; epileptic;
rs on right hip; weak heart; normal electric reaction; grip, rt.
0-It. 0.
15214, male, age 26 years; brain weight 1065 grams; father alcoholic;
five children died young; first child a ‘lunatic,’ followed by two mis-
carriages; two other children mentally defective; this, the sixth child;
probably congenitally luetic; spasms at nine months; walked at three
years; talked at four years; had measles and whooping-cough, ab-
scesses, scrofula, Pott’s disease; hunchback, T.B. of vertebrae; weak
heart; was apparently normal until nine months; not strong enough to
work; grip, rt. 18-It. 22; could dress self; learns quickly, forgets soon.
This case is presented by itself in table 4.
QUANTITATIVE STUDY OF THE PURKINJE CELLS Das
15250 (G. 285), male, age 43 years, mental age three; nervous family;
epileptic, had measles and whooping-cough, brain fever at two years;
side of face deformed from atrophy of jaw; stooped shoulders; left arch
fallen; irregular heart; poor vision, eye fatigue and headache; no
abnormal movements; exaggerated reflexes; grip, rt. 21-It. 23.
15297 (G. 203), male, age 26 years, mental age eight, apparently
normal heredity; two deaths in infancy and two other children feeble-
minded; physicians believed defect due tio congenital lues; did barn
and garden work, grip, rt. 33—It. 31.
15299 (G. 195), male, age 36 years, mental age two; brain weight 335
grams; father probably feeble-minded; this case the first of seven
conceptions; two later ones being miscarriages; had whooping-cough
and scarlet fever; large head, left scapula higher than right, left testicle
undescended; fallen arches; knock-kneed; did no work; grip, rt. 15-
lt. 17; could not talk, but would swear freely.
15310 (G. 258), male, age 18 years, mental age four; brain weight
1051 grams; ancestry possibly neuropathic; this case Mongolian in type;
Wassermann positive; knock-kneed and stoop-shouldered: died of T.B.
after extreme wasting away; left testicle undescended; normal reflexes;
ate well and played well; left-handed; grip about .18 for each hand; did
not alternate feet in climbing stairs; speech poor; subject to headaches.
15320 (G. 327), female, age 20 years, mental age one; mentality of
ancestry doubtful; had large tumor in anterior part of the right hemi-
sphere of the cerebellum; began to grow very weak at eighteen years;
walking became poor, and a year later had difficulty in swallowing;
helpless in both hips and hands, left hand especially weak; pupil
reflexes lost, other reflexes exaggerated.
Subnormal infants. For purposes of comparison, the cerebella
of six infants, four negroes and two whites, were studied. These
cases are of the type found in charity hospitals where nearly all
are below normal and where a very large percentage are illegiti-
mate. Out of twenty such cases, where the brain weights were
taken, only two brains weighed as much as the average for their
age. Consequently, though the point could not be determined
individually, we are safe in assuming that they represent a
distinctly subnormal group of the population.
Senescents. The cerebella of five cases of senescence were
studied. Enough is known of each of these to make it fairly
certain that the changes observed were due primarily to old age
rather than to other causes. The age, color, and sex of these
are given in table 5, which shows the results of the cell counts.
Paresis. While making this study, the cerebellum from a
234 ROBERT S. ELLIS
case of paresis was sectioned before the cause of death was known,
and it seemed desirable to include it as an example of what may
happen in this disease. It is from a white man who was of
average physique and who died at the age of thirty-three years.
The data are given in table 6.
In the course of this investigation careful counts have been
made on parts of about forty cerebella and sections from about
twenty-five others have been less carefully examined. The
twenty-five cases on which the figures given in this paper are
based are believed to be typical of the classes under which they
are listed.
PREPARATION OF MATERIAL
The cerebellum was removed from these brains, weighed, and
the weight in grams recorded. In cases where the specific
gravity had changed materially from being for several years in
alcohol, the specific gravity was determined and the brain weight
was corrected to the normal weight for a cerebellum of that
volume. No change of consequence was observed in the specific
gravity of those cerebella which had been preserved in formalin.
Consequently, no corrections were made on the weights as taken.
After weighing, three’ blocks were taken from each cerebellum.
The entire vermis was removed and cut so that a sagittal section
could be made of it entire. Then blocks were cut which would
give sections nearly through the middle of each hemisphere and
at right angles to practically all the folia. The plane of these
was so selected as to cut the lobus biventer, or paramedianus
(Bolk), on the under surface, to pass through the dentate nucleus,
and to cut the anterior dorsal edge of the hemisphere near the
vermis. A section so made shows for comparative purposes
the areas of each hemisphere to which different functions have
been assigned by Bolk (05), Rynberk (’07, ’12), et al.
Figure 1 shows the plane in which the sections were taken,
and figure 2 the appearance of such a section and gives the
localization pattern advanced by the above writers.
‘The blocks of hardened tissue were cut about 5 mm. thick.
Their maximum length and maximum breadth were then meas-
QUANTITATIVE STUDY OF THE PURKINJE CELLS 235
ured in millimeters and the measurements recorded on the cards
with the weights.
The procedure in dehydrating and embedding the blocks from
material, preserved in 10 per cent formalin was as follows:
Alcohol 50 per cent + Formalin 6) per) cent; < « éc4%.). 0c. oe cece 2 hours
Alcohol 70 per cent + Formalin 4 per cent.................... 15 hours
Aleohol 90 per cent + ‘Formalin 2 per cent:................... 8 hours
Ploomorre wer Cent: «..\) -.'skih a Meats ere Eek Lhe co tee 24 hours
Alcohol + Ether, equal parts..... oe eS OF ee Se 24 hours
eI tO. DEM CONL, 2 ¥ ay. mace einai ivclaeie's) 015 ete: 2 to 4 days
UL ING SCA 0 ae ae nO iE te okt 1 oR a 6 hours
SEED 4 58S Se eR ER ee he os at Larter ae eee 1 hour
Benzole-paraffine....... a Sif, 20h bye Reg aught eee ea ire 18 hours at 40°C.
BACAR CRETE Gisic1s cc cs ciecsa cls: craie este RIG aoe 3 to 6 hours at 52-58°C.
Fig. 1 Human cerebellum seen from above. The line S.S. marks the plane
of the section which is shown in figure 2.
In the experience of the writer, old formalin cerebella tend to
macerate easily, and the molecular layer especially shows a
tendency to break away from the internal granular layer. Where
this occurs there is always the possibility that some of the
Purkinje cells will drop out during the process of cutting and
staining. When the above procedure is followed, however, this
maceration is not usually found and good preparations are
secured. The common practice’ of washing formalin material in
water before dehydrating makes the molecular layer more likely
to break away from the rest of the cerebellum, and it should
consequently be avoided. Material preserved in alcohol was
treated as above with the exception that the blocks were put at
once in 90 per cent alcohol without formol. ;
236 ROBERT 8S. ELLIS
After embedding, the sections were cut on a rotary microtome
at 25 u, fixed to the slide with Meyer’s albumen, and stained with
carbol-thionine and eosin. Better results are sometimes secured
with the cerebella of infants if Delafield’s haematoxylin and
Orange G are used instead. Care must be taken that the
sections be not treated with absolute alcohol, as this dissolves
the parlodion. Equal parts of absolute alcohol and chloroform
may, however, be safely used (King, 10).
The most satisfactory fiber preparations from old formol
material were secured by Bielschowsky’s pyradine method. With
a modification of this the pericellular basket fibers have been
clearly shown in cerebella that had been in formol for thirteen
Fig. 2. Section through a hemisphere of the human cerebellum in the plane
shown in figure 1. The designations are explained on page 237.
years. If the pyradine is not thoroughly washed out before
putting into silver, and a weak solution of silver nitrate, 0.5
to 0.75 per cent, is used and changed frequently, the fibers will
be impregnated with silver in one or two days. The sections
are then treated as usual. These fiber preparations give valuable
assistance in determining the character and causes of the
deficiency in cells.
After staining, the sections on the slide were measured and
the measurements recorded on the cards with the original
measurements of the blocks as cut from the cerebellum at the
time of weighing. From these two measurements the percentage
of shrinkage is calculated. This will be referred to later.
QUANTITATIVE STUDY OF THE PURKINJE CELLS Zoe
METHOD OF STUDY
The purpose of this study, as has been stated above, was to
compare the number of Purkinje cells in different types of brains,
and also to find what numerical differences, if any, exist between
different areas of the same cerebellum. Reasoning by analogy
from what is known of the cell losses in the cerebrum, it seemed
not improbable that some areas might be found more subject to
degeneration than others, and, in view of the localization theory
of Bolk, it appeared all the more desirable to make a comparison
of the different areas to which different functions have been
credited. For this purpose each hemisphere has been divided ~
into six areas,? as shown in figure 2.
1. Lobus anterior (Bolk); head area; the region anterior to
the sulcus primarius (S. pr.)
2. Lobulus simplex (Bolk); neck area; from the sulcus
primarius to the sulcus postclivalis (S. pel.)
» 3. Lobulus semilunaris superior; arm area; from the sulcus
postclivalis to the sulcus horizontalis magnus (S.h.m.)
4. Lobulus semilunaris inferior; arm or leg area; from the
sulcus horizontalis magnus to the sulcus pregracilis (S. prg.)
5. Lobulus gracilis; leg area; from the sulcus pregracilis to the
suleus postgracilis (S. psig.)
6. Lobulus biventer; leg area (?); part posterior to the sulcus
postgracilis, exclusive of the amygdala.
In order to secure exact and strictly comparable measurements
of the relative numbers of cells in these different areas and of
the relative numbers in the same areas of different cerebella, the
following method has been used:
_ The sections were projected at a magnification of 30 diameters
by means of the Edinger projectoscope, and tracings made in
pencil of the line of Purkinje cells to be counted. The length
of this line as traced was measured in millimeters by means of a
map measurer. This divided by 30 gave the actual length of
? The first two designations (1-2) are given according to Bolk ’05 while the
remainder (3-6) are those used in Quain’s Elements of Anatomy, vol. 3, Neu-
rology, part 1, eleventh edition, 1908.
238 ROBERT S. ELLIS
the line on the slide. The number of cells in the line was then
counted under the microscope, only those cells showing the
nucleolus being counted, with the exception that when the cells
were so disintegrated that the nucleolus would not appear in
the preparation, the cells were counted if they appeared to
belong properly to the section. Dividing this total by the
length of the line in millimeters gives the number of cells in a
line 1 mm. long on the slide, which, as has been stated, is for a
section 25 uw thick. To correct this for shrinkage during dehy-
dration and embedding, the value thus obtained for 1 mm. on
the slide has been multiplied by the square of the percentage
obtained by dividing the sum of the length and breadth of the
section on the slide by the sum of the length and breadth of the
block as first cut from the cerebellum. This gives the number
of cells found in a line of Purkinje cells 1 mm. long and 25 yu
thick, this being really a surface 25 » by 1 mm. in the cerebellum
as weighed.
For comparable quantitative results this correction is neces-
sary because it is evident that when the cerebellum shrinks,
the line of Purkinje cells shrinks also, and it is necessary to correct
both for the change in the length of the line and for the change
in the thickness of the section on the slide—it being evident
that the 25u thickness of the section represents a greater thick-
ness in the cerebellum as weighed. The sum of the length and
breadth of the section have been used rather than one dimension
alone, because careful measurements show that the section is
compressed in breadth to some extent when cut on the micro-
tome. The sum of the two measurements consequently gives a
more accurate indication of shrinkage.
The number of cells found in a line 1 mm. long and in a section
25u thick would afford a satisfactory unit for comparing different
areas of the same brain; it would not do, however, asa unit for com-
paring different brains. If the molecular layer of the cerebellum
should be removed, leaving the cell bodies of the Purkinje cells
intact, it is evident that these would extend in a much convoluted
sheet over the surface of the remaining part of the cerebellum.
The area occupied by the Purkinje cells in a large brain is thus
larger than that occupied by them in a small brain. The sur-
QUANTITATIVE STUDY OF THE PURKINJE CELLS 239
faces of similar solids vary, we know, as the squares of the cube
roots of their volumes. A priori we should perhaps not be safe
in assuming that large and small cerebella are exactly similar
solids, it being possible that small cerebella might be more
convoluted than large ones so that the disparity in surfaces
might to some extent at least be removed.
Careful examination, however, of large and small cerebella—
where the latter are from cases at least one month old—does not
justify this a priori assumption. Individual differences do exist,
but size has not been found to alter materially the convolution
pattern after birth (Berliner, ’05). Furthermore, our results
based on the assumption that large and small cerebella are
similar solids will be found to agree with theoretical expectations.
In order, then, to compare areas which represent the same
fraction of the total number of Purkinje cells in different cere-
bella, it is necessary to take areas which vary as the squares of
the cube roots of their volumes. As a convenient indication of
volume we have used the weight in grams of the fixed cerebellum,
this having been corrected, if necessary, for change in specific
gravity. A constant fractional part of each cerebellum, an
equivalent unit area (HUA), has been secured by taking a line of
Purkinje cells as many millimeters long as the numerical value
obtained, by taking the square of the cube root of the cerebellar
weight in grams—this being from a section cut 25, thick and
corrected for shrinkage as already explained.
If, then, we let
N=the number of cells per EUA
L=the length of the line of cells as projected in millimeters
M=the magnification
C=the total number of cells counted
S:=the sum in mm. of the length and breadth of the block
as first measured
S.=the sum in mm. of the length and breadth of the section
on the slide
W =the weight of the cerebellum in grams after fixation then
Wy foekes = W3
n=(C+5)x(Z pak
240 ROBERT S. ELLIS
or
CM ey 7
N= == as W2
L x S. x 3
As an example we may substitute the values for Case 15035,
Area 3, left hemisphere (table 1).
N = xe) meas
2220 89.1
= 6.47 x 0.86 X 25
= 138
RESULTS OBTAINED FROM COUNTS
Neurohistologists who have made a study of the Purkinje
cells are familiar with the fact that these cells are not evenly
distributed, but tend to be more or less irregularly spaced, and
this seems to be particularly true for the human cerebellum. In
the bottom of sulci they are not normally so numerous as they
are at the summits of the folia. It would not be far wrong, in
fact, to say that they increase in frequency as one passes from
the bottom of the sulcus to the summit of the folium. This is
probably to be interpreted as incident to the phenomena of
growth. The bottom of a sulcus represents an arrest in growth;
the summit of the folium is the last to fill out and develop. It
is, then, not surprising that a greater number of Purkinje cells
should appear in the region characterized by the greatest amount
of growth change.
Considerable variation will be found in: the frequency of
Purkinje cells, even in adjacent folia. This is especially true of
defective brains, less so of normal ones. This makes it necessary
to count several hundred cells in order to get a satisfactory
average for any given area. Each value given in the tables in
this paper is based on an average of about 400 cells, and it is
believed that by taking a number so large the effects of acci-
dental selection have very largely been eliminated.
If perfectly satisfactory normal cerebella had been used as
the basis for the norms given in table 1, it seems certain that
QUANTITATIVE STUDY OF THE PURKINJE CELLS 241
all the values for the different areas would be somewhat higher
than those given there. It has seemed wiser, however, to use
the values actually observed rather than to attempt the more or
less dangerous expedient of guessing at the extent to which these
have fallen below the norm.
The number of Purkinje cells in equivalent unit areas (HUA)
of the two hemispheres of the four normal cerebella is shown in
table 1.
From table 1 it appears that the cells are. closer together in
the lobus anterior (area 1) and that they become progressively
less numerous in the anteroposterior direction, with the exception
that in the lobus biventer, area 6, the number is slightly greater
than in area.5. .
In the lobus biventer the cells are usually more numerous
in the posterior part. This is not shown in the table because
the two parts are considered together; the fact is of importance,
however, in relation to some of the changes found in subnormal
and senescent cerebella, for in our cases the anterior part of this
lobe has sufféred the greatest amount of degeneration.
TABLE 1
Number of Purkinje cells per HUA in normal cerebella. Areas as in fig. 1.
R=right hemisphere, L=left hemisphere
W.I.No.| RACE | yu ,| AREA 1 | arpa 2 | anea3 | arma d | area 5 | arma 6 | Toran
fee apeoit [2B (AE |p
os a0 Gg et ee oa Et
ee len Wee eae. idea uae
esa aS Pet ae wae | EF. | Blog| allan
AVIOLO SOx on ch te 162=8 | 185+2 | 135+10| 182+14) 125+6 | 127+5 Peta
Note. The numeral preceded by + equals one-half the average difference
between the right and the left hemispheres.
242 ROBERT S. ELLIS
Differences between right and left hemispheres. A glance at the
table shows that in most cases there is a considerable difference
between the values for the right and left hemispheres. In three
cases the values for the-right hemispheres are larger, while in
one case the value for the left is the larger. Unfortunately, it is
not known whether these cases were right or left handed, and it
is consequently impossible to show here a positive correlation
between a greater number of cells and superior unilateral skill.
As a suggestion, however, it is not improbable that we have here
three right-handed individuals and one left-handed. In any
case it may be noted that the differences between the right and
left hemispheres are greatest in areas 3 and 4, the supposed arm
and hand areas, and that they are least in area 2, the neck area.
On purely a priori grounds this is what is to be expected if there
is a correlation between number of cells and fineness of muscular
coordination. The greatest difference between the right and left —
hemispheres should be in the hand and arm areas, because it is -
in the two hands that there is normally the greatest difference in
motor skill. In the neck area there is obviously less reason for
expecting one side to be more developed than the other. How-
ever, further studies must be made here on cerebella from cases
with known histories before this question can be answered with
any assurance of correctness.
SUBNORMAL CASES
In order to compare the relative frequency of Purkinje cells
in the cerebella of subnormal individuals, the histories of which
have been given on page 232f, with normal ones, counts have
been made on areas 1, 3, 4, and 6 of the right hemisphere and
on areas 3 and 4 of the left hemisphere. These results are
presented in table 2.
It is obvious at a glance that in this group of mental defectives
the number of Purkinje cells is notably deficient. ‘The con-
stancy of the results and the wide difference between the values
for the normal and those for the defective cerebella as shown
by’ the fact that the latter are only 63 to 78 per cent of the
former make it impossible to assume that accident, chance, or.
ee a ee le
aa ae
QUANTITATIVE STUDY OF THE PURKINJE CELLS 243
TABLE 2
Number of Purkinje cells per HUA in subnormal cases. Designations as in
table 1
W.I. NO. Fae AREA 1] AREA 3|AREA4/AREA6| TOTALS
R OF: be 980:| 98 sel FORE
14880 i 102 | 100 ‘ 598
R Bee 126 122" 100K TT
se L 114 | 102 fdee
R 123 85 Baur lage sl
15145 i 95 | 105 630
R 508) e109.) 127
15250 { L 108 | 108 + \ 660
R 135 93 89 95 |\
52
15297 my 38 97 f 597
R 109 92 79 47
D)
15299 L 95 81 \ 503
R 85 | 100 94 98 |).
15310 * 77 73 f 527
R 93 80 95 67 |\
2
15320 { i 85 75 f 495
PNG OTA Ce etc A Pie Cs oe eal ss Me 105 96 94 97
Per cent of normal as in table 1......... 63 iil “al 78
personal equation have played any considerable part in de-
termining the differences observed. Moreover, the number of
cases is large enough to make it practically certain that typically
a decidedly smaller number of Purkinje cells is present in the
low grades of feeble-minded individuals. Here, then, we have
an anatomical deficiency as a basis of the observed deficiency
in motor codrdination. To this should be added the fact that
many of the cells counted were atrophied and showed con-
solidation and degeneration of the nucleus to such an extent
that they could hardly have been of any functional value. Many
244 ROBERT S. ELLIS
other cells, although not so far degenerated, showed clearly
the early stages of the process. With many of these very im-
portant elements of the neuromuscular mechanism lacking, and
with many more in a degenerating condition, it i is inevitable that
poor coérdination should result.
After the cell counts for the different areas were completed,
the clinical histories of the subnormal cases were consulted and
an effort was made to see how far the variations in the different
areas would give evidence either for or against the localization
theory of Bolk. Too much must not be expected from such a
comparison, because the case histories were not made with a
view to their being used in this manner and they are conse-
quently incomplete. It will be interesting, however, to consider
briefly the facts in each case.
No. 14880 shows as compared to the normal the lowest number .of
cells per EUA in the head area. The personal history shows that he
had a small head and poor speech. It is consequently not improbable
that the entire head musculature was under-developed and poorly
coérdinated. The record says he was right handed. The. cell count
shows a slightly greater number of cells in the left hemisphere. It will
be shown later in discussing senescence and paresis that in right-handed
individuals the right hemisphere suffers more loss than the left. The
case history indicated postnatal degeneration, and this, if true, would
account for the smaller number of cells in the right hemisphere.
No. 15144. The history shows a bullet-shaped head and very poor
speech. He was able, though, to thread needles and tie his shoes,
which shows that his hands at least were capable of difficult motor
coddinations. The cell counts show the head area very low, but the
arm and hand area nearly normal. This agrees well with the locali-
zation theory.
No. 15145. The history shows epilepsy—a degenerating disease—an
abscess on the right hip, and a failure in the grip test. The cell counts
show the arm and hand areas lower than the others, with the right side
lower than the left, this being due probably to degeneration.
No. 15250. The record shows poor speech, atrophy of the jaw, and
eye fatigue. The cell counts show the head area the lowest. The
grip test shows the left hand slightly stronger and the cell count for
area 3 agrees with this.
No. 15297. The record shows a mental age of eight, which, as far
as the cerebellum is concerned, would indicate a higher development
of the head area with respect to eye movement, speech, and facial
expression. The cell count for area 1 is the highest of the eight cases
listed. This agrees with theoretical expectations. The record for grip
QUANTITATIVE STUDY OF THE PURKINJE CELLS 245
shows the right hand stronger; the cell count for area 3 agrees with
this, but for area 4 is opposite to it.
No. 15299. This is a low-grade case, unable to talk or work; also he
was knock-kneed. ‘The cell counts for both hemispheres are very low,
and this is especially true in area 6, the leg area (?). A large part of
the anterior part of the lobus biventer in the right hemisphere was
completely atrophied. The grip test shows both hands about equally
weak and there is little difference between the cell counts for the two
hemispheres.
No. 15310. The record shows that he was left handed and that a
Wassermann test was positive. This probably accounts for the extreme
loss of cells in the left hemisphere through degeneration. The grip
tests shows the right hand stronger and the cell count shows more cells
in the right hemisphere.
No. 15320. The low cell counts here agree well with the low men-
tality and the general lack of motor power and coérdination. The
tumor in area 6 of the right hemisphere probably accounts for the loss
of ability to walk well after the age of eighteen. Also it is important
to note that many of the remaining Purkinje cells were in the process
of degenerating. General motor deficiency was consequently inevitable.
In view of the shia? meagre clinical data available, it was not
expected that a very high degree of correlation would be found
between the reported defects in motor codrdination and the
numerical deficiency in Purkinje cells. Furthermore, a state-
ment of the relative number of cells, if taken alone, is not neces-
sarily a complete indication of functional efficiency; for the cells,
although present, may be so far degenerated that they are
without functional value. It is consequently interesting to find
that in the cases considered there is a considerable difference,
even in cell number, between the normal and the subnormal
and that the losses in cells agree very well with Bolk’s locali-
zation theory.
The question naturally arises as to whether this deficiency in
cells is due to agenesis or to some toxin or other agency present
during intra-uterine life, or to postnatal disease or injury,
acting on the cells already formed.
To throw some light on this problem, counts were made on
area 3, the arm area, of both hemispheres of six low grade in-
fants. These may safely be said to be of subnormal ancestry;
also they can hardly be said to have had the most favorable
nutritional conditions during fetal life. How the number of
cells in these cases compares with the normal is shown in table 3.
246 ROBERT S. ELLIS
TABLE 3
Number of Purkinje cells per EUA in subnormal infants. Designations as in
table 1
AREA 3
Ww.tI. NO. RACE AGE
18? L
months
14428 B 1 120: 105
E22 W 5 126: 115
14250 W. 3350) 106: 120
H23 B 5 94: 103
14215 B 9 110: 102
14427 B 36 LOA ae
Average.:.... Mi Oracle cs, 08,3 fd, ce Gace TEES Rea eR 110
Normalan mene eos ot cd score. 2 Rita) Re ee i Pe aD 2 ote Isis
BRemcentiol normal as inevapler) 15 eee eeeeen ee a eens 82
The sections do not show that cells have degenerated and
dropped out, and yet in every case the number of cells is dis-
tinctly below normal.. The cause of the deficiency must then
have acted antenatum. Whether this is to be charged to
agenesis or to early. destruction is, however, not easily deter-
mined. ‘The writer’s opinion, based on the evidence at hand,
is that the cell deficiency is due primarily to agenesis. Beyond
this it is not possible to determine from the present study the
etiology of the conditions observed.
In some cases, however, mental and motor deficiency appears
to be due to causes which operate postnatum. This is shown in
case no. 15214. The cell counts as compared with the normal
are shown in table 4.
These values in table 4, the reader will note, are almost normal.
Yet the case is a distinctly subnormal one., The clinical history
shows that the ancestry is free from feeble-mindedness as far as
known; also it shows that the child was normal until nine months
of age. At the time he began to have spasms and to show
evidences of subnormality, due perhaps to the delayed effects
of congenital lues. The cerebellum is of normal weight, but the
—E———————
QUANTITATIVE STUDY OF THE PURKINJE CELLS 247
TABLE 4
Number of Purkinje cells per HUA in case No. 15214, an example of
postnatal arrest of development. Designations as in table 1
w.I. NO HEMISPHERE ARBA 1 AREA 3 AREA 4 AREA 6
R 159 124 106 143
pet { L 147 151
ATEN AC ee ger}. eve ayer 159 1385+12 128 +23 143
ISG aes ae eee 162=8 135+10 1382+14 127-5
cerebral hemispheres are much below weight, which further
indicates that the cause of the arrest took effect after birth.
An examination of the cells reveals on the anatomical side at
least a reason for the defect. Many cells are atrophied and are
so far degenerated that they cannot have any functional value.
Other cells are less degenerated and some are apparently normal,
but others have completely disintegrated and have been ab-
sorbed. This is clearly a case of postnatal degeneration, but
even so, it represents the exception rather than the rule among
subnormal individuals. The entire case history shows this.
In this as in the other cases, then, we find that in low-grade
mental defectives there is a distinct deficiency, either numerically
or cytologically, in a large percentage of the Purkinje cells.
SENESCENCE
It is a familiar fact that very old people usually suffer a con-
siderable loss of motor control. The results of counts made on
five cases of senescence of different ages are given in table 5 and
show how the cells drop out with increasing age.
The results of table 5 are shown graphically in chart 1. The
value for the normal is based on table 1 and the starting-point
for the drop in the curve is placed somewhat arbitrarily at the
age of forty years.
In connection with this curve it is interesting to note that it
is based on the cerebella of two men of superior mentality, on
one negro male autopsied in a general hospital, and on one white
male and one white female dying at extreme old age in hospitals
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, No.2
248 ROBERT S. ELLIS
TABLE 5
Number of Purkinje cells per HUA in senescent cerebella.
Designations as in table 1
HEMI-
W.I.NG.| RACE| SEX AGE | ueRE| BEA 1 AREA 3 AREA 4 AREA 6 Total
Noma 42? 162+8 { ee f Hig 2725
14464 | B. | M. | 62 { - it an Bs a } 667
16107 | W. | M. | 65 { x es ee 3 oe \ 591
14340 | w. | M. | 79 : ot a os a } 500
E27 | w.| F. | 94 x & x ei ee \ 462
14544 | w. | M. |100 - ae * = mi \ 403
for the insane. The uniformity of the curve is consequently
surprising. Much speculatidn might be based on such results
as these, but here it will suffice to call attention to two points:
first, the average loss in area 1, the head area, as compared with
the values from table 1, is relatively the greatest; second, in
areas 3 and 4, the right hemisphere suffers more than the left.
PARESIS
It was not originally intended to include any cases of paresis
in this study, but as the cerebellum of one case was prepared
before the cause of death was known, and as it is of interest
because of its similarity in cell losses to the other types of cases
presented, it seemed worth while to include it for purposes of
comparison. The results of the cell counts for this case are
presented in table 6.
Here again we find the head area very low and the right
hemisphere much lower than the left. This was a right-handed
man and the cell losses have been greatest on the right side.
QUANTITATIVE STUDY OF THE PURKINJE CELLS 249
The figures for cell losses do not of course fully indicate the
loss in functional efficiency of such cerebella. Many of the
cells that remain are in the process of disintegration and there-
fore many of those included in the cell count probably have
little or no capacity for functioning.
NUMBER OF PURKINJE CELLS
“iS ae RE RRA
ao 1 a a i 2
mmiamisietstet ele yaa) sia
SEA SESE Saisie arias eb eee e eet
DSRS RECGGS OS Oo Cece ee
SESS 5 SG00S 00s aso U Ae SS
: ERO GRS BEG Ioo. s
HUSSRSSR TOE SE ss ios = tees
SCGSERNS DOTS C IIe ee
BSE ou hacia eae betta boon
BA ae ee
) G0 Ee es Noe
SUT UteRResoec cose eee
ja a Dm
CESAR SSS Ie iN oe
2S aE SIT a
{oe (62S RSS oo Oe ese aes
- SCRE n SNE) oO Sdn
ZOBUSEESEES a3 /c5 oye
ens Toft halal de UL)
a a a cy ead Ve
O08 re a I Ls
LO oor comes So. FOS e100
Chart 1 Showing the decrease in the number of Purkinje cells with advanciny
age. Based on the totals of the four areas given in table 5. The dots show the
relation of the six observed values to the graph as drawn.
THE VERMIS
Sections of the vermis were prepared in all cases and a few
preliminary counts were made to determine whether or not it
would be profitable to make counts on all cases. The results for
the tuber vermis in five cerebella—two normal, one senescent,
and two subnormal—are given in table 7, and for comparison,
the averages of area 4 of both hemispheres are given also.
No significant. differences in the relations found in the' two
sections appear from this table. The sections of the vermis
from the other cases were examined under the microscope with-
out counting, and as it seemed clear that they would not differ
materially from what had been found in the hemispheres, no
further counts were made.
250 , ROBERT Ss. ELLIS
TABLE: 6
Number of Purkinje cells per HUA in a case of paresis. Designations as in
table 1
W.I. NO. AGE AREA | AREA 3 AREA 4 AREA 6
‘ R 107 99 68 90
area ee { ie 133 118
Normalan. cte ec ee ir cies: 162+8 135+10 132+14 127+5
COMPARISON WITH THE CEREBRAL CORTEX
The writer has had occasion to make cell counts on a number
of areas in the cerebral cortex, and has found, in agreement with
most others who have conducted investigations along this line,
that there is ordinarily a distinct deficiency in cells in cases of
amentia. Among those authors who may be mentioned here
are, for example, Hammerberg (’95), Roncoroni (’05), Tredgold
(03), and Bolton (’08).
The motor area of the cortex in several cases of paresis and
also of senescence has been carefully examined, and it is interest-
ing to note that in all of these there has been found a deficiency
in Betz cells. How general-this particular loss is, the writer
does not know; it is, however, important to recognize that
disintegration takes place in the cerebrum in a manner similar
to that found in the cerebellum. It is therefore not probable
that the motor deficiencies observed in the cases studied have
been due solely to deficiencies in the cerebellum. Probably in
the majority of cases a defective cerebellum is accompanied by a
defective cerebrum, and vice versa.
It would be beyond the scope of the salar paper to attempt
a more detailed statement of the character of the cell changes in
these atypical brains; enough has been done, however, if we have
succeeded in pefablahine the numerical differences found in these
different types of cases.
QUANTITATIVE STUDY OF THE PURKINJE CELLS O51
TABLE 7
Number of Purkinje cells per EUA in the tuber vermis, compared with the
average for area 4
NORMAL SENESCENT SUBNORMAL
W.I. No. | W.1I. No. | W.I. No. | W.I. No. | W.I. No.
14485 15073 16107 15310 15320
PT DCI VETS) a1. ios eee ee 157 137 112 83 91
PATO A Beat ssi sidc natheteac tee ernee 138 119 87 84 85
SUMMARY AND CONCLUSION
The main purpose of this paper was to show the numerical
differences in Purkinje cells in normal, subnormal, and senescent
cerebella. |
From the data submitted it is evident that in cases of extreme
mental defect due to agenesis or to the early action of toxins
during intra-uterine life, there is an evident deficiency in the
number of these cells.
Similar reductions in the number of cells due to various causes
are found in senescence (and paresis).
In the subnormal cerebella the evidence indicates that the
normal number of cells has never been present in a developed
form. In the senescent (and paretic) cases, however, the small
number is due to disintegration.
The anterior lobe of the cerebellum shows the greatest de-
ficiency in cells in both the subnormal and’ senescent cerebella.
The biventral lobe shows the greatest variation in both types
of cases. In some cerebella it shows the greatest loss of cells
and in others the least loss.
The differences between the two hemispheres in respect of
cell number average less in subnormal cerebella than in normal
ones. This probably has a relation to the differences in the
degree of unilateral dexterity found in normal and subnormal
individuals, i.e., normal people are usually more distinctly right
(or left) handed than are the subnormal, who tend to be more
ambidexterous.
ae ROBERT §S. ELLIS
The deficiency in cell number affords in large measure an
explanation of the motor defects found in subnormal individuals.
It shows, furthermore, that in idiocy and in imbecility we may
expect to find the whole brain defective rather than the frontal
lobes only, while the higher grade of defectives (morons) proban
show very slight deviations from the normal.
Further studies in this field on better material with better-
known clinical histories in which is included a study of blood-
vessels, nerve fibers, neuroglia, and a cytological study of all
the important types of cells are necessary to bring out the more
detailed differences between these types of cerebella.
For a review of the literature on the clinico-pathological study
of the cerebellum with a detailed report of a single case see
Archambault (’18).
LITERATURE CITED
ARCHAMBAULT, LASautite 1918. Parenchymatous atrophy of the cerebellum.
Jour. of Nerv. and Ment. Disease, vol. 48. ~ >
BERLINER, K. 1905 Beitrige zur Histologie und Entwickelungsgeschichte des
Kleinhirns. Archiv. f. mikr. Anat., Bd. 66. :
Botx, L. 1905-07 Das Cerebellum der Siugetiere. Petrus Camper, Neder-
landsche Bijdragen tot de Anatomie, Bd. 3 u. 4.
Botton, J.S. 1903 The histological basis of amentia and dementia. Arch. of
Neurology, vol. 2.
1910-11 A contribution to the localization of cerebral function based
on the clinico-pathological study of mental disease. Brain, vol. 33.
Gopparp, H. H. 1914 Feeble-mindedness. New York.
HAMMARBERG, ©. 1895 ‘tudieniiber Klinik und Pathologie der Idiotie. (Trans.
from Swedish into German by W. Berger and pub. by S. £. Henschen).
Upsala.
Kine, Heten D. 1910 The effect of fixatives on rats’ brains. Anat. Rec.,
vol. 4. ‘
Roncoront, L. 1905 Lo sviluppo degli strati molecolari del cervello, ete.
Arch. di Psichiat., vol. 26.
RynBerk, G. Van 1907-08 Die neuren Beitrige zur Anatomie und Physiologie
des Kleinhirns der Siiuger. Folia Neuro-biologica, Bd. 2
1912 Weitere Beitrige zum Localizations-problem im Kleinhirn.
Folia Neurobiologica, Bd. 6, Supplement. 1
TrepcoLty, A. F. 1903 Amentia. Arch. of Neurology, vol. 2
1914 Mental deficiency. 2nd Ed. London.
*
ee er
AUTHOR’S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, MARCH 17
THE EFFECT OF OVER-ACTIVITY ON THE MORPHO-
LOGICAL STRUCTURE OF THE SYNAPSE
KIYOYASU MARUI
Sendai, Japan
Neurologial Laboratory of the Henry Phipps Psychiatiic Clinic, Johns Hopkins
Hospital, Baltimore, Md.
FOURTEEN FIGURES
INTRODUCTION
Investigations on the histological manifestations of the nerve
cell in fatigue have long been familiar to us. Numerous authors
have brought contributions to this important and interesting
subject. If we take for granted, as Sjoeval declared, that the al-
teration of nerve cell in tetanus is to be regarded as the effect of
activity, the number of references becomes even more abundant.
It would suffice to point to those works of Hodge (14, 15, 16),
Vas (33), Mann (23), Lugaro (29), Sjoeval (82), Dolley (17, 18,
19), and many others. There is almost a complete agreement on
the point, that over-activity causes appreciable histological
alterations in the nerve cell. Kocher (19), who studied the same
subject in our laboratory, could not, strangely enough, find any
qualitative and quantitative changes in the histological characters
between the fatigued and resting nerve cells. Recently Saito
undertook an exploration in the same line in this clinic, but the
results are not published yet.
Compared with the abundant researches on the neurocyto-
logical manifestations, there has been no attempt made to inves-
tigate the histological alteration of the synapse in fatigue, as far
as I know. Not only in regard to over-activity, but also in other
pathological conditions, the histopathological changes of the
synapse have been the topic of very few investigations. And no
wonder, for the structure of the synapse had not yet been con-
clusively demonstrated even in the normal condition, despite
253
254 KIYOYASU MARUI
many investigations by different authors (Golgi, Bethe, Held,
Bielschowsky, Auerbach, and many others). The lack of our
knowledge concerning the pathological manifestations in the
synapse and especially of experimental pathological data seems to
be dependent on the difficulty of technique on the one hand and
inaccessibility of adequate material on the other. In my recent
publication (24), I described more minute structures of the
synapse of the Mauthner cell of teleosts (bony fish). Per-
haps I have not made the structure of the synapse clear beyond
discussion; however, my results have been sufficiently definite
for me to attempt experimental work on the pathological con-
dition, especially in fatigue. Careful and thorough investigation
on this topic led me to very interesting results, which throw
light on the histological condition of the synapse in functional
activity, although it was forced far beyond the physiological
limit. In the present paper I will describe mainly the manifesta-
tions in the synapse as it was not my purpose to investigate the
neurocytological alterations of the Mauthner cell itself. But
wherever notable manifestations of the cell body appear, they
will be described as well as the findings of the synapse. ‘To Dr.
Adolf Meyer I beg to express my high appreciation here for his
help and kind suggestions in this study.
MATERIAL AND METHODS OF STUDY
In the present investigation Ameiurus nebulosus was the ma-
terial of experimentation; Carassius auratus, which offered me
many interesting results in my previous work, proved to be un-
favorable material, owing to the fact that it is weak and often
dies before it comes to the utmost exhaustion. It was always at-
tempted in this experiment to work with strict control of the nor-
mal structure of the Mauthner cell as well as its synapse, and
care was always taken to avoid the formation of artefact as much
as possible.
As a resting non-fatigued control, which was provided in each
experiment, I used a fish of about the same size which was kept
in a resting condition during the experiment of the other fish.
The body length of the fish used varied from 5 to 7 inches.
EFFECT OF OVER-ACTIVITY ON SYNAPSE 250
The experiment consisted in forced activity of the fish, carried
to the most advanced stage of fatigue; the experimental fish was
placed in a jar 7.5 inches in diameter and about 8.5 inches in
depth and then the water in the jar was continually stirred by
running water, which came with high pressure from a faucet. The
fish swims in the stirred water, trying all the time to hold its
equilibrium. According to Bartelmez (3), the Mauthner cell
participates in the equilibratory reflex; so it was supposed that
in this experiment the Mauthner cell would be forced into con-
tinuous activity. The duration of the experiment varied re-
markably in each fish (from 24 to 98 hours). At first: the fish
swims actively and is able to hold its equilibrium very well, but
gradually it gets tired and in the final stage of the experiment it
is deprived of the ability of balancing, so that it is moved pas-
sively tumbling around in the stirring water. By this time if we
stop the running water for a moment, the fish would lie on its back
or on one side, showing no attempt to maintain its upright pos-
ture. As the sign of utmost exhaustion, I chose a test, which con-
sisted in holding the fish upside down by its tail in the water as
well as in the air. In the most advanced stage of exhaustion the
fish did not flap at all even in this test.
The fish was then decapitated and bled, the brain was quickly
but carefully dissected out and fixed in 10 per cent formalin, for-
mol-Zenker fluid, and alcohol (95 per cent), respectively. The
resting control fish was killed at the same time and the brain was
manipulated in .quite the same way as the fatigued brain. It
must be emphasized also that the material was always fixed fresh
and that material from fish which died was not examined.
From the material thus obtained, the following preparations
were made with the same technique as was described in my recent
publication:
1. Thionin-eosin preparation (1 series each of normal and
fatigued brains).
2. Toluidin-blue preparation (5 series each of normal and
fatigued brains).
3. Heidenhain preparation (6 series each of normal and
fatigued brains).
256 KIYOYASU MARUI
4. Levaditi preparation (14 series each of normal and fa-
tigued brains).
5. Bielschowsky preparation (6 series each of normal and
fatigued brains).
6. Cajal preparation (5 series each of normal and fatigued
brains).
Besides these I used the following stains in the present work:
7. Scharlach stain (5 series each of normal and fatigued brain).
The fish brain was fixed in 10 per cent formalin twenty-four
hours, rinsed with water, cut at 154 with the freezing micro-
tome. The sections were stained in saturated solution of Schar-
lach R in 70 per cent alcohol, rinsed with distilled water, counter-
stained with diluted Ehrlich’s hematoxylin solution, washed
again, and mounted in glycerin. As there is only one pair of
Mauthner cells in a fish brain, it was rather hard to get the sec-
tions in which the Mauthner cell is found.
8. Mallory preparation (5 series each of normal and fatigued
brain). In this preparation the brain was fixed first in 10 per
cent formalin twenty-four hours and then placed in formol-Zenker
fluid twenty-four hours. Paraffin sections 5 to 8 thick were
stained with a diluted solution of Mallory’s hematoxylin.
9. S.-fuchsin-light green stain (5 series each of normal and
fatigued brains). Brains were first fixed in 10 per cent formalin
twenty-four hours and then placed for eight days in the chromic
acid acetic acid mixture. Paraffin sections of 54 were treated as
follows: 1) Remove paraffin from the sections and pass to 96
per cent alcohol. 2) Place sections for one hour in a saturated
aqueous solution of S-fuchsin in the incubator at 58°C. 3) Wash
twice with distilled water, till no more stain comes out of the
sections. 4) Dip the slides in motion 10 to 20 seconds in the
following solution:
Saturated alcoholic solution of picric acid 30
Aqua destillata 50
5) Rinse carefully twice in water. 6) Place the slides in a sat-
urated aqueous solution of light green twenty minutes. The sec-
tions were then washed with water, dehydrated very quickly and
passed into xylol.
EFFECT OF OVER-ACTIVITY ON SYNAPSE 257
In all 122 series of normal and fatigued brains form the basis of
the present article.
ON THE INTERNAL MORPHOLOGY OF THE MAUTHNER CELL AND ON
THE MINUTE STRUCTURE OF THE SYNAPSE
Under this heading I will make a few notes on the internal mor-
phology of the Mauthner cell (figs. 1 and 2) and on the structure
of the synapse, which are necessary as the foundation of the fol-
lowing statement and were not yet described in my recent pub-
lication. Nissl bodies are distributed evenly through the cell
body and bases of the dendrites, leaving free only the axone hil-
lock. They are relatively small as compared with those in the
motor cells and very numerous, as Bartelmez (3) stated. The
Nissl substance is found in the shape of variably long striae and
is arranged generally parallel to the contour of the cell body and
in part also to the surface of the nucleus. The remaining stain-
able substance is irregularly scattered and is more or less short;
some of it is spheroidal. The spindles were found especially on
the surface of the cell and in the larger dendrites. The so-called
nuclear caps did not come to my observation. The axone hillock
is entirely free from stainable substance and marked off by a tol-
erably sharp-curved plane from the granular protoplasm of the
cell body and shows at its margin a layer of especially fine gran-
ules. The nucleus of the Mauthner cell differs in no essential
from the typical nuclear structure of the nerve cell.
As was precisely stated in my recent communication, the
synapse of the Mauthner cell is penetrated by the Golgi network,
which is formed by the arborization and the reunion of the deli-
cate processes of the neuroglia cells in and about the ‘axone cap.’
It was also accepted that the Golgi network is to be attributed
to that category of the neuroglia tissue, which Held (24) termed
as the reticular glia tissue formed by the somewhat modified plas-
ma of the glia cell. Furthermore, I paid special attention in that
paper to the histological structure of the nervous elements of that
synapse and the relation of the latter to the Golgi net. We may
therefore pass directly to the description of the finer structure
of the glia cells themselves and the condition of the capillaries in
this synapse.
uu 4,
+, of
EA
EFFECT OF OVER-ACTIVITY ON SYNAPSE 259
The neuroglia nucleus is surrounded by a variously wide bor-
der of protoplasm, which latter sends protoplasmatie processes
in different directions. The process itself ramifies and becomes
more and more delicate, until it passes into the beams of the Golgi
net or becomes attached to the wall of a capillary. This picture
could very clearly be observed in the Levaditi preparations,
Heidenhain preparations of formalin material, Mallory and acid-
fuchsin-light green preparations. In the Heidenhain and thionin-
eosin preparations of formol-Zenker material a similar condition
was demonstrated, although it was not so clear as in the above-
_ mentioned preparations.
The protoplasm of glia cells was brought out favorably in the
thionin-eosin preparations, Heidenhain preparations, Mallory and
acid-fuchsin-light green preparations. There are two different
types of neuroglia cells; in one type the protoplasm is very scanty,
so that the glia cell shows a small round cell body, while in another
type we find a large mass of protoplasm around the nucleus. The
neuroglia nuclei are sometimes connected with each other not by
means of the Golgi-net substance, but by a variously broad mass
of granular protoplasm. I should interpret this as the result of
amitosis which is observed now and thenin thesynapse. It must
be positively emphasized here, however, that all the glia cells of
the synapse belong to one reticulum and that there is no cell indi-
vidual among them in normal brains. The structure of the glia
nucleus hardly calls for a description except that sometimes it
shows evidences of amitosis, as Bartelmez (3) also stated. Cap-
illaries are found here and there in and about the synapse of the
Mauthner cell. They do not offer anything particular in their
The figures 1 to 8 are the unretouched photomicrographs taken from different
preparations of both the control and the fatigued Ameiurus brains. In figures
2 and 5 an apochromatic Zeiss ocular no. 4 and Zeiss objective D were used; others
were taken with the same ocular and a Zeiss immersion js. The length of bel-
lows was 60 em. in all the photomicrographs. The figures 6 to 14 were drawn from
different preparations of fatigued Ameiurus brains, using the Abbe camera lucida.
(Zeiss apochromatic ocular no. 4, Zeiss oil-immersion ;'z, tube length 20 em.)
Fig: 1 Toluidin-blue preparation (alcohol material) (control fish).
Fig. 2. Thionin-eosin preparation (formol-Zenker material) (control fish).
Fig. 3 Toluidin-blue preparation (alcohol material) (fatigued).
260 KIYOYASU MARUI
structure and consist of endothelium and adventitia with very
few nuclei. As already remarked, the adventitia of the capillary
is connected with the bases of the reticular beams; but I was not
able to distinguish the perivascular limiting membrane of Held as
a membrane separated from the adventitia, so that I could not
find the so-called perivascular lymph space between them in
normal fish brains.
THE HISTOLOGICAL MANIFESTATIONS OF THE MAUTHNER CELL
IN FATIGUE
To recapitulate and discuss the results of other authors here
would go beyond the purpose of the present work; my remarks
will be restricted to my main results of investigation.
The cell body of fatigued cells was found either in the state of
turgescence (figs. 3 and 4) or of shrinkage (fig. 5); in the former
case the cell border was convex between the dendrites and the den-
drites appeared shorter, whereas in the latter the cell border was
concave and the dendrites looked longer. I agree with the opin-
ion of Vas (33), Mann (23), Lugaro (21), Pugnat (26) and Holm-
gren (17), that the enlargement.of the cell body is to be considered
as the manifestation of activity and the shrinkage as that of ex-
haustion. In this way the results of Hodge (14, 15, 16), which de-
viate from those of others, might well be interpreted. Dolley (7,
9) described the fluctuations of the size of the cell body in the
course of activity, but I. have a little doubt about his statement
that in later stages the absolute size of the cell body increases
steadily to the end. The alteration of the Nissl substance mani-
fested itself in more or less advanced stages of chromatolysis, as
was described by Vas (33), Lambert (20), Mann (23), Lugaro (21,
22),and many others, and thereby the cytoplasm was stained vari-
ously deeply (figs. 3, 5). The Nissl bodies were found in a state
of fragmentation, shortening, and irregular distribution; in the
advanced stage of chromatolysis they were reduced to fine gran-
ules or even to a homogeneous substance. Sometimes I observed
the central beginning of the chromatolysis, as was stated by Vas
(33), Mann (23), and, in tetanus, by Sjoeval (32).
EFFECT OF OVER-ACTIVITY ON SYNAPSE 261
nt
s
-
t
Ls tle
tre AY, A
n>
os Sas Se Aaa Ait pte <
Mere Pes televise 3 QOS ARE Rae eK
Fig. 4 Toluidin-blue preparation (alcohol material) (fatigued).
Vig. 5 Thionin-eosin preparation (formol-Zenker material) (fatigued).
262 KIYOYASU MARUI
The nucleus was sometimes located in the center of the cell body,
but in many cases it was situated more or less eccentrically, as was
described by many others—Vas (33), Lambert (20), Sjoeval (32),
Holmgren (17). Morphologically, the nucleus showed now and
then no particular change, but in other cases it was found either
swollen or shrunken. The swollen nucleus appeared large and
round and showed a smooth nuclear membrane, whereas the
shrunken nucleus showed irregular shape and a crenated mem-
brane. Vas (33) observed always an enlargement of the nucleus,
while Hodge (14, 15, 15, 16) found regularly the shrinkage of the
latter. Lugaro (21), Mann (23), Pugnat (26), and Holmgren (17)
declared on the basis of their study that activity causes turges-
cence followed by shrinkage in exhaustion. Sjoeval (32) denied
the pathological significance of this finding for the reason that he
observed those nuclei also in the cell which showed no change in
particular. I think those manifestations of the nucleus are to be
attributed to different stages of activity; Dolley (9) also declared
that the nucleus shows fluctuations of size in the course of
activity.
The nucleolus was found mostly eccentric in the nucleus, al-
though this was the case also in some resting nerve cells. It was
sometimes observed swollen (fig. 4); in other cases it showed an
irregular shape, oblong or angular (fig. 5). Mann (23), Lugaro
(21), Luxemburg (22) observed the enlargement of the nucleolus
during activity; it disappears later in exhaustion (Mann (23),
Luxemburg (22) ). Goldscheider and Flatau (12, 13), and Sjoeval
(32) confirmed this in tetanus. Mathes (25) observed the nu-
cleolus of angular shape as I did; the objection, that it might be
the deceptive appearance caused by the granules above or be-
neath the nucleolus, could not come in question in my cases.
As a most peculiar manifestation of the nucleus figure 5 was
presented; the nucleus is very large compared with the size of the
cell body, the nuclear membrane is marked sharply and the nucle-
olus itself is small and shows an ellipsoid shape. Besides that
there is a deeply blue-stained substance of triangular shape and
the remaining space of the nucleus is filled with compact acido-
phile substance. The nucleus shown in figure 6 might come under
EFFECT OF OVER-ACTIVITY ON SYNAPSE 263
Fe
at 4
7
Figs. 6 and 7 Levaditi preparation (fatigued).
264 KIYOYASU MARUI
the same category of nuclear change; we observe here besides the
nucleolus a rod-shaped substance which shows a staining reac-
tion similar to that of the nucleolus, although I am not sure
about this, as it is from a Levaditi preparation. As far as I know,
no such picture of the nucleus has been described before; figure 8
of Dolley’s (7) publication demonstrates a cell, the nucleus of
which shows a deeply stained spheroidal substance besides the nu-
cleolus, but Dolley did not mention anything about that in the
text or in the description of the plate. Whether this kind of mani-
festation of the nucleus is to be regarded as the alteration of double
nucleolus, which latter is met not infrequently, or can be attrib-
uted to a special appearance of the nucleus in fatigue, I can-
not tell.
Holmgren (17), Sjoeval (32), and others found the stainable
substance massed about the nuclear membrane, and forming either
an irregular or a complete ring. I also observed the same phenom-
enon in many cases (fig. 3); the nuclear membrane was out-
lined by a delicate blue-stained line or a large mass of stain-
able substance in a different portion of its circumference, giving
the picture of a half-moon. Sjoeval and Holmgren interpreted
this as a restitution phenomenon of the tigroid substance; Dolley
(7) also regarded it as a sign of greater nuclear activity. I agree
with the opinion of these authors. Holmgren (17) observed besides,
that the nucleolus and the nuclear granulation emigrate from the
nucleus into the cell body. The emigration of the nucleolus never
came to my observation; the case, however, from which figure 4
was reproduced may indicate the emigration of stainable sub-
stance from the nucleus into the cell body. The nucleus as well
as the cell body is swollen in this case, and the nucleolus is also ex-
tremely swollen, and we find many blue-stained granules going
out of the nucleus into the cell protoplasm. On the other hand,
I found also the accumulation of the acidophile substance in the
nucleus (fig. 5). On the basis of these findings I should agree
also with Holmgren, who came to a conclusion that a mutual inter-
change of substance takes place between nucleus and cell proto-
plasm in activity. On the ground of Richard Hertwig’s doctrine
of nucleus-plasm ratio, Dolley (6, 7, 8, 9) measured the size of cell
EFFECT OF OVER-ACTIVITY ON SYNAPSE 265
body and nucleus of nerve cells in activity and divided the cells.
into many stages of alteration. As I did not undertake the meas-
urement of the cell body and nucleus, I will not go further into
details of Dolley’s work; but his method of division is, as he him-
self admitted, an arbitrary one, and I found many cells, which can-
not be assigned to any of his stages. Furthermore, I am afraid
that in his interpretation of things, facts are linked with hypo-
thetical considerations which are not directly observable. I will
be satisfied in the present work, if I can make sure that the
Mauthner cell, the synapse of which is the material of this study,
manifests appreciable changes in fatigue.
THE HISTOLOGICAL MANIFESTATIONS OF THE SYNAPSE IN FATIGUE
In these experiments which consist in forced activity, although
it goes far beyond the physiological limit, attention was direct-
ed from the first, not only to the nervous constituents of the
synapse directly, but also and especially to the manifestations in
the glia tissue, which shows the changes of the functioning nerve
tissue in an indirect way. It was hoped that through the study
of the histological manifestations in fatigue some light would be
thrown upon the problem, concerning the function of the neurog-
lia cells in the state of physiological activity. I shall first de-
scribe the findings in the synapse of the Mauthner cell in fatigue
and later go over to the consideration of the significance of the
manifestations.
A. Manifestations in the Pericellular Reticular Structure of the
Synapse
The gha reticulum of the axone cap and of the cell surface pre-
sented itself in fatigue in a more or less advanced stage of devia-
tion from its normal configuration. In the Levaditi preparations,
which bring out the net figure very clearly and sharply in a dark
brown or black color, the alteration of the net configuration is
most distinctly noticeable. Figures 3 and 4 of my recent pub-
lication (24) demonstrate the normal condition of the glia reticu-
lum in the Levaditi preparation. The Golgi network, which is
visible on the cell surface (3) as well as in the axone cap (4),
266 KIYOYASU MARUI
stands out sharply, and the net beams are demonstrated in rigid
dark lines.
In many eases of fatigue this net configuration appears more or
less irregular and less sharply marked. The net beams present
themselves swollen and thick here and there; in other places they
are thinner than usual and in some places even broken up. The
substance of the net beams looks loose and less compact and in the
extreme state of decay the net beams are reduced to variably
large amorphous corpuscles and look not unlike silver precipitates
distributed irregularly in the synapse. 3
Figure 7 was reproduced from the Levaditi preparation of
fatigue case; it shows the surface section of a slightly shrunken
Mauthner cell and the cell surface is covered partially with the
Golgi network. The reticular structure of the synapse stands in
this case in a more advanced state of alteration. The spheroidal
or star-shaped structures, which correspond to the nervous ter-
minal feet, are demonstrated very distinctly in the nodal points
of the Golgi network (especially clear in the lower part of the fig-
ure). Now, the net beams connecting these terminal feet with
each other are in part marked tolerably sharply, but most of them
are swollen and are not clean cut. Some others look, on the con-
trary, very thin and loose or even broken up. Here and there we
find the terminal feet, which lie isolated on the cell surface, as a
consequence of the breaking up of the radiating net beams. The
reticulum of the axone cap (upper part of the figure) is in a similar
state of alteration; the net beams are extremely swollen and
loosened and appear thick. The reticular figure is partly well
preserved, although we find here and there the decay of the net
beams.
Figure 8 was produced from another case prepared by means
of Levaditi’s method; the Golgi network in the axone cap as
well as on the cell surface shows essentially similar changes.
The net figure here and there is preserved tolerably well, but the
meshes are irregular as compared with those in the resting condi-
tion. In other parts of the synapse the net beams show a more or
less advanced state of alteration and some beams are really re-
duced to amorphous black fragments.
EFFECT OF OVER-ACTIVITY ON SYNAPSE 267
The above described findings were repeated in a different inten-
sity In many cases of fatigue by means of Levaditi’s method,
although there were several cases with negative findings in the
synapse. It must be emphasized here that in the resting control
animal the pericellular reticular structure was always brought out
clearly and sharply. Of the manifestations of the glia cells in the
synapse I shall give a more precise description later. As espe-
Fig. 8 Levaditi preparation (fatigued).
cially worthy of note here, I want to discuss the relation between
the glia cells and the reticular glia structure. The protoplasm
and the protoplasmatic processes of the neuroglia cells increase
and swell in mass, as will be mentioned below. Some processes
show their relation to the reticulum even in their swollen condi-
tion, but some of the processes are no longer connected with the
glia reticulum. Some of them fall to pieces so that we find pro-
toplesm masses of different shape and size around the glia cells.
268 KIYOYASU MARUI
Some glia cells even lie freely in and about the synapse; to this I
shall come back again.
In the thionin-eosin preparations of the formol-Zenker material
it is very difficult to find such delicate alterations of the reticular
structure. In a slight alteration it is almost impossible to distin-
guish the difference between the resting condition and the fatigue
case, so far as the net figure is concerned. In the most advanced
stage of alteration, however we are able to find similar manifesta-
tions as those described in Levaditi preparations, although it is
not so easy to see as in these preparations. Figures 2 and 5 were
reproduced from thionin-eosin preparations. In figure 2, which
demonstrates the resting condition, we find regularly arranged
dark points around the cell, which evidently represent the net
beams of the pericellular reticulum. In the fatigued cell (fig. 5)
we find around the cell body a number of similar dark points, which
are, however, scattered without any order at the cell periphery.
The microscopic observation revealed clearly a change of peri-
cellular network similar to that described in Levaditi prepara-
tions; the net beams were in part swollen and others broken up.
The findings of the glia cells I shall describe later. In the Heiden-
hain preparations and other preparations a similar manifestation of
the reticular structure was observable, although it is not so clearly
demonstrable as in the Levaditi preparations so that we ean find
the alteration only in a far advanced state.
B. Manifestations of the neuroglia cells in the synapse
Among the neuroglia cells of the synapse of the Mauthner cell
I found in many cases Alzheimer’s amoeboid glia cells (figs. 9, 10,
11, 12, and 14). These amoeboid glia cells, as characterized by
Alzheimer (1), show a large protoplasmatic cell body with a spe-
cial morphological structure and small dark nuclei, and in typical
Figs. 9, 10, and 14 Thionin-eosin preparation (fatigued).
Fig. 11 Acid-fuchsin-light green preparation (fatigued).
Fig. 12 Mallory preparation (fatigued).
Fig. 13 Scharlach R stain (fatigued).
269
270 KIYOYASU MARUI
forms they have striking resemblance to an amoeba. Besides the
typical forms I found many others which do not resemble amoeba.
Before I pass to the description of the amoeboid glia cells, I will
pay some attention to the manifestations of glia cells which go
hand in hand with the appearance of the former. Although not
so numerous, we find glia cells in every stage of regressive and pro-
eressive change (figs. 9, 10, and 14). The regressive nuclei
appear sometimes extremely swollen and pele and at other times
they show a zigzag shape and deep stain. The homogeneous
stain of the nucleus and protoplasm, the breaking up of the nu-
cleus into spherules or small masses are other histological prop-
erties of the regressive nuclei. Furthermore, I observed in the
same sections production of youne amoeboid glia cell; karyokinesis
of the glia nucleus was found now and then, and amitosis of the
nucleus, observable occasionally in the physiological condition,
seems to appear oftener in fatigue preparations (fig. 6).
In young amoeboid glia cells the shape of the cell body is
simple and we find sharply marked protoplasm around the nu-
cleus. In older cells, however, the protoplasm grows larger and
sends processes of irregular shape in different directions. The
process itself has at first a simple shape, but later it shows a more
or less complicated shape; sometimes I observed that the gha
cells send the processes to nerve fibers or capillaries, holding the
latter between their ramifications. Generally speaking, the
amoeboid glia cells were found relatively more numerous near
the blood-vessels than in the other parts of the synapse. In
young amoeboid glia cells the protoplasm is first quite homogene-
ous, but in the further course of life many kinds of manifestation
become noticeable in the cell body.
In the Mallory preparations (fig. 12) the protoplasm of large
amoeboid glia cells contains variously large vacuoles; the size of
the vacuoles varies considerably, but generally speaking they are
not very large in my preparations. The number of the vacuoles
is also variable with the s ze and age of the cell. The content of
the vacuoles is quite clear in my preparations; I assume that
these vacuoles are lipoid cysts, the content of which was extracted
in the process of embedding. In my thionin-eosin preparations.
EFFECT OF OVER-ACTIVITY ON SYNAPSE 271
(fig. 10) and fuchsin-light green preparations I could also find
those vacuoles. Besides these I observed in the Mallory prepa-
rations dark violet or blue-stained granules in the cell protoplasm.
There is no doubt that these granules are identical with the
methyl-blue granules of Alzheimer (1). This kind of granules
varies considerab y in size, but in any one cell they are in general
of similar size as Alzheimer described. With the production of
this kind of granule the loosening of the protoplasm structure of
the amoeboid glia cells takes place. In the acid-fuchsin-light
ereen preparations, in which the methyl-blue granules cannot be
brought out, the cell body shows a granular or bubble-like ap-
pearance in this stage. At the same time marked changes be-
come noticeable in some nuclei; they are stained either homo-
geneously deeply or remarkably pale. Even the neuroglia cells
in the synapse, which are not in possession of the proper attributes
of an amoeboid glia cell, but have long narrow processes instead
of a large protoplasm mass, sometimes display alterations; then
the processes look as though they were dissolved into granules,
which show the same staining reaction as the methyl-blue
granules.
In the acid-fuchsin-light green preparations (fig. 11) the amoe-
boid glia ‘cells were brought out very clearly; in the evenly.
green-stained cell protoplasm the Alzheimer fuchsinophile gran-
ules appeared as large red spherules. The number of these gran-
ules varied in different cells; some large and old cells have gran-
ules scattered through the whole protoplasm. The sizes of the
granules are almost equal and the considerable size distinguishes
these granules from the fine fuchsinophile granules, which are only
occasionally observable in the normal glia cells. As already re-
marked, I found also in this preparation vacuoles of different size
in the cell body of the amoeboid glia cells. The contents of these
vacuoles in my preparations were always clear; the large lipoid
cysts with yellow substance described by Alzheimer (1) did not
come to my observation. The Alzheimer light green granules
were not observed either in my preparations.
In my thionin-eosin preparations (fig. 10) of formol-Zenker ma-
terial I found also a number of typical as well as atypical amoeboid
THE JOURNAL CF COMPARATIVE NEUROLOGY, VOL. 30, NO. 3
272 KIYOYASU MARUI
glia cells, which were met more frequently around the blood-
vessels than in the other parts of the synapse. Large and old
cells showed vacuoles of different size in their bodies. Some of
these glia cells showed another kind of granules, which were dem-
onstrated by means of the thionin-eosin stain in a characteristic
metachromatic or more or less blue-violet color. The granules
fill the cell body as well as the processes; the size is variable, but
in any one cell they are of almost equal size. The shape of these
granules is round or that of irregular lumps. Another charac-
teristic of this kind of granules is that in the illumination by elec-
tric light they are especially beautifully observable. In the space
around the blood-vessels and also in other parts of the synapse I
often noticed a group of these granules; this is to be interpreted
as the section of a cell or its process, bearing this kind of granules.
As far as my observation went, these granules do not lie freely
in the tissue.
What is the nature of these granules? Reich (27, 28, 29, 30)
demonstrated in the Schwann cells of the peripheral nerve fibers
rod- or comma-shaped fairly large granules (7-granules), which
were brought out in a characteristic metachromatic stain by
means of thionin, toluidin-blue or kresyl-violet, and he identified
these granules with the protagon of Liebreich on account of the
similarity of the staining reaction and of the solubility in warm
alcohol (45°) and in warm ether. He found, moreover, that they
are soluble also in warm xylol, that they are not at all stainable
in acid stains, and also that they are especially beautifully ob-
servable in the illumination by electric light.
Later, in certain pathological conditions, Alzheimer (1) dem-
onstrated in the neurogolia cells granules which gave a char-
acteristic metachromatic basophile stain by means of toluidin-
blue and thionin; he identified these granules with the 7-granules
of Reich and called them metachromatic basophile granules, al-
though the granules differed somewhat from the z-granules mor-
phologically. Idid not test the properties of solubility of the gran-
ules, which I observed in the glia cells; but on the basis of
their staining reaction and-morphological characteristics I as-
sume that they are identical with the metachromatic basophile.
EFFECT OF OVER-ACTIVITY ON SYNAPSE Sie
granules of Alzheimer. Whether these granules consist of pro-
tagon or not, is another question; recent studies raised doubt
against the real existence of protagon as a uniform substance.
(Rosenheim, Tebb, Thudicum, cited in (18)). I should add here
that I always used bergamot oil instead of xylol in the process
of paraffin embedding.
Alzheimer (1) declared that he did not find, or he found at the
most only indications of, these granules in the amoeboid glia cells;
but as far as my observation went, I found a number of amoeboid
gha cells with this kind of granules. The finding that these gran-
ules are observable in the cells of blood-vessels, as will be de-
scribed later, and in the glia cells around the blood-vessels relatively
more numerously, and the fact that in Scharlach stain fat drops
are found in those cells, as will be related below, make one assume
that they are transported toward the blood-vessels and that these
granules give rise to the production of fat as Alzheimer did.
Reich assumed that the appearance of this kind of granules has
a relation to the decay of the myelin sheath; according to Alz-
heimer (1), it is not, however, necessary for the appearance of this
kind of granules. As the neuroglia cells of the synapse of Mauth-
ner’s cell lie mostly at the border of the axone cap, where the nerve
fibers lose their myelin sheaths, I cannot decide this question
from my own observation. Moreover, as the granules appear
only in fatigue, it is probable that they have something to do
with a pathological nutrition condition of the nerve tissue; the
fact that they are a catabolism product is acceptable because they
are transported toward the blood-vessel, and it is also probable
that they are changed into fat, just like the other catabolism
products.
On the basis of the above-described facts, it is quite clear that
in fatigue a number of amoeboid glia cells are produced in and
about the synapse, which carry different kinds of catabolism prod-
ucts in their cell body as well as in their processes. As already
repeated, I found these amoeboid glia cells relatively more
numerous around the capillaries in and about the synapse. It was
also remarked that many a conglomeration of metachromatic
basophile granules was demonstrated around the blood-vessels
274 KIYOYASU MARUI
with a different .size and shape (fig. 10). This was also the case
with the methyl-blue granules and the fuchsinophile granules.
All these granules were embedded in the evenly and lightly
stained ground substance, which is to be interpreted as the sec-
tion of the cell body or the process of the glia cell. As far as
my observation reached, these granules did not le free in the
space around the blood-vessel or in the tissue.
The cells of the adventitia of blood-vessels in the normal brain
show very little protoplasm around their nuclei; but in fatigue
cases, when a number of amoeboid glia cells are present around the
vessels, they show a larger mass of protoplasm. In the thionin-
eosin preparation I observed occasionally the metachromatic
basophileg ranules in them. Scharlach R (fig.13) revealed many
fat drops in the latter. It must be added here that in the
neuroglia cells around the vessel and in the synapse fat drops were
demonstrated within the protoplasm. According to my obser-
vation, there was, however, very little fat lying in the tissue or
in the space around the vessel.
THE ‘FUELLKOERPERCHEN’ OF ALZHEIMER
After the description of the changes of the glious reticulum and
the glia cells of the synapse, one more manifestation, which goes
hand in hand with the appearance of the amoeboid glia cells is to
be mentioned. I have already described that in sections in which
a number of typical amoeboid glia cells were observable, many a
protoplasm mass of different size and shape appears in the peri-
vascular space and in the other parts of the synapse. These pro-
toplasm masses were found to be sometimes homogeneous and
sometimes granular and they occurred usually near the amoeboid
glia cells. Some of these masses must be regarded as the section
of the cell body or the process of an amoeboid glia cell (fig.10) ;
but some of them show evidently that they were cast off from the
body of the amoeboid glia cells (fig. 15). It was also related
above that the processes of the glia cells, which are not in possession
of the proper attributes of an amoeboid glia cell are broken up in
their substance and are reduced to fragments. I assume that I
EFFECT OF OVER-ACTIVITY ON SYNAPSE BID
can homologize these protoplasm pieces with the ‘Fuellkoerper-
chen’ of Alzheimer (1); as far as the origin of these corpusceles is
concerned, I agree with Alzheimer, in so far as the reticular beams
of glia tissue swell and are loosened in their substance so that’ fi-
nally they fall here and there to pieces. The probability of a post-
mortem alteration cannot come into consideration; it must be
emphasized here that the fish brains were always fixed in their
_fresh condition.
THE RELATION OF THE AMOEBOID GLIA CELLS TO THE GLIA
RETICULUM
As stated before, the glia reticulum of the synapse of the Mauth-
ner cell in fatigue was found in a more or less advanced state of
deviation from its physiological configuration. The net figure
appeared less sharp and the meshes were found more irregular;
the net beams were observed either swollen or lessened in thick-
ness. The substance of the net beams appeared loosened, and in
the state of extreme deterioration the net beams were here and
there broken up and reduced to fragments so that in some parts
of the synapse the net figure was no longer in evidence. The
question now arises whether these manifestations of the peri-
cellular reticulum are to be attributed to artefacts caused by the
process of preparation of the sections or to be claimed as ante-
mortem phenomena caused by the over-activity. The follow-
ing circumstances speak explicitly for the latter; first, I got the
same results in different kinds of preparations; second, I did not
find any case of resting control fish, in which a similar picture of the
reticulum was observed, and, third, I noticed a number of amoe-
boid glia cells with various catabolism products and the so-called
‘Fuellkoerperchen. It must be emphasized here that the nu-
clei of the glia cells of the synapse showed not only regressive,
but also progressive changes.
Buscaino (5), Rosental (31), and others studied the postmortem
appearance of the amoeboid glia cells. Wohlwill (35) declared
that through the postmortem decay of neuroglia tissue glia cells
take occasionally the amoeboid appearance, and the occurrence
of the methyl-blue granules and the ‘Fuellkoerperchen’ does not
276 KIYOYASU MARUI
always indicate the previous existence of amoeboid glia cells. In
the present work all the fish brains, both normal and fatigue, were
placed in the fixing solutions in their fresh condition, and the like-
lihood of postmortem production of the amoeboid glia cells cannot
at all come under consideration. Attention must especially be
called to the fact that in all my control preparations not a single
amoeboid glia cell did come to my observation in and about the
synapse.
What would then be the mechanism of the breaking up of the
glia reticulum? It is very hard to answer this question definitely.
Hisath (10), who studied and demonstrated the protoplasmatic glia
structure very well by his own method, suggested that in the sec-
tions in which amoeboid glia cells occurred either in the marrow
only or in both the marrow and the cortex, the glia cells with deli-
cately arborized protoplasm processes do not come to observation.
Alzheimer (1) also made the same observation; instead of the glia
cells with arborized protoplasm processes, he found by means
of the same method glia cells with a little increased plasm with-
out any process or those with large cell body carrying different
kinds of granules. He took for granted that with the appearance
of the amoeboid glia cells the other glia structure also sustain
some alteration. On another occasion Alzheimer (1) made the
observation that in a cortex, which showed no glia cells with pro-
toplasmatic processes by means of Mallory’s method, the. Golgi
method revealed such cells with processes. On the ground of
this finding, he carefully expressed his opinion, declaring that the
processes did not here go into decay or were not withdrawn by
the cells, but merely did not show the affinity to Mallory’s
hematoxylin.
Now, as already remarked, the processes of the glia cells of the
synapse arborize and unite into a uniform glia reticulum in the
synapse. This condition can be beautifully demonstrated in the
Levaditi preparations. In fatigue a number of the glia cells are
converted into the amoeboid glia cells and they lie free from the
reticular structure of neuroglia tissue; and in this case attention
must be called to the fact that some other glia cells in the synapse
still show their relation to the reticulum and that I could observe
a EEE
EFFECT OF OVER-ACTIVITY ON SYNAPSE vk
almost every stage of the dissolution of an amoeboid glia cell
from the diffuse glia reticulum. Near the amoeboid glia cells
the ‘Fuellkoerperchen’ or the fragments of the Golgi net beams
were also observable, as already stated. So I came to the
conclusion that in the extreme stage of fatigue. a number of amoe-
boid glia cells are produced and are set free from the reticulum.
Jakob (18) recently described in his article on secondary degenera-
tion the detachment of the glia cells from the diffuse glia reticu-
lum. Of course this process took place only slightly and on a
small scale in my material, but the mechanism would here be the
same. At the same time the substance of the reticular struc-
ture becomes looser, and finally some of the beams undergo disso-
lution and fall to pieces, so that the above-described alteration
of the reticulum takes place. What effect the fixing solutions
have here on the more or less loosened reticular beams, I can
not say; but I believe that the mechanism of the breaking up
of the glia reticulum could well be interpreted by thé above-
described facts.
As far as I know, the studies of the pathological snes of
the pericellular retictaltin are very few; the studies of Eisath (10)
and Alzheimer (1) were not especially directed toward the peri-
cellular reticulum, but merely to the glia reticulum in general.
Besta (4) investigated the behavior of the pericellular reticu-
lum in certain pathological conditions, but his description does
not explain the mechanism of the deterioration clearly enough
and he did not mention the appearance of the glia cells at all. As
far as the manifestation of the structure in question in fatigue is
concerned, this investigation is unique.
THE BIOLOGICAL SIGNIFICANCE OF THE AMOEBOID GLIA CELLS
AND THE CONCLUSION FROM THE RESULTS |
As far as my observation went and so far as the histological
’ technique used brought out the facts, over-activity caused no def-
inite change in the nervous structure of the synapse of the Mauth-
ner cell. Considering the nature of the experiment, one should
not be surprised at this; moreover, the structures in question are
so fine and the results of the technique for those structures are
278 * KIYOYASU MARUI
sometimes so inconstant that one should be very careful to attrib-
ute any pathological significance to any slight histological mani-
festation in those structures. The appearance of the amoeboid
glia cells may indicate, however, that some catabolism process
takes place in and about the Mauthner cell. Alzheimer (1),
who investigated thoroughly the amoeboid glia cells and the
catabolism processes in the nerve tissue, said about the cases in
which a number of amoeboid glia cells were found without any
finding on the side of nervous structure, that some catabolism
products which escape the microscopical demonstration at the
present time would be produced on account of the disturbance
of the nutrition of the nerve tissue. The finding of the amoeboid
glia cells in the synapse might show that in over-activity some
catabolism products are produced as the effect of pathological
nutrition condition or owing to dilapidation of the nervous
structure, not demonstrable at the present time. These would
stimulate the formation of the amoeboid glia cells to serve as
scavengers. That postmortem production of the amoeboid gla
cells cannot come under consideration, I already remarked.
Rosenthal (31), who wanted, to interpret the appearance of
amoeboid glia cells with methyl-blue granules as a sign of necro-
biosis of neuroglia tissue, regarded the formation of the amoeboid
glia cells with fuchsinophile granules as that of increased scay-
enging activity; the amoeboid cells with these granules were also
found in fatigue, as remarked. According to Wohlwill (34),
different kinds of diseases which show the amoeboid glia cells have
edema as their common cause. ‘The question whether over-activ-
ity causes edema in the region of the synapse and a swelling
process of the glia cells can come under consideration or not,
must be left undecided here. So far I may assume that in over-
activity a catabolism process in a wide sense takes place in the
synapse, which comes under the fourth category of catabolism
processes of Alzheimer (2). But I cannot state definitely whether °
the catabolism products come only from the synapse or from
both the synapse and the cell; the latter appears to me more
probable.
EFFECT OF OVER-ACTIVITY ON SYNAPSE 279
The question whether the amoeboid glia cells come from newly
produced glia cells or from those which were present in the
synapse, is very hard to answer; but the finding of many amoeboid
glia cells in spite of very little increase of the glia cells may in-
dicate that at least some of the old glia cells give rise to amoeboid
glia cells. As far as the transition of one kind of catabolism prod-
uct into another within the protoplasm of the amoeboid glia
‘cells is concerned, I cannot add much to the description of Alz-
heimer (1). The fact that we find the amoeboid glia cells with
different granules and lipoid substance relatively more numerous
around the blood-vessels than the other parts indicates that the
catabolism products are assimilated into different granules by the
amoeboid glia cells and carried to the blood-vessels and de-
posited in the cells of blood-vessels as fat and later are gradually
disposed of.
In my experiments I could not find a single case in which the
picture of the ‘neuronophagia’ of the Mauthner cell was observed;
this might be interpreted to mean that the alteration of the cell
body did not go so far in over-activity. Dolley (8) described
cell death as the effect of over-activity; it is rather strange to
me that no attention was directed to the changes of the glia cells
in his article.
SUMMARY
Careful investigation of the cell body as well as the synapse of
the fatigued Mauthner cell in Ameiurus revealed a number of in-
teresting findings, which can be summarized as follows:
1. The cell body was found either swollen or shrunken; the
turgescence was regarded as the result of over-activity and the
shrinkage as that. of exhaustion.
2. The Nissl substance was in a more or less advanced stage
of chromatolysis and thereby the cytoplasm was stained vari-
ously deeply. On the border of the nucleus a mass of stainable
substance was observed as the restitution phenomenon of the
Nissl bodies on the side of the nucleus. It was also accepted
that a mutual interchange of substance occurs between nucleus
and protoplasm in activity.
280 KIYOYASU MARUI
3. The nucleus was also either swollen or shrunken; in the
swollen nerve cell it was sometimes removed to one side of the
cell body. The nucleolus was also found swollen sometimes and
other times shrunken and it was of angular or otherwise irregular
shape. . .
4. In the nervous structure of the synapse no definite altera-
tion could be brought out; but the synapse showed a number of
amoeboid glia cells with methyl-blue, fuchsmophil, and meta-
chromatic basophile granules. Also fat drops were demonstrated
in the glia cells and in the cells of the blood-vessels.
5. The reticular glia structure of the synapse appeared in
many cases of fatigue in more or less advanced deviation from
its normal configuration and even broken up in sonte parts;
this was interpreted as the result of the detachment of the
amoeboid glia cells from the reticulum, as also the effect of the
loosening and dissolution of the net beams.
6. The appearance of the amoeboid glia cells showed that some
catabolism process occurs in the synapse as the effect of patho-
logical nutrition conditions in fatigue.
Ld
EFFECT OF OVER-ACTIVITY ON SYNAPSE 281
LITERATURE CITED
1 Auzueimer, A. 1910. Beitrige zur Kenntnis der pathologischen Neuro-
glia und ihrer Beziehungen zu den Abbauvorgiingen im Nervengewebe.
Nissl’s und Alzheimer’s Histologische u. Histopathologische Arbeiten,
3.3.
1913 Uber die Abbauvorgiinge im Nervensystem. Referat auf d.
VII Jahresvers. d. Ges. Deutscher Nervenirzte in Breslau, Sept., 1913.
Deutsche Zeitschr. f. Nervenheilkunde, Bd. 50, 1914.
3 BartTeLMEzZ, G. W. 1915 Mauthner’s cell and the nucleus mortorius teg-
menti. Jour. Comp. Neur., vol. 25.
4 Besta, C. 1910 Sul modo di comportarsi dei plessi nervosi pericellulari in:
aleuni processi patologici del tessuto. Riv. d. pat. nerv. e ment., 15.
5 Buscatno, V. M. 1913 Sulla genesi e sul significato delle cellule ameboidi.
Riv. di patol. nerv. e. ment. 18. (Referat, Zeitschr. f. ges. Neurol. u.
Psycht., Referate u. Ergebnisse 8, 1914.)
6 Dotury, D. H. 1909 The morphological changes in nerve cells resulting
from over-work in relation with experimental anemia and shock.
Journ. of Med. Research.
bo
7 1910 The neurocytological reaction in muscular exertion. Ameri-
can Journ. of Physiology, 25.
8 1911 Studies on the recuperation of nerve cells after functional
activity from youth to senility. Journ. of Med. Research, 24.
9 1913 The morphology of functional activity in the ganglion cells of
the crayfish, Cambarus virilis. Arch. f. Zellforgchung, 9.
10 Ersatu, G. 1911 Weitere Beobachtungen iiber das menschliche Nerven-
stiitzgewebe. Arch. f. Psychiatrie u. Nervenkrankheiten, Bd. 48.
11 Eve, F. C. 1896 Sympathetic nerve cells and their basophile constituent
in prolonged activity and repose. Journ. of Physiology, vol. 20.
12 GoLDScHEIDER UND Fxiatrau 1898 Normale und Pathologische Anatomie
der Nervenzellen. Berlin, 1898.
13 1898 Weitere Beitriige zur Pathologie der Nervenzellen. IV. Mit-
teilung. Uber Verainderungen der Nervenzellen bei menschlichem Te-
tanus. Fortschritte der Medizin, 1898.
14 Hopaez, C. F. 1892 A microscopical study of changes due to Ainetional ac-
tivity in nerve cells. Journ. Morph., vol. 7.
15 1894 A microscopical study of the nerve cell during electrical stim-
ulation. Jour. Morph., vol. 9.
16 1895 Changes in ganglion cells from birth to senile death. Obser-
vation on man and honey bee. Journ. of Physiology, vol. 18.
17 Hotmeren, F. 1900 Studien in der feineren Anatomie der Nervenzellen.
Anatomische Hefte, 15.
18 JaAxos, A. 1912 Uber die feinere Histologie der sekundiren Faserdegenera-
tion in der weissen Substanz des Riickenmarks (mit besonderer
_ Beriicksichtigung der Abbauvorgiinge). Nissl’s u. Alzheimer’s His-
tolog. u. Histopatholog. Arbeiten, Bd. 5
19 Kocurr, R. A. 1916 The effect of activity on the histological structure of
nerve cells. Journ. Comp. Neur., vol. 26.
282 KIYOYASU MARUI
20 Lampert 1893 Note sur les modifications produites par l’excitation elec-
trique dans les cellules nerveuses des ganglions sympathiques. Soe.
de Biolog. (Cited in (32) ).
1 LuGaro Cited in (82).
2 Luxempure 1899 Uber morphologische Verinderungen der Vorderhorn-
zellen des Riickenmarks wihrend der Tatigkeit. Neurolog. Central-
blatt.
23 Mann 1895 Histological changes induced in sympathetic motor and sen-
sory nerve cells by functional activity. Journ. of Anatomy and Phys-
lology.
24 Marur, K. 1918 On the finer structure of the synapse of the Mauthner cell
; with especial consideration of the Golgi net of Bethe, the nervous ter-
minal feet, and the ‘nervous pericellular terminal net’ of Held. Jour.
Comp. Neur., vol. 30.
25 Matures 1898 Riickenmarksbefunde bei zwei Tetanusfiillen. Deutsche
Zeitschr. f. Nervenheilkunde.
26 Pucnat 1898 Des modifications histologiques de la cellule nerveuse dans
ses divers etats fonctionels. Bibl. Anat. (Cited in (32).
27 Reicu, F. 1905 Uber die feinere Struktur der Zelle des peripheren Nerven.
Allgem. Zeitschr. f. Psychiatrie.
oR)
28 1907 Diskussion zu dem Vortrage von Alzheimer. Allgem. Zeitschr.
f. Psychiatrie.
29 1907 Uber den zelligen Aufbau der Nervenfaser auf Grund mikro-
histiochemischer Untersuchungen. Journ. f. Psych. u. Neurolog.,
Bd. 8.
30 1905 Zur feineren Struktur der Zelle des peripheren Nerven Allg.
Zeitschr. f. Psych., 62.
31 RosenrHaL, S. 1913 Experimentelle Studien tiber amoeboide Umwandlung
der Neuroglia. Nissl’s u. Alzheimer’s Histolog. u. Histopatholog. Ar-
beiten, Bd. 6. Referat: Zeit. f. d. ges. Neurolog. u. Psychiat., Referat
u. Ergenbnisse, 8, 1914.
32 Ssonvatyt, E. 1903 Die Nervenzellenverinderungen bei Tetanus und ihre
Bedeutung (im Anschluss an einen Fall von menschlichem Tetanus).
Jahrbiicher f. Psych., 23.
33 Vas, F. 1892 Studien iiber den Bau des Chromatins in der sympathischen
ganglienzelle. Arch. f. mikroskop. Anatomie, Bd. 40.
34 WouLwiLL, F. 1914 Uber amoeboid Glia. Virchow’s Archiv f. patholog.
Anat., 216.
35 Woutwit., F. Diskussion zu Alzheimer (2).
' be a rh G H
NEDA Ages 5.
.- “aed : | ied f ; 4
ye me tee it ars
' ; te + f, ies
' hae Sly a) ware ¢ A ART
hie alt ee eee
0) 4) oT ena nT RS ee Waa ,
1D of
Resumido por C. Judson Herrick, por el autor, O. Van der Stricht.
El desarrollo de las células de los pilares, el espacio del ttinel y
los espacios de Nuel del 6rgano de Corti.
El espacio del ttinel se desarrolla alrededor del fasciculo del
nervio espiral, el cual camina entre las porciones nuc!eadas de las
células internas y externas de los pilares. En su origen es una
hendidura intercelular cuyo contenido liguido es elaborado por el
citoplasma vacuolar de las células de los pilares y se vierte en el
espacio adyacente. Algunas partes de este protoplasma secretor
sufren un proceso de citolisis, de tal modo que la hendidura crece
y su contenido liquido aumenta en cantidad a sus expensas. El
autor describe con detalle el desarrollo ulterior de las células de
los pilares y sus cabezas. El primer espacio de Nuel aparece en
forma de una hendidura longitudinal situada entre los pilares
externos y las células ciliadas externas y en su interior se acumula
el liquido segregado por los pilares externos. En las superficies
laterales de estos Ultimos se proyectan vesiculas de seerecciém
claras, las cuales experimentan un proceso de citolisis y liquefac-
cidn. Las células externas de los pilares verifican na emigracién
embrionaria desde la primera fila de células ciliadas externas
hacia las células internas de dichos pilares. El autor describe
el desarrollo del segundo, tercero y cuarto espacio de Nuel. El
contenido liquido del ttinel y el del primer espacio de Nuel se
mezclan a través de hendiduras existentes entre los pilares exter-
nos y comunican con las de los segundo, tercero y cuarto espacio
de Nuel. El liquido de todos los espacios de Nuel esta separado
de la endolinfa del conducto coclear por los techos, muy delgados,
de estos intersticios. Tales estructuras realizan, indudablemente,
la propagacion de las ondas vibratorias desde la membrana basilar
hasta la membrana tectoria, contenida en el canal coclear.
Translation by José F. Nonidez
Columbia University
1: eis
“ig elf a & ete
Sa eda ethavie: yer te tobeeieotis rn 3
eat eyaeth STI We oH 7 (1% ! lana
Pppdetiger goibisnsd' At ia Canidae He dian
: Peal iain 0 ; sa TG BS, it: <9 ae leM: ci 3 Teteshig p
id ‘onbbhnOht inc oj 1 odcial Abin aul ul Se egauisoes
die 12 pes me bas Pagid sl ha ‘ting ACliy ta Seo a
aE ee SG ikea ade orale Wt ft sepsis
ey babies) pi Xu . a ae
alo: ‘otitoonaliiys pert eyes
He Ione SUL AT siriayt itis
a) ae epei herp aL OUTS OG
iy we Ht Sip upg oe utile ete B10?
Ser beokeniety LT 4 i
“nano bie cel Sun
Bebe aint dreary
; be titre
Resumido por el autor, Leslie B. Arey.
Un mecanismo retinal para la vision eficiente.
Las células visuales y el pigmento retinal de muchos vertebra-
dos inferiores exhiben sorprendentes movimientos a la luz y en
la oscuridad. Los experimentos comunicados previamente han
establecido que la rapidez de estos cambios ha sido muy exagera-
da. La validez de lasuposici6n, también muy extendida, respecto
a su umbral de sensibilidad extremadamente bajo, ha sido in-
vestigada después por el autor. Tal determinaci6én es importante
en vista de otra suposicién discordante que considera que los
cambios en la posicién de las células visuales a la luz intensa y
a la difusa favorecen la visién de los conos y bastones, respectiva-
mente, mientras que los movimientos correspondientes del pig-
mento retinal también aumentan mecdnicamente la eficiencia
visual. En otras palabras, si los beneficios reputados como
adaptativos se derivan de tales cambios fotomecanicos, las re-
acciones a la luz difusa deben ser esencialmente idénticas con
las que se sabe ocurren en la oscuridad total. Esta suposicién
sin embargo, es en su mayor parte gratuita. Las reacciones de
estos elementos bajo la acci6én de la luz de intensidad graduada
prueban que el umbral de estimulacién es notablemente elevado.
En general, el maximo de reaccién hacia la luz aparece primero
en una intensidad luminosa que permite justamente la lectura
de los caracteres de imprenta ordinarios. De aqui que la su-
puesta sensibilidad fética elevada de las células visuales y pig-
mento retinal no queda probada, mientras que las condiciones
mecénicas para una visién de la penumbra, tedricamente mas
eficiente, se establecen sobre una base experimental.
Translation by José F. Nonidez
Columbia University
AUTHOR’S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, MAY 1
A RETINAL MECHANISM OF EFFICIENT VISION!
LESLIE B. AREY
The Anatomical Laboratory of the Northwestern University Medical School
TWO TEXT FIGURES
PRELIMINARY
It is a well-established fact of retinal physiology that the
visual cells and retinal pigment in many of the lower vertebrates
exhibit striking movements in light and in darkness (’15 a, 716),
although in man and other mammals such changes apparently
are slight (’15 b).
The pigment cells have processes, probably fixed, which inter-
digitate with the visual elements and in which the pigment
granules migrate to and fro (figs. 1 and 2). The visual cells
likewise modify their positions due to the contractility of the
so-called myoid, which is that portion of the inner member of
the rod or cone between the ellipsoid and the external limiting
membrane. The extensibility of the myoid is variously de-
veloped, but in some animals the extremes of change may be as
one is to ten (’16).
There is represented in figures 1 and 2 the condition of the
retina, with respect to these changes, as it is characteristically
found in a fish, the common horned pout, Ameiurus. In dark-
ness (fig. 1) the pigment withdraws toward the chorioid, thus
exposing the visual rods and cones; of the two types of visual
elements the cone is extended, whereas the rod is so retracted
that its ellipsoid lies close to the external limiting membrane.
In light (fig. 2) the appearance is reversed; the pigment now
1 Contribution No. 69, March 3, 1919. The experimental data were obtained
at the Fairport Biological Station while a guest of the United States Bureau of
Fisheries.
343
344 _LESLIE B. AREY
Fig. 1 The effect of total darkness on the position of the retinal pigment and
of the visual rods and cones of the fish Ameiurus nebulosus. The pigment is
withdrawn toward the chorioid; the cone myoid elongates, the rod myoid
shortens.
Fig. 2. The effect of bright, diffuse daylight on the same retinal elements of
Ameiurus. The pigment moves forward toward the external limiting mem-
brane, thereby masking the visual rods and cones. At the right of the figure
the positions of these visual cells are indicated—rods elongated, cones shortened.
ABBREVIATIONS
chr., chorioid; mb.lim.ex., external limiting membrane; my.bac., rod myoid;
my.con., cone myoid; pd.cl.pig., base of pigment cell; scl., sclera; st.nl.ex., exter-
nal nuclear layer; st.pig., pigment layer.
oa
se
— ee
a
A RETINAL MECHANISM OF EFFICIENT VISION 345
pushes outward to mask the visual cells, while these latter oceupy
mutually reciprocal positions—rods elongated, cones shortened.
Such positional changes have been interpreted as of use in
furthering efficient vision. It seems logical that the masking
pigment of the light phase would serve to protect the delicate
visual elements from the overstrong influence of light (Chiarini,
06), while the insinuation of such pigmented processes between
the individual visual elements effects for the latter a certain
degree of optical isolation by acting as an absorbent of dis-
persed rays refracted from neighboring cells (Garten, ’07).
From the standpoint of sensory reception, the sharpness of the
retinal image is in this way enhanced. The withdrawal of the
pigment in dim light might be thought to involve a response
which allows the, visual cells to utilize all the weak light available.
Furthermore, there is good reason to believe that the cones are
concerned with bright-light vision, the rods with dim-light or
twilight vision. Hence a shortening of the cone in bright light,
drawing it down nearer the source of illumination, while the rod
at the same time elongates and is thus moved out of the way,
would appear to be a useful maneuver (Herzog, ’05; Exner and
. Januschke, ’06). The converse procedure in dim light—by which
the rods are shortened and the highly refractile cones, with their
lens-like ellipsoids no longer masked by pigment, are length-
ened—would be equally advantageous (Garten, ’07; Arey, ’15 a).
For the detailed applications of these apparently adaptive
responses the reader is referred to Garten (’07), who gives an
extended consideration of the correlations within the vertebrate
classes between the morphology, optical qualities, and distribu-
tional ratios of the rods and cones on the one hand, and, on the
other, their movements together with the migrations of the
retinal pigment.
Interesting and logical as these speculations may be, they nev-
ertheless lack a sound experimental basis. It is certain that the
movements as summarized in a preceding paragraph occur re-
spectively in daylight and in darkness; but since responses in
total darkness cannot be useful in the manner suggested, it is
obvious that in order to derive the reputed adaptive benefits
346 LESLIE B. AREY
from such photomechanical changes, the responses in dim light
must be identical with those demonstrably occurring in the dark.
This assumption, however, is largely gratuitous.
The supposition of an identity between the set of responses
ensuing in dim light and in darkness is further complicated by
certain conflicting statements and beliefs. There is a general
impression current that the retinal pigment and visual cells react
to the slightest traces of light; this is reflected in the statements
of many workers who have feared their results would be im-
paired unless the strictest precautions were observed.
On the basis of actual experimentation, however, the earliest
observations are encountered in the writings of Angelucci (90),
who reported that five minutes of candle light caused the frog’s
cones to be highly shortened, whereas the pigment remained as
in darkness. On another page, nevertheless, he records that
after twenty minutes of twilight the pigment assumed the light
position, but the cones are influenced to a less degree!
Somewhat later, Pergens (’97) wrote that after a five hours’
exposure to colored lights (red, yellow, green, and blue) of an
intensity such that colors could be distinguished by an observer
after one minute of dark adaptation, the cones of the white fish,
Leuciscus rutilis, are strongly retracted. The weakest response
was reported from the blue—a result, however, not in agreement
with Herzog’s (’05) later findings on the frog. The latter
worker found the blue-violet most effective, although it should
be added that the duration of exposure employed by him was
only two minutes.
In a further communication (’99) Pergens confirmed his earlier
conclusion that the pigment migrates least extensively in red
light (equal intensities being used), but mod fied his previous
belief regarding the inefficiency of the blue to provoke cone
retraction.
Exner and Januschke (’05) performed some experiments,
which, unfortunately, are not trustworthy as evidence. Speci-
mens of the fish Abramis brama were exposed during the late
afternoon, the experiment continuing through the period of
failing ight and terminating at dusk. Examination showed the
cones to be in the position characteristic of light.
A RETINAL MECHANISM OF EFFICIENT VISION 347
In a few experiments Garten and Weiss (’07) found that light
in which colors could not be recognized, acting for five or more
hours, caused the cones of the frog to assume a position inter-
mediate between that characteristic of bright light and of total
darkness. A limited pigment migration was reported as well.
These same workers made observations on the fish Abramis
brama, contained in a dish lighted by reflection from the cover.
Two grades of illumination were chosen: in one colors held within
the container could not be distinguished; in the other they were
recognizable. According to the results given, in the first grade
the cones were maximally retracted in nine retinas and partially
soin seven. In the second grade the cones were found shortened
in all the eight retinas used, whereas the pigment exhibited no
noteworthy change except in the sector constituting the ventral
one-third of the eye.
The foregoing statements reveal the following conditions:
Angelucci’s several pronouncements in the same publication, if
not actually contradictory, at least serve to befog the issue.
The use of colored lights by Pergens was unfortunate; moreover,
the reliability of certain of his conclusions is questionable. Exner
and Januschke’s experiments were so ill devised as to furnish no
crucial evidence. The results of Garten and Weiss suggest an
extreme sensitivity of the cones to low light intensities, whereas
the pigment patently has a higher threshold. Finally, there
exist no data concerning the responses of the rods to weak light.
It is clear that if the visual elements and retinal pigment are
as highly sensitive to mere traces of light, as often has been
held, they can assume no useful positions in the ordinary dim
light of rod vision, while the utility of a response evoked under
conditions of virtual darkness will still await an explanation.
Accordingly, it was with the intent of learning the true conditions
that this investigation was undertaken.
348 LESLIE B. AREY
PROCEDURE
Essential to success in a determination of this kind is the
choice of appropriate experimental animals. Previous experi-
ence with a variety of forms led to the selection of two fishes and
the frog. The cones of the golden shiner, Abramis crysoleucas,
have large, conspicuous ellipsoids and are so highly mobile that
the light-adapted myoid can shorten to one-tenth its maximum
length in darkness (16). The cones of the common grass frog,
Rana pipiens, are easy to observe, but show a narrower range of
movement, the limits of extensibility being as one is to four.
The rods of the horned pout, Ameiurus nebulosus, not only are
of exceptional size in comparison with the usual diminutive ele-
ments of fishes, but they also undergo changes in length in the
ratio of one to ten (16); the value of Ameiurus for experimenta-
tion of this sort cannot be overestimated. There is a further
inherent advantage in the animals chosen, inasmuch as their
visual cells remain at a uniform level during the characteristic
positional changes. The retinal pigment of ail three animals
exhibits extensive movements: in darkness it is confined to a
narrow zone at the bases of the pigment cells, whereas in the
light it migrates nearly to the external limiting membrane.
Temperature is an additional factor which must be carefully
controlled, although the necessity of this has been recognized
only recently (16). In dark-adapted fishes a temperature
near the freezing point brings about a maximal contraction of
the cone myoid, such as is characteristically associated with the
action of light, while raising the temperature to the limit which
is compatible with life proportionately elongates the myoid.
Light, however, is so much more effective than temperature that
the latter factor does not enter as a complication in the light
adaptation of cones. The relatively slight quantitative effects
of temperature acting in darkness on the position of the rods
and on the distribution of pigment may also be disregarded.
In the frog, on the contrary, high and low temperatures evoke a
maximal expansion from the pigment during dark adaptation,
complete contraction being obtainable only at an intermediate
grade of about 15°C.; the cones, however, are shortened at the
A RETINAL MECHANISM OF EFFICIENT VISION 349
upper temperature limits alone. From these statements it fol-
lows that to obtain significant results from experimentation
upon frogs a temperature of about 15°C. must be maintained,
while with fishes ordinary summer temperature of 22°C., or
higher, is favorable. |
Experiments were conducted in a large, windowless room into
which weak daylight of a non-directive nature could be introduced
from a second room; the latter, in turn, received its light directly
from windows located at the end far from the single communi-
cating door. Animals were tested under five? different conditions
of illumination: 1) total darkness; 2) light in which the presence
of objects could just be determined; 3) light of an intensity
which allows the certain identification of bright colors; 4) light
which just permits the reading of ordinary journal print; 5)
bright, diffuse daylight. Exposures lasted two or three hours or
more.
As experience proved, these seemingly rough criteria of light
intensity are sufficiently accurate for the purposes required;
with a little practice such grades can be kept fairly uniform.
Permanent preparations of Perenyi-fixed, paraffin-infiltrated sec-
tions served as a basis for study.
To further brevity and clearness, the results obtained from
experimentation will be condensed to mere summaries.
EXPERIMENTATION
A. Retinal pigment
1. Frog. In total darkness, and in light of just sufficient
strength to allow objects or colors to be discerned, the pigment
lies in a narrow stratum near the chorioid. When the illumi-
nation is increased just sufficiently to allow the reading of ordi-
nary print, the pigment, for the most part, becomes expanded,
in some cases completely, in others in a zone only three-fourths
the maximal breadth.
2 For Ameiurus another intensity—one which enables ordinary print to be
easily read—was also utilized.
350 LESLIE B. AREY
2. Ameiurus. The pigment of this fish is at first more sensi-
tive than that of the other animals studied, migration being
distinctly initiated in many (or perhaps most) individuals at an
intensity by which the presence of objects can be detected. At
the next grade (colors distinguished), expansion is well ad-
vanced, although the pigment does not extend the maximal dis-
tance, for it is dense at the cell bases, but sparse distad. In
light in which one can just read, the expansion is nearly com-
plete, but does not become maximal until the illumination is
sufficient to allow easy reading.
3. Abramis. The effect of light is not apparent until it is of
such a strength that colors can be recognized. At this intensity
the pigment in about half the individuals was essentially in a
condition of greatest contraction; the remainder showed the pig-
ment well started, but not extended at most more than half the
way to the external limiting membrane. Owing to a sudden
failure in my supply of animals at the end of the season, I secured
but few observations on the effect of light which makes reading
possible; the pigment in those animals studied, however, was in
a state of partial expansion only, so that it appears safe to con-
clude that stronger illumination is necessary to allow expansion
to proceed to completion.
B. Visual cells
1. Cones. The cones of Abramis and the frog remain fully
elongated, in the characteristic dark positions, until the illumina-
tion is so increased that colors are recognized. At this intensity
the cones of Abramis show distinct indications of incipient re-
traction, while those of the frog are less than half their former
length. When illuminated sufficiently to allow reading, the frog’s
cones shorten maximally; the cones of the few Abramis studied?’
were likewise greatly shortened, but not completely so.
® As on the tests on the retinal pigment of this animal, further experimenta-
tion would have been desirable, but the supply of available material suddenly
ceased.
A RETINAL MECHANISM OF EFFICIENT VISION 351
2. Rods. Until an intensity is used at- which printed matter
can be read, the rods of Ameiurus remain in the typical short-
ened condition of darkness; at about this grade, however, they
apparently begin to elongate slightly. Any considerable degree
of lengthening must first appear only in stronger light.
DISCUSSION AND CONCLUSION
These results, as compared with the findings of Garten and
Weiss (compare p. 347), indicate that there exists a rather lower
degree of photic sensitivity in the visual cells and retinal pig-
ment than they maintain. Nevertheless, it must be apparent
to all, not only that in the subjective choice of arbitrary grades
of light a variable personal factor enters, but also that the de-
scription of stages thus chosen cannot be expressed in accurate
terms. It is possible, of course, for any one person to standardize
these grades fairly well for his own experiments; on the other
hand, the determination of a critical intensity, such as that in
which ‘colors can be distinguished,’ depends on. one’s individual
acuity in color discrimination, on the brightness of the test
colors, and on the decision as to whether these colors are to be
just distinguishable with intense scrutiny or to be identified
with ease. In any event, it appears that my first grade of
illumination (that in which objects were discernible) was un-
doubtedly of lower intensity than the weakest employed by Gar-
ten and Weiss' (compare p. 347); it lay far below the point
where colors cease to be recognizable, this latter constituting their
lowest grade.
Moreover, it is not impossible that the intensity of light to
which their animals were actually subjected was higher than
Garten and Weiss supposed. Their animals, contained in a
‘spacious basin,’ received light (electric) reflected downward
from the cover. After the experimenter had accustomed his
eye to total darkness for five minutes, the condition of illumina-
tion was judged by looking downward into the dish at colors
placed directly over the surface of the water. As to whether
this is a method which tends toward underestimating the light
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 4
352 LESLIE B. AREY
conditions actually obtaining in the basin, the reader may decide
for himself.
There is one further circumstance which on casual considera-
tion might be held responsible for the quantitative divergence
between my results and those of Garten and Weiss. They con-
tinued their experiments in most cases for five hours, while my
determinations lasted on the average perhaps three hours. It
seems plausible that the long-continued action of very weak
light might register an effect not manifest in shorter periods.
That this possibility is not operative in the cases under consid-
eration follows from certain other observations of Garten and
Weiss. They report maximal cone retraction in the fish after
the weakest grade of light employed had acted in one series for
three hours and in another crucial series for one and one-half
hours. Garten also records that in light too weak to distin-
guish colors by, a shortening of the cones occurred (‘eintritt’)
in one hour.
The facts developed in this investigation may for conve-
nience be consolidated into the following statement: although re-
sponses of the visual cells and retinal pigment may be initiated
at lower intensities of light, an approach to.a maximal response
is first elicited at an intensity which permits the reading of ordi-
nary print. This signifies that the threshold of stimulation of the
visual cells and retinal pigment ts high; or, in other words, the
assumed great photic sensitinty of these elements 1s disproved.
Furthermore, since the responses in weak light are substantially
identical with those in darkness, the mechanical conditions are
present for a theoretically more efficient dim-light and bright-
light vision, as postulated (compare p. 345).
SUMMARY
The threshold of stimulation of the visual rods and cones and
of the retinal pigment, at which they exhibit their characteristic
photomechanical changes, is high.
The alleged great sensitivity of these elements to light of ex-
tremely low intensity is consequently disproved. .
A RETINAL MECHANISM OF EFFICIENT VISION 353
Although responses may be initiated at lower intensities, in
general an approach to a maximal light response is first elicited
at an intensity which makes ordinary print legible.
Since the responses in dim light approximate those in total
darkness, the mechanical conditions are present for a theo-
retically more efficient dim-light and bright-hght vision than
would otherwise obtain. This increased efficiency depends
upon the assumption by these elements of correlative advan-
tageous positions.
LITERATURE CITED
Ancetucci, A. 1890 Untersuchungen iiber die Sehthatigkeit der Netzhaut und
des Gehirns. Untersuch. zur Naturlehre d. Menschen u. d. Thiere
(Moleschott), Bd. 14, Heft 3, S. 231-357.
Arey, L.B. 1915a The occurrence and significance of photomechanical changes
in the vertebrate retina—an historical survey. Jour. Comp. Neur.,
vol. 25, no. 6, pp. 535-554. é
1915 b Do movements occur in the visual cells and retinal pigment of
man? Science, N.S., vol. 42, no. 1095, pp. 915-916.
1916 The movements in the visual cells and retinal pigment of the
lower vertebrates. Jour. Comp. Neur., vol. 26, no. 2, pp. 121-201.
Curarini, P. 1906 Changements morphologiques qui se produisent dans la
rétine des vertebrés par l’action de la lumiére et de l’obscurité.
Deuxieme partie. La retine des reptiles, des oiseux et des mammi-
féres. Arch. ital. de Biol., T. 45, fasc. 3, pp. 337-352.
Exner, S., unp JanuscHkKE, H. 1905 Das Verhalten des Guanintapetums von
Abramis brama gegen Licht und Dunkelheit. Ber. d. k. k. Akad. d.
Wissensch. zu Wien, Math.-naturw. Kl., Bd. 114, Abth. 3, Heft 7,
S. 693-714.
1906 Die Stibchenwanderung im Auge von Abramis brama bei
Lightverinderung. Sitzungsb. d. k. k. Akad. d. Wissensch. zu Wien,
Math.-naturw. K1., Bd. 115, Abth. 3, S. 269-280.
Garten, S. 1907 Die Verinderungen der Netzhaut durch Licht. Graefe-
Saemisch-Handbuch der gesammten Augenheilkunde, Leipzig, Aufl. 2,
Bd. 3, Kap. 12, Anhang, 130 S.
Garten, S., UND Werss, R. 1907 [Results incorporated in Garten, ’07, pp.
39 and 40.]
Herzoc, H. 1905 Experimentelle Untersuchung zur Physiologie der Beweg-
ungsvorgiinge in der Netzhaut. Arch. f. Anat. u. Physiol., Physiol.
Abth., Jahrg., Heft 5 u. 6, S. 413-464.
Percens, E. 1897 Action de la lumiére colorée sur la rétine. Ann. Soc. roy.
d. sci. med. et nat. de Bruxelles, T. 6, pp. 1-38.
1899 Vorgiinge in der Netzhaut bei farbiger Belichtung gleicher
Intensitait. Zeitschr. f. Augenheilk., Bd. 2, S. 125-141.
Resumido por el autor, Swale Vincent.
Contribucién al estudio de los reflejos vasomotores.
La estimulacién de los nervios sensoriales en perros aneste-
siados con étero cloroformo produce generalmente un aumento
en los movimientos respiratorios, cuando la narcosis no es muy
fuerte o cuando no se emplea el curare. Este aumento de movi-
mientos respiratorios produce una disminucién de la presién
sanguinea y cuando son pronunciados no se puede obtener
ningtin reflejo presor, aun estimulando fuertemente. Con nar-
cosis fuerte o compresi6n del cerebro no se produce la disminucién
de la presién sanguinea debida a esta causa. La interferencia
mecdnica con la circulacién es la causa de esta disminucién,
puesto que se elimina abriendo el torax. Este fendmeno complica
seriamente los experimentos sobre los reflejos vasomotores. En
los perros anestesiados con éter o cloroformo o con el cerebro
comprimido, una débil estimulacién de los nervios produce genera-
Imente una disminucién, y una fuerte estimulacién un aumento
en la presién sanguinea. La frecuencia de la estimulacién ejerce
efectos sobre el reflejo; cuando es rapida se obtiene un aumento,
y con menor rapidez una disminucién. Los nervios mas volu-
minosos responden de un modo mas marcado al estimulo que
los menores. La estimulacién de la piel, mtisculos e intestino
origina generalmente una disminuci6n en la presién sanguinea,
pero si se estimula la piel de un modo violento y extenso se pro-
duce un aumento en dicha presién. Bajo la accién de la morfina
y el curare, por el contrario, un aumento en la presién tiene
lugar generalmente, aunque con la morfina la estimulacién débil
puede producir una disminucién de la misma. La influencia
de las glindulas endocrinas sobre los reflejos vasomotores no
es clara (vedse sin embargo Pearlman and Vincent ‘Endocrin-
ology,” en prensa). El] cambio de reflejo en la estimulacién
de los nervios somaticos se produce principalmente por los efectos
sobre los vasos sanguineos del drea espldcnica.
Translation by José F. Nonidez
Columbia University
ES
AUTHOR’S ABSTRACT OF ''HIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICF, JUNE 2
A CONTRIBUTION TO THE STUDY OF VASOMOTOR
REFLEXES
D. OGATA AND SWALE VINCENT
Physiological Laboratory, University of Manitoba, Winnipeg, Canada
NINETEEN FIGURES
CONTENTS
PRETORIA te oe cele eke oon eee te EE See ee mee eee 355
2. The influence of respiratory movements upon blood-pressures............ 357
3. The effect of the strength of the stimulus upon vasomotor reflexes....... 351
4, The influence of the frequency of stimulation upon vasomotor reflexes... 364
5. The effects upon vasomotor reflexes of stimulating nerve trunks of dif-
ferent categories (sensory, motor, and mixed nerves) and of different
SUAS ON arent cabs cede PEN EE TSS CIEE RATE IN COTES IT ORT Ca Io Ehao a oes fies Oe 366
6. Vasomotor reflexes from nerve terminations..............6.....00e0eeees 370
7. The influence of the ductless glands upon vasomotor reflexes............ 374
8. The question as to which vascular areas are constricted or dilated on
central stimulation of somatic merves.....) 20.2680. oe AOS. 375
MESES ISIS AE Yes AMRIT My in Witw' fe bay Pp Se Ae © Tan's te reece Nhe) Ree eR 376
I, INTRODUCTION
General blood-pressure is affected reflexly by central stimula-
tion of various sensory nerves (reflex vasomotor action). This
subject has been studied already by a number of authors. A
complete list of the older investigations may be found in Tiger-
stedt’s Lehrbuch der Physiologie des Kreislaufs,“! papers by
Asher! and Bayliss? in Ergebnisse der Physiologie, and in Nagel’s
Handbuch der Physiologie (Hofmann"). The history up to
November, 1914, is given by Vincent and Cameron. As to
more recent important investigators of this problem, we may
refer to Porter,2°—*4 Martin, 22-2 Ranson,**—** Gruber,*? and their
respective co-workers, also to Domitrenko® and Hunt.'*:16
Even among these recent investigators there seems to be con-
siderable difference of opinion as to what may be regarded as
the usual or normal response to afferent impulses. Thus Porter
355
356 D. OGATA AND SWALE VINCENT
and Quinby* say: ‘‘It is sometimes urged that in shock the
blood-pressure falls instead of rising on stimulation of afferent
nerves. This abnormal reaction was observed in several of our
experiments.”’ This statement clearly involves the assumption
that a rise is the normal effect, though it is recognized that the
fall is a not very unusual occurrence. Vincent and Cameron*®
seem to be of opinion that the usual effect of stimulating the
central end of the cut sciatic nerve is a rise, and the fall due to a
pure vasomotor reflex is rather rare. So also Hunt. On the
other hand, Martin and Lacey” having observed regularly a
definite drop in blood-pressure by weak stimulation and a rise
by far more strong stimulation, became doubtful of the truth
of the generally accepted doctrine that pressor responses are the
normal results of sensory stimulation.
These differences of opinion must be due to some factor or
factors other than the strength of stimulus. Besides the fac-
tors most usually considered, such as different modes of stimu-
lation, different nerves, different conditions of the same nerve,
different narcotics, drugs, etc., there are two important consid-
erations recently brought forward which unmistakably affect
the vasomotor reflexes or complicate the problem of their
elucidation.
In 1915 Vincent and Cameron“ called attention for the first
time to a fall of blood-pressure caused by increased respiratory
movements. ‘They write: ‘While anaesthesia is fairly complete
the effect of stimulating the central end of the cut sciatic nerve
is a pure and distinct rise. As the effect of the anaesthetic be-
gins to pass off, the effect of stimulation will be a rise of blood-
pressure followed by a more or less pronounced fall. Respira-
tory movements will now be found to have been markedly in-
creased, and the extent of the fall of pressure appears to be at
any rate proportional to the violence of the respiratory activity.”
Martin and Lacey” investigated the influence of the inter-
ruption of the primary current at widely varying rates, but
failed to notice any effect, as also did Hunt™ in his earlier work.
Only quite recently it was clearly pointed out by Gruber’ that
with the same strength of stimulus, pressor and depressor results
VASOMOTOR REFLEXES 357
were obtainable by varying the rate of stimulation from 1 to 20
- stimuli per second. This was later incidentally confirmed by
Hunt.!*
Thus, for the investigation of the complicated problem of
vasomotor reflexes, it became very necessary to investigate each
possible factor separately. In this way only would one be able
to answer correctly for the normal vasomotor response.
The present investigation was undertaken for the purpose of
studying some of the factors separately, of confirming previous
investigations, and of trying, if possible, to reconcile contradic-
tory views as to the conditions which determine any particular
vasomotor response.
We beg to acknowledge our indebtedness to Mr. John Car-
michael for his valuable assistance in all our experiments.
2. THE INFLUENCE OF RESPIRATORY MOVEMENTS UPON BLOOD-
PRESSURES
It is well known that the respiratory center can easily be
affected by central stimulation of sensory nerves. Thus Howell'*
writes in his Textbook of Physiology that ‘‘stimulation of any
of the sensory nerves of the body may affect the rate or the
amplitude of the respiratory movements.’ But no mention is
made of the influence of these movements upon blood-pressure.
The same applies to other text-books, except Starling’s,?* in
which we find, ‘‘The increased respiratory movements will also
aid the venous circulation and have a similar effect in increasing
the systolic output,” which would necessarily bring about a rise
of blood-pressure. But, ‘‘A constant and immediate result of
exaggerated respiratory movements is a fall of blood-pressure,”’
and not a rise, as Vincent and Cameron pointed out. They
found that ‘‘the extent of the fall of pressure appeared to be at
any rate largely proportional to the violence of the respiratory
activity,” that the fall of blood-pressure was ‘‘brought about by
performing rapid artificial respiration by compression of the
thorax,” that ‘‘deep voluntary breathing in the case of the
human subject produced a regular and pronounced lowering of
the blood-pressure,” that ‘‘the more widely the thorax is opened
358 D. OGATA AND SWALE VINCENT
the more the fall of pressure tended to become replaced by a
rise,’ and that the effect of artificial respiration was ‘‘a rise, and :
not a fall, when the animal was under curare,”’ i.e., when a stop
was put to the spontaneous respiratory movements. Thus the
fall of blood-pressure as a result of increased respiratory move-
ments seems to have been sufficiently established.
By the majority of previous observers curare was thought to be
an indispensable drug in the study of the problem of vasomotor
reflexes, with or without any consideration of its action on the
vasomotor center itself. But we know that narcotics and other
drugs are not always free from influence upon these reflexes, as
pointed out by various previous investigators,':9? and, there-
fore, in experimental work they should be reduced to as few as
possible or altogether eliminated (Vincent and Cameron). The
change in character of the respiratory movements, especially
their increase, becomes thus an almost unavoidable complica-
tion in the study of vasomotor reflexes when curare is not used
and the narcosis is not deep enough. If this complication be
left out of consideration, erroneous conclusions may be reached.
Since the appearance of Vincent and Cameron’s paper several
writers have referred to the influence of the increased respiratory
movements. Unfortunately, they are not in complete harmony
with one another. Ranson and Billingsley****> say, ‘‘ With
stronger stimulation the greatly increased respiratory movements _
may no doubt play an important part in the drops in blood-
pressure,’ but Gruber and Kretschmer® write that their “‘ex-
periments do not support Vincent and Cameroh’s theory that the
fall in blood-pressure is brought about by movements of respi-
ration which interfere with the heart’s activity.’”’ This latter
statement seems to deny definitely the respiratory role upon
blood-pressure. Vincent and Cameron did not positively deny
that there is a true vasomotor fall of blood-pressure under cer-
tain conditions as a result of central stimulation of afferent
fibers. But they insisted, and rightly, too, that many apparent
vasomotor falls are really due to increased respiratory move-
ments. We shall see later that the fall of blood-pressure with
weak stimuli is a commoner occurrence than Vincent and
VASOMOTOR REFLEXES 359
Cameron were inclined to believe. At any rate, the matter is
so importan! that we have repeated the experiments to test the
effects of respiration on blood-pressure.
_ We have stimulated electrically the central stump of several
cut nerve trunks (saphenous, tibial, peroneal, sciatic, ulnar, and
median) with various strengths of stimulus.
When the narcosis (with ether or chloroform) was not deep
enough, the respiratory movements were always increased by
strong stimulation. The most frequent response of the blood-
pressure to stimulation, e.g., of the sciatic nerve, may be illus-
trated by figure 1.
With weak stimulation there is practically no increase of res-
piratory movements either in amplitude or in frequency, and
the blood-pressure is either a fall or a fall followed by a more
or less marked rise. When stronger stimuli are applied, the re-
spiratory movements increase either in amplitude or in fre-
quency, or in both, and the blood-pressure rises, instead of
falling, and is followed by a marked fall. The rise of blood-
pressure increases usually in proportion to the development of
the strength of stimulation.
In figure 2 the anesthesia was made much deeper with the
same animal as in figure 1, and stimuli of several strengths were
applied to the same nerve.
There is very little increase of respiratory movements on each
stimulation, and the response of blood-pressure is also small in
degree. The latter, as seen from the figure, is either a fall or a
rise according to the strength of stimulation, and the marked
fall after the rise which is observed in figure 1 simultaneously
with the increased respiration cannot be seen. This suggests
at once that the marked fall accompanied by remarkably in-
creased respiratory movements might be ascribed, at any rate,
mainly, to the influence of the latter movements caused by sen-
sory stimulation. Moreover, under brain compression it is not
very difficult to stop the respiratory movements entirely, and in
this case only very strong stimulation will initiate spontaneous
respiratory movements. Under these conditions a fall of blood-
pressure of the same character as that observed with an in-
360 D. OGATA AND SWALE VINCENT
ereased respiration never occurs. Thus it would not be un-
reasonable to assume that figure 2 shows real vasomotor reflexes,
even though weak, not complicated by the increased respiratory
movements, while figure 1 represents the vasomotor reflex
masked by the effects of increased respiration.
It is often very difficult or almost impossible to obtain any
rise of blood-pressure when the respiratory movements are very
violent. In those cases a marked fall is the only result of cen-
tral stimulation of afferent fibers.
How these increased respiratory movements affect the blood-
pressure was very carefully investigated by Vincent and Cam-
eron. After pointing out several possible causes, they came to
the conclusion that this fall is due to direct mechanical interfer-
ence with the heart’s action and with the return of the blood to
the heart.
In order to confirm this theory, we opened the thorax in the
middle line as did Vincent and Cameron, and found that the falls
disappeared. The contrast is clearly shown in figures 3 and 4.
In these two cases the same nerve of the same animal was
stimulated with the same strength of stimulus, in figure 3 in the
intact animal and in figure 4 with thorax open.
In addition to opening the thorax we cut both vagi, both
phrenici, and as many intercostals as possible on both sides,
without obtaining very different results from those obtained by
merely opening the thorax.
In a very few cases a marked fall of a similar character to that
due to the increased respiratory movements, was observed in
animals with thorax wide open. It was, however, soon dis-
covered that this fall was produced by compression of the in-
ferior vena cava by the heart which became more freely movable
than before through opening the thorax. The heart fell back
upon the soft-walled vein, and thus diminished the flow of blood
to the right heart.
But it is certainly true that by means of almost pure vaso-
motor reflex, i.e., without any or with very little increase of
respiratory movements one can obtain a marked fall preceded
by a rise, as shown in figure 5.
VASOMOTOR REFLEXES 361
Therefore we do not conclude that such a fall of blood-pressure
is always produced by increased respiratory movements. But
the important point for us at present is the undoubted fact that
increased respiratory movements can and do cause a fall of
blood-pressure, and that this fall can be easily eliminated by
opening the thorax sufficiently wide.
These observations, along with various others quoted above
from the paper by Vincent and Cameron both on animals and
on human subjects, confirm fully their statement as to the occur-
rence of a fall of. blood-pressure brought about by increased
respiratory movements, and probably explain the nature of this
fall. We believe that the increased respiratory movements
caused by sensory stimulation form a very important com-
plication which has often led to misunderstanding of the true
vasomotor reflexes.
Gruber and Kretschmer, as mentioned before, deny this re-
spiratory effect upon blood-pressure. They used a slow rate of
stimulation and the fall of blood-pressure was the usual effect.
But the fall is generally thought to be a result of weaker stimu-
lation, and they do not deny that the increased respiratory
movements to a certain degree cause a fall of blood-pressure
when the stimulus is strong enough as to produce them.
Our experiments were made on thirty-three dogs.
3. THE EFFECT OF THE STRENGTH OF THE ey UPON
VASOMOTOR REFLEXES
After repeated experiments by numerous investigators, the
generally accepted view as to the effect upon the vasomotor
reflexes of different strengths of stimulus seems to coincide with
Knoll’s!® original statement, i.e., that a depressor effect is usu-
ally the result of a weak stimulation, while a pressor effect fol-
lows, as a rule, a stronger stimulation. Reid Hunt" pointed out
that weak stimulation was one of the methods of obtaining a
reflex fall of blood-pressure, and Vincent and Cameron noticed
the same fact.
Among more exhaustive investigations on this point we should
refer to those by Porter,?°-*4 Martin,”~**4° and their respective
362 D. OGATA AND SWALE VINCENT
co-workers. The former writer seems to regard a rise of blood-
pressure as the normal vasomotor response, while the latter holds
a different view. Martin and. Lacey’s”? experiments were con-
ducted on cats either under brain pithing, decerebration or
brain compression, or under ether or urethane. The nerves
stimulated were the sciatic, radial, median, ulnar, and saphenous.
The results of their experiments were very definite. ‘‘In every
one of the experiments the stimulation was repeated many
times over a range of stimuli from the threshold value to three
or four times the threshold. Well-marked drops of pressure
followed all such stimulations,’ save in one exceptional case.
Thresholds for pressor reflexes were much higher than those for
depressor reflexes. Thus the experiments of these workers
support Knoll’s statement.
Our own experiments consisted in stimulating various nerves
(sciatic, tibial, peroneal, saphenous, median, ulnar, and vagus)
with induction shocks on dogs under ether, chloroform, and
brain compression. As to the method, we have to mention that
the different effects of weak and strong currents, respectively,
were satisfactorily attained by means of sliding the secondary
coil up to or away from the primary, but that on many occasions
more than one battery was used to obtain a stronger stimulus.
The rate of stimulation was 388 to 54 in a second.
The fall of blood-pressure due to the increased respiratory
movements being taken into consideration, the main results of
our experiments may be summarized as in the following table:
FALL WITH
Ste Peay eS) SAME SS UNE
ASEM AIR NOES RISE WITH AND STRONG errata
STRONG STIMULATION
STIMULATION
Bithvercets crs See St Re ee 20 2 4
Chloroforms: ts. eRe eee eee, 14 4 4
Brain Compression .«.405...¢4.04 00n ee Ve con 12 2 0
Totaly AS ee ok 46 8 8
The term ‘fall’ in the table comprises also a fall followed by a
rise and ‘rise’ also a rise followed by a fall.
alt
VASOMOTOR REFLEXES 363
In forty-six cases out of sixty-two in total, weak stimulation
produced a fall or a fall followed by a rise, and strong stimula-
tion caused a rise or a rise followed by a fall. A typical response
is shown in figure 6.
The animal was under ether and the thorax was very wide
open in the middle line in order to eliminate the disturbance
from increased respiratory movements. Figure 7 shows a similar
response under chloroform.
In the remaining sixteen cases the response was either a fall
or a rise through all strengths of stimulation which we used, and
the different effects with weak and strong stimulation were not
observable.
Thus it does not seem to us unreasonable to conclude that
weak stimulation of the central stump of the cut nerve produces
usually a fall of blood-pressure and a strong stimulation produces
usually a rise.
From these conclusions it may naturally be understood that
from the threshold of stimulation up to a certain point the fall
of blood-pressure increases with the development of the strength
of stimulus, and then the fall gradually decreases until a neutral
point is reached, where the vasoconstriction and dilatation just
counterbalance each other, and finally the rise appears, which
increases usually with the increase of the strength of stimulus,
but cannot continue very long, since powerful stimuli would
elicit vigorous reflex movements of the animal and obscure the
true vasomotor reactions unless indeed the animals were deeply
under curare. As we have been unable so far to find any at-
tempt by previous investigators except Stiles and Martin‘? to
describe this rather peculiar course of vasomotor responses, we
though it worth while to emphasize it in this place (fig. 8).
In our experiments we have employed also stimuli of other
kinds than electric induction shocks, namely, mechanical, ther-
mal, and chemical. In this series thirty-eight out of sixty-seven
stimulations were effective, and of these thirty-five caused a fall
of blood-pressure and only three produced a rise. As the cali-
bration of these stimuli was not so practicable as with induction
shocks, we cannot draw any very positive conclusions, but we
364 D. OGATA AND SWALE VINCENT
are inclined to believe that a greater number of pressor re-
sponses could be obtained if we could improve the method of
stimulation so that the sensory fibers might be stimulated more
strongly.
It thus appears from our experiments that the depressor effect
of weak stimuli is much more common than Vincent and Cam-
eron thought, though these observers were careful not to deny
its occurrence. Reid Hunt,'* in a recent paper, seems to have
had the same difficulty that Vincent and Cameron encountered
in obtaining the depressor effect of weak stimulation, which he
ascribed to the different frequency of stimulation they employed.
The fact that a weak stimulation of a sensory nerve causes,
as a rule, a reflex fall of blood-pressure and a strong stimulation
a reflex rise, together with the statement of Bayliss* that the
orthodox effect due to the stimulation of the depressor nerve
(nerve of Cyon®) can be converted into a rise by the action of
strychnine, led us to inquire whether a pressor response could
be obtained by strong stimulation of the depressor nerve. So far
as our results inform us, neither such a strong current as would
injure the nerve nor inductién shocks up to eighty per second
frequency could reverse the depressor response. The response
to the stimulation after injection of strychnine was sometimes
increased and sometimes decreased, but the reversal of the re-
sponse did not appear in our experiments even with a dose
which caused general convulsions on weak stimulation.
4, THE INFLUENCE OF THE FREQUENCY OF, STIMULATION UPON
VASOMOTOR REFLEXES
That the frequency of stimulation has a certain effect upon
vasomotor reactions seems to have been known to the older in-
vestigators. In 1883 Kronecker and Nicolaides” noticed that the
vasomotor centers could more easily be affected by changing the
frequency of stimulation than by changing its strength. They
write: “‘One can never attain such a strong vasoconstriction by
increasing the intensity of the stimulating current as by increas-
ing the frequency of the current of moderate ‘ntensity.”’ We
have not been able to consult the original paper of these writers.
——
- VASOMOTOR REFLEXES 365
From a reference in Ergebnisse der Physiologie by Asher,! it is
not clear whether the stimulation was directly upon the nerve
centers or reflexly through afferent nerves. —
But the credit of pointing out clearly that the frequency of
stimulation has an effect upon vasomotor reflexes must be
ascribed to Gruber.’ This writer remarks: ‘‘That summation
takes place with rapid rates of stimulation is undisputable, but
it does not seem probable where the strength is more than 400
times threshold that the phenomenon of summation can explain
the different effect obtained with these rates of 1 per two seconds
and 20 per second interruptions.’’ The similar effect of fre-
quency of stimulation was afterward proved incidentally by
Reid Hunt,!* who considers it convenient to use the infrequent
rate of stimulus to obtain a reflex fall of blood-pressure.
Our experiments on this subject have been carried out on
dogs with rates of stimuli of 1, 2, 5, 10, 20, 40, and 80 per sec-
ond upon various nerves, under chloroform and curare or under
brain compression. ‘Though our results were not so conclusive
as those obtained by Gruber (in fifteen out of forty stimulations
similar results to his were obtained), still we do not hesitate to
ascribe an important rdle to the frequency of stimulation. Ac-
cording to Martin’s” investigations, the intensity of stimula-
tion in Z-units is directly proportional to that of the current
in the primary circuit. We arranged the apparatus in such a
way as to get a current of a certain strength and one ten times
stronger as we desired. With the former current we obtained a
fall by stimulating five times per second, and a distinct rise by
stimulating ten times per second, while with the latter current
we observed a fall with the rate of stimulus one per second and
a rise with five per second stimulations. A selected record is
shown in figure 9, where one and the same nerve was stimulated
with the same intensity but with different frequency.
Much more remarkable were the rates of stimulation at which
the maximum pressor response was reached.
366 D. OGATA AND SWALE VINCENT 7
RATE OF STIMULI PER SECOND
Number of experiments whose maximum pres-
sor response was reached at the rate of
stimuli mentioned above......0......5..0.....1 O} OF Of 4 138" Mies
As is seen from the table, in 78.9 per cent the maximum re-
sponse is reached between twenty to forty per second stimula-
tion, and in one-third at the rate of forty per second. Beyond
these points the effect increased only in four cases. This phe-
nomenon may be seen a!so in figure 9.
Kronecker and Nicolaides?® observed the fact that the effect
of stimulation of the vasomotor centers increased with the fre-
quency of stimuli up to twenty to thirty per second, but not
beyond this point. Tur” also pointed out that the effect of
stimulation of the lingual nerve increased until the stimuli
reached forty per second, beyond which, however, the effect
diminished. These observations coincide fairly well with our
own.
,
5. EFFECTS UPON VASOMOTOR REFLEXES OF STIMULATING NERVE
TRUNKS OF DIFFERENT CATEGORIES (SENSORY, MOTOR, AND
MIXED NERVES) AND OF DIFFERENT SIZES
According to the investigations of some authors, different
nerves, apart altogether from the depressor nerve, respond dif-
ferently to central stimulation. Hofmann," in Nagel’s Hand-
buch, says: ‘‘There are single nerves, which for the most part
(glossopharyngeal) are depressor, and others which are exclu-
sively (splanchnic) or preponderatingly (sciatic, facial, infra-
orbital, cervical nerves) pressor.’’ Vincent and Cameron studied
the effect of stimulating the main trunk of the sciatic, as well as
its common peroneal, lateral cutaneous, and purely muscular
branches, the saphenous, median of axilla, the hypoglossal, the
glossopharyngeal, the superior laryngeal, and the vagus. But
the different nerves all produced similar or comparable results
on the blood-pressure. They were strongly tempted to the
hypothesis that an equivalent stimulation of a roughly equal
number of afferent fibers will yield similar reflexes.
~~ ee ee eee
he a deal
~ VASOMOTOR REFLEXES 367
Our experiments also have led to the conclusion that there is
no essential qualitative difference between the various nerves
subjected to stimulation (sciatic, tibial, peroneal, median, ulnar).
A possible exception may be made in the case of the saphenous.
It may be that there is a greater tendency to a fall on stimu-
lating this nerve than in the case of others. Whatever the nerve
may be, whether a purely sensory nerve, as the saphenous; a
mixed nerve, as the sciatic, peroneal, tibial, ulnar, or median,
or a purely motor nerve, such as a muscular branch of the fe-
moral, the course of response to weak and strong stimulation was
in most cases a fall and a rise as already described in section 3.
A few examples are quoted in tabular form where a selected
purely sensory nerve and a purely motor or a mixed nerve ap-
parently of the same size were stimulated in turn under entirely
similar conditions.
Sensory and motor nerves compared
BRANCH OF RIGHT
FEMORALIS
ARRANGEMENTS OF STIMULATION RIGHT SAPHENOUS
1 battery coil at 30 em. 15 seconds...) —4 mm. Hg. 0, mm. Hg.
1 battery coil at 25 em. 15 seconds...| —10 mm. Hg,| —1, +2 mm. Hg.
1 battery coil at 20 cm. 15 seconds...| —10, +2 mm.Hg.| —4, +2 mm. Hg.
1 battery coil at 15 em. 15 seconds...| +4, —12 mm. Hg.| +4, 0mm. Hg.
1 battery coil at 10 em. 15 seconds...| +6, —14mm.Hg.| +8, —2 mm. Hg.
1 battery coil at 5 em. 15 seconds...| +16, —18 mm. Hg. | +14, —12 mm. Hg.
1 battery coil at 0 cm. 15 seconds...| +22, —14 mm. Hg. | +26, —14 mm. Hg.
(From experiment 47. Under brain compression.) ;
The sign ‘—’ means a pure fall of blood-pressure, ‘+’ a pure rise, and ‘—, +’
or ‘+, —’ a mixed response, namely, a fall followed by a rise or a rise followed
by a fall, respectively.
Sensory and mixed nerves compared
ARRANGEMENTS OF STIMULATION RIGHT SAPHENOUS RIGHT PERONEAL
1 battery coil at 30 em. 15 seconds... 0 mm. Hg. 0 mm. Hg.
1 battery coil at 25 em. 15 seconds...) —6, O:mm. Hg.|.=—1, 0 mm. Hg.
1 battery coil at 20 em. 15 seconds...| +2, —10 mm.Hg.| —4, +4 mm. Hg.
1 battery coil at 15 em. 15 seconds...| +6, —12 mm. Hg.| +2, —6 mm. Hg.
1 battery coil at 10 em. 15 seconds...| +8, —12 mm. Hg. | +14, —10 mm. Hg.
1 battery coil at 5 em. 15 seconds...| +22, —10 mm. Hg. | +20, —14 mm. Hg.
1 battery coil at 0 em. 15 seconds. ..| +22, —12 mm. Hg. | +22, —20 mm. Hg.
(From experiment 47.
Under brain compression.)
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 4
368 D. OGATA AND SWALE VINCENT
With the increase of the strength of stimulus the respiratory
movements also increased, though not very markedly, and
therefore some of the falls following the rise on stronger stimu-
lation might have been more or less due to this complication.
But in the main it seems that the purely sensory nerves have
somewhat lower threshold than other kinds of nerves (fig. 10).
Whether this is due to a large number of afferent fibers con-
‘tained in the sensory nerve than in those of the other kinds of
the same size could only be decided by more numerous experi-
ments and more elaborate methods than those we have employed,
as, for example, the measurement of resistance of each nerve
and more satisfactory methods of controlling the intensity of
stimulation in each case.
in connection with the problem as to different kinds of nerves
we have studied the influence of the size of the nerve upon vaso-
motor reflexes. The hypothesis of Vincent and Cameron is
quoted at the beginning of this section. A similar problem was
taken up also by Stiles and Martin,‘? who compared the effect
of stimulating two nerve paths at the same time with that of
exciting each. by itself. They found that “stimulation of two
afferent paths at the same time has often a more marked vaso-
motor effect than the stimulation of either path alone with an
equivalent strength of current. The degree of summation was
only moderate.’’ This shows that the stimulation of a larger
number of afferent fibers will produce often a more marked
effect than that of few fibers.
We stimulated two nerves of the same category but of different
sizes separately one after another under conditions as similar as
possible, a different number of afferent fibers being assumed to be
present in the nerves of different sizes.
The results may be represented as follows, page 369.
These few examples show that the results were not very con-
clusive. We can say only so far with some confidence that when
the responses were in the same sense, i.e., when the fall or the
rise was the result of corresponding equivalent stimulations, the
reflex change of blood-pressure was on the whole more marked
with the nerve of larger size than with those of smaller size (fig.
11).
Oi ia la ae alice i eal
Mixed nerves compared with each other
VASOMOTOR REFLEXES
ARRANGEMENTS OF STIMULATION
1 battery coil at 30 cm.
1 battery coil at 25 cm.
1 battery coil at 20 cm.
1 battery coil at 15 em.
1 battery coil at 10 cm.
1 battery coil at 25 cm.
1 battery coil at 30 cm.
1 battery coil at 15 cm.
1 battery coil at 10 cm.
(From experiment 47.
Sensory nerves compared with each other
15 seconds. ..
15 seconds. ..
15 seconds. ..
15 seconds. ..
15 seconds...
15 seconds...
15, seconds...
15 seconds...
15 seconds...
RIGHT SCIATIC
0
ba
—14,
mm. Hg.
4 mm. Hg.
0 mm. Hg.
12, —22 mm. Hg.
18, —16 mm. Hg.
RIGHT SCIATIC
—2,
—6,
369
RIGHT PERONEAL
2 mm. Hg.
4 mm. Hg.
16, —18 mm. Hg.
22, —22 mm. Hg.
Under brain compression.)
ARRANGEMENTS OF STIMULATION
1 battery coil at 25 cm.
1 battery coil at 20 cm.
1 battery coil at 15 cm.
1 battery coil at 10 cm.
1 battery coil at 5 cm.
1 battery coil at O cm.
1 battery coil at 25 cm.
1 battery coil at 20 cm.
1 battery coil at 15 cm.
1 battery coil at 10 cm.
1 battery coil at 5 em.
1 battery coil at O cm.
(From experiment 48.
10 seconds...|
10 seconds...
10 seconds...
10 seconds...
10 seconds...
10 seconds...
10 seconds...
10 seconds...
10 seconds...
10 seconds...
10 seconds...
10 seconds...
RIGHT SAPHENOUS
0
5)
=18,
iho
e118;
+4,
mm. Hg.
0 mm. Hg.
0 mm. Hg.
0 mm. Hg.
0 mm. Hg.
0 mm. Hg.
RIGHT SAPHENOUS
0
a
se)
=a
+4,
+2,
mm. Hg.
0 mm. Hg.
,0 mm. Hg.
0 mm. Hg.
0 mm. Hg.
0 mm. Hg.
Under brain compression.)
0 mm.
—10, 0 mm.
—8, 4 mm.
8, —6 mm.
12, —10 mm.
RIGHT TIBIAL
—2, 2mm.
—2, 0 mm.
10, —8 mm.
20, —16 mm.
SAPHENOUS
A BRANCH OF THE RIGHT
0 mm. Hg.
es
Lg.
+2,
="
0 mm. Hg.
0 mm. Hg.
0 mm. Hg.
0 mm. Hg.
A BRANCH OF THE RIGHT
SAPHENOUS
These conclusions coincide with the experience of Stiles and
Martin and lend some support to the hypothesis of Vincent and
Cameron.
It may not be amiss to add to these conclusions that in few
cases when the stimuli were very strong the smaller nerve re-
acted more vigorously than the larger one, which phenomenon
may perhaps be explained partly by different resistances of dif-
ferent-sized nerves to currents of similar strength.
370 D. OGATA AND SWALE VINCENT
Thus, so far as our experiments go, we are inclined to conclude
provisionally that among nerves of different categories there are
no essential qualitative differences of response, and the greater
the number of afferent fibers stimulated, the more marked is
the response of blood-pressure within a limited range of strength
of stimulation.
6. VASOMOTOR REFLEXES FROM NERVE TERMINATIONS
Several investigators have stimulated the nerve terminals
instead of the nerve trunk itself.
When we apply a stimulus to a surface such as the skin we
should bear in mind that we may actually be stimulating either
the end-organs alone or these structures as well as the nerve
fibers, according to the mode of stimulation. Any physiologi-
cally appropriate stimulus, though mild, applied to the end-
organs would give rise to more highly effective impulses than
inappropriate ones. Thus the study of vasomotor reflexes in
response to stimulation of the sense organ with its most ap-
propriate stimulus is highly desirable. But even with other
kinds of stimulation we may learn much that is valuable, be-
cause any stimulus which plays a part in our normal daily life
comes usually through the end-organs on the outer or inner
surface of the body, and not by way of exposed nerve trunks,
as in the foregoing experiments.
The skin, the mucous membrane of the nose, muscles, the in-
testine, and other abdominal organs were employed frequently
by previous investigators. We have selected the skin, muscles,
and the intestine as representative of regions containing differ-
ent modes of nerve endings. The stimuli used were mechanical
(incision, scratching, pinching, kneading), thermal (hot or boil-
ing water and cold water or lumps of ice), chemical (10 per cent
solution of sulphuric acid), and electrical (induction shocks of
various strengths). The animals (dogs) were under ether,
chloroform, or brain compression.
The results of a first series are presented in the following
tables:
~
VASOMOTOR REFLEXES
Mechanical stimulation of the skin
- : STIMULATED r
ANESTHESIA Scere MODE OF STIMULATION
L. inn. thigh | Pinching
ithersss.i-. ages: | R. inn. thigh | Scratching
R. inn. thigh | Incision
| L. inn. thigh | Pinching
R. inn. thigh | Scratching
| Abdomen
Chloroform....... Over saphen- | Incision
. ous, ulnar,
| and sciatic
[| nerves
Brain compres- { L. inn. thigh | Scratching
Gl ey Se eee 8 R. inn. thigh | Incision
SNAEPCRR AEN: 5s SU Sos sts cus ORE e SoMee Ques eee via See
Thermal stimulation of the skin
L. inn. thigh | 65°C.—boiling water
i A aie { L. inn. thigh | Ice—15°C. water
L. inn. thigh | Boiling water
Chloroform....... { meee jing ice
Brain compres- L. inn. thigh | Boiling water
BLODLEE satiate. L. inn. thigh | Ice
POMEL See Sch ath ich Reta Bini Sys oie 2 ate alk sath Sita cues (ark ARs hin 0 0s
Electrical stimulation of the skin
UES cad ote separ L. inn. thigh | Strong induction shocks
Chloroform......... L. inn. thigh | Strong induction shocks
Brain compression..| L. inn. thigh | Strong induction shocks
371
REFLEX RESPONSE
OF BLOOD-PRESSURE
chet | Fall | Rise
0 2 0
2 5 0
0 6 0
1 1 0
2 3 0
3 4 0
0 2 0
0 3 0
8 | 26 0
0 5 0
3 0 0
2 3 0
0 1 0
0 1 1
2 0 0
Fie a) 1
8 5 0
4 2 0
4 2 0
8 4 D. OGATA AND SWALE VINCENT
As is clear from the tables, almost every stimulation in this
series, produced a reflex fall of blood-pressure, and no signifi-
cant qualitative difference is observable either with different
modes of stimulation or with different methods of anaesthesia
or with different portions of the skin.
That this statement is applicable almost without any modifi-
cation to the results of stimulation of muscles and intestine
will immediately be understood from the tables on page 373.
Thus it is fairly clear that the stimulation of nerve terminals
in the skin, muscles, and the intestine produces usually a reflex
fall of blood-pressure, as was reported by Vincent and Cameron.
But the threshold of stimulation for the nerve-terminals in
the skin is very much higher than that for the exposed nerve-
trunks. Thus a stimulus which is to be reckoned a strong one
for the exposed nerve-trunk is to be considered a weak one for
the surface of this skin. This fact explains the previously de-
scribed results. So far the effects have all been those of a weak
stimulation, namely, a fall of the blood-pressure.
If, now, we take steps to secure a considerably greater amount
of stimulation by simultaneous scratching of large areas in dif-
ferent regions, it is not difficult to satisfy oneself that the same
general law applies for the nerve-terminals as for one exposed
nerve-trunk. Thus, if we scratch a limited area with a mod-
erate degree of vigor, we get a fall, while more violent applica-
tion of the instruments to a large area, will give a rise (fig. 19).
In the last section we compared the effects of stimulating two
nerves of the same category but of different sizes, and showed
that the nerves of greater size usually surpass those of smaller
size in their power of evoking vasomotor reflexes, and referred
to Vincent and Cameron’s hypothesis that the number of af-
ferent nerve fibers is an important factor. In our stimulation
of nerve endings, as a rule, we could only apply the stimulations
to a small portion of the surface. Now the nerve fibers spread
widely from the nerve trunk, and the stimulation of a nerve
would be equivalent to the stimulation of the entire surface to
which the nerve is distributed. In other words, the stimula-
tion of a small portion, e.g., of the skin, corresponds to that of a
ee ee rr SS
ANESTHESIA
Chloroform.......
Brain compres-
BION: ELLER oe
VASOMOTOR REFLEXES
Mechanical stimulation of muscles
STIMULATED MUSCLE
R. add. mag.
R. sartor.
R. sartor.
L. semitend.
R
MODE OF
STIMULATION
Scratching
Scratching
Kneading
Scratching
. add. mag. or 1 semitend. | Scratching
Stimulation of the intestine
Small intestine
Interior surface
intestine
Interior surface
intestine
Interior surface
intestine
Small intestine
Small intestine
Small intestine
Interior surface
intestine
Interior surface
intestine
Interior surface
intestine
Interior surface
intestine
of
of
of
of
of
of
of
small
small
small
small
small
small
small
Kneading
Induction shocks
Pinching
'
Boiling water ap-
plied
Distension
Kneading
Kneading
Induction shocks
Scratching
Boiling water ap-
plied
Ice piece applied
373
REFLEX
CHANGE OF
BLOOD-
PRESSURE
0| 1,0
0| 2)0
0} 3/0
3| 0,0
4/ 0|0
7 | 15) 0
0; 3,0
1) 01.0
0; 1,0
£7). 28
0; 40
0; 21
0; 4,0
Of 22
0/ 1/0
0; 1,0
DA iaied 4
2 | 20) 3
374 D. OGATA AND SWALE VINCENT _
small number of sensory nerve fibers. If the stimulation of a
few fibers be the equivalent of a weak current, the fall of blood-
pressure caused by stimulating the nerve terminals may be
ascribed to the fact that we are stimulating only a few fibers.
Under morphia and curare a rise of blood-pressure is more
easily obtained than a fall on stimulation of nerve-endings.
But under morphia at any rate it is not difficult to obtain a rise
with a strong stimulus and a fall with a weak one (fig. 19).
Gaskell’s’? discovery that in mammals ‘‘A large dose of curare
will remove both the contraction of the muscle and the dilatation
of its blood-vessels upon stimulation of the nerve,’’ may possi-
bly account for the greater tendency towards a rise when the
animal is under this drug.
The paralysis of the vasodilator nerves by curare seems to
necessitate the taking of certain precautions in interpretation
of the results of experiments.
It seems probable that if it were found possible to increase
very considerably the energy and extent of the stimulation in
the cases of kneading of muscle and of the intestine, we should
have to record a rise of pressure instead of the fall with which
we are familiar.
7. THE INFLUENCE OF THE DUCTLESS GLANDS UPON VASOMOTOR
REFLEXES
The extracts of some of the ductless glands (adrenal body,
thyroid, and pituitary) have been alleged to affect the vaso-
motor irritability on one way or another.!7:43.12.27.28 Since the
results of the previous investigations are not conclusive, we
thought it might be worth while to investigate the matter
again. It is to be feared that our experiments are not much
more convincing than those of previous workers on this subject.
The change of blood-pressure (augmentation or diminution)
due to the injection of the extracts of these glands is an undesir-
able complication. In cases where an augmented blood-pres-
sure is the result, as with adrenin and pituitrin, the decreased
pressor reaction to the stimulation of a nerve may most properly
be ascribed to the diminished response of the already more or
VASOMOTOR REFLEXES 375
less contracted blood-vessels, or possibly to the additional con-
traction of the blood-vessels of the small areas other than those
previously affected by the drug.
But the comparison of the vasomotor reflexes before and after
the injection of adrenin (‘‘adrenalin” Parke, Davis & Co.) seems
to show that the pressor reflex is slightly decreased in the latter
case. The results of Hoskins and Rowley were similar and
more definite. The elimination of the function of the supra-
renal glands by tying them off gave no clear results.*
The injection of thyroidin (Parke, Davis & Co.) and the extir-
pation of both thyroid glands do not appear to have any distinct
influence upon vasomotor reflexes.
Pituitrin (Parke, Davis & Co., surgical) showed scarcely any
significant results.
All these experiments were performed on dogs under brain
compression for the purpose of excluding any influences from the
increased respiratory movements and those of anaesthetics and
other drugs.
8. THE QUESTION AS TO WHICH VASCULAR AREAS ARE CON-
STRICTED OR DILATED ON CENTRAL STIMULATION OF
SOMATIC NERVES
The fall of blood-pressure produced by stimulation of the de-
pressor nerve is effected chiefly by dilatation of the splanchnic
area,” though, as Bayliss has shown, the vessels of the limbs,
head, and neck also partake in the relaxation. The latter
writer showed also that the rise of blood-pressure on stimula-
tion of the central stump of the splanchnic (?) nerve was, for
the most part, due to the constriction in the splanchnic area.
The reflex rise of blood-pressure due to the stimulation of the
*That is to say, when the nerves to the limb are intact. In the denervated
limb there is a very important difference according to whether or no the ad-
renal bodies are eliminated. Mr. Pearlman and myself have recently found
that when the central end of the sciatic is stimulated in such a way as to give
a pressor response the intact limb follows passively the blood-pressure while the
denervated limb constricts. After removal of the adrenal bodies the dener-
vated limb also dilates. These results are explained more fully in a paper about
to be published in ‘Endocrinology.’—S. V.
376 D. OGATA AND SWALE VINCENT
sensory nerves of the skin, too, depends mainly on the constric-
tion of the blood-vessels in the same area,** and Hofmann"
writes: ‘‘The rise of blood-pressure on stimulation of sensory
nerves is produced by the constriction of the blood-vessels of
abdominal organs as in asphyxia. At the same time the blood-
vessels of the brain, skin, and muscles dilate, as a rule, and an
increase of the volume of limbs takes place.”” Thus the splanch-
nic area plays a principal part in the reflex changes of blood-
pressure on stimulation of the somatic as well as the splanchnic
nerves.
We have made some experiments on this point, and can con-
firm the above statements. The dogs had both vagi cut and
were under morphia and curare or brain compression, and the
sciatic or saphenous nerve was stimulated with induction shocks.
The volume changes of the limbs (hind and fore) and of the
abdominal organs (small intestine, kidney, and spleen) were
recorded.
The rise of blood-pressure, when sufficiently high, was always
accompanied by a remarkable diminution of the volumes of
abdominal organs and a pronounced dilatation of limbs (fig. 18).
The pronounced fall of blood-pressure with weak stimulation
when the animal was under brain compression was seen to be
accompanied by a distinct increase of the volume of the intes-
tine. Thus it appears clear that a reflex rise and fall of blood-
pressure on stimulation of a somatic nerve (sciatic, saphenous)
is brought about chiefly by constriction and dilatation of the
blood-vessels in the splanchnic area.
9. SUMMARY
1. In dogs under ether or chloroform, stimulation of sensory
nerves (saphenous, tibial, peroneal, sciatic, ulnar, and median)
causes usually increased respiratory movements when narcosis is
not profound or curare is not employed. These increased move-
ments produce a fall of blood-pressure, and when they are very
violent, one cannot obtain any pressor reflex even with a strong
stimulation. When much increased respiratory movements are
prevented by very deep narcosis or brain compression, fall of
VASOMOTOR REFLEXES Yi |
blood-pressure due to this cause does not occur. Mechanical
interference with the circulation as a result of the increased
movements of the thoracic walls seems to be the main cause of
this fall, since it can be eliminated by opening the thorax.
When this complication is not taken into careful consid-
eration the results of vasomotor experiments are liable to be
misinterpreted.
2. In dogs under ether, chloroform, or brain compression, a
weak stimulation of the central end of the cut nerves (sciatic,
saphenous, tibial, peroneal, median, ulnar, and vagus) pro-
duces usually a fall and a strong stimulation, a rise of blood-
pressure. With a gradual increase of the strength of stimulus
up from the threshold, the reflex fall of blood-pressure first
increases, then decreases, and gradually becomes converted
into a rise, passing through a neutral point. We have failed to
obtain a pressor effect by the strongest stimulation of the de-
pressor nerve of Cyon.
3. The frequency of stimulation has an effect upon vasomotor
reflexes. With a rapid rate of stimulation a rise is obtained
and with a slow rate of stimulation in many cases a fall of blood-
pressure. Of the different rates of stimulation we employed
(one to eighty per second), the maximum pressor response is
reached at twenty to forty per second.
4. No essential qualitative difference was found among vari-
ous nerves (sciatic, tibial, peroneal, median, ulnar, branch of
femoral nerve) subjected to stimulation. The saphenous nerve
has a greater tendency to give a fall than those above mentioned.
A purely sensory nerve seems to have a somewhat lower thresh-
old than other kinds of nerves. Between nerves of the same
category but of different sizes, the larger one produces usually
a more marked response within a limited range of the strength
of stimulation.
5. When the animal is under ether, chloroform, or brain com-
pression, stimulation (mechanical, thermal, chemical, and elec-
trical) of nerve terminations, such as those in the skin, muscles,
and the intestine, causes a fall of blood-pressure in the great
majority of cases, but violent or extensive stimulations of the
378 D. OGATA AND SWALE VINCENT
skin produce a rise. Under morphia and curare, on the con-
trary, a rise is a usual response, due clearly to a specific pharmaco-
dynamical influence of these drugs. But under morphia a weak
stimulus will produce a fall.
6. The influence of the ductless glands (adrenal, thyroid, and
pituitary) upon vasomotor reflexes is not clear.* The injection
of adrenin, thyroidin, and pituitrin and tying off or extirpation
of the glands produced in our experiments no distinct effect.
7. The reflex change (fall or rise) of blood-pressure on stimu-
lation of the somatic nerves (sciatic, saphenous) is produced
chiefly by the dilatation or constriction of the blood-vessels in
the splanchnic area, as in the cases of the stimulation of the
splanchnic and the depressor nerve.
*See footnote, page 375.
Oowrwnds
© oO NI
10
it
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
VASOMOTOR REFLEXES 379
BIBLIOGRAPHY
Asner, L. 1902 Ergebnisse der Physiologie, Jg. 1, Abt. II.
Bayuiss, W. M. 1906 Ergebnisse der Physiologie Jg. 5, S. 319.
1908 Proc. Roy. Soc., B., vol. 80, p. 353.
BRUNTON AND TUNICLIFFE 1894 Journ. Physiol., vol. 7, p. 373.
Cron, E. pe 1900 In Richet’s Dictionnaire de Physiologie, p. 774.
DomitrrEenKo, L. F. 1912 Dissert., Odesse, p. 312. (Physiol. Abstracts,
1917, vol. 2, p. 30).
GaskELL, W. H. 1916 The involuntary nervous system. p. 90.
Grouper, C.M. 1907 Amer. Journ. Physiol., vol. 42, p. 214.
GruBer, C. M., anp Kretscumer, O. 8S. 1918 Amer. Jour. Physiol., vol.
46, p. 222.
GrUTzNER AND HripENHAIN 1878 Arch. f. d. gesammt. Physiol., Bd. 16,
(quoted from Stiles and Martin) .*4
Hormann, F.B. 1909 In Nagel’s Handbuch der Physiologie des Menschen,
Boel Sais:
Hoskins, R. G., anD Rowtey, W.N. 1915 Amer. Journ. Physiol., vol. 37,
fie ak 7A le
Howe..t, W. H. 1915 Text-book of physiology, p. 688.
Hunt, R. 1895 Journ. Physiol., vol. 18, p. 406.
1918 Amer. Journ. Physiol., vol. 45, p. 197.
1918 Amer. Journ. Physiol., vol. 45, p. 231.
Kepinow 1912 Arch. f. exper. Pathol., Bd. 67, 8. 247 (Zentralbl. f. Physiol.,
1913, Bd. 27, S. 129).
KLEEN, 1887 Skand. Arch. f. Physiol., Bd. 1, 8. 247 (quoted from Reid
Hunt).
Kwnoitu 1885 Sitz. Acad. Wiss., Wien, Math. Naturur. Kl., Bd. 92 (iii),
S. 448 (quoted from Reid Hunt). ’
Kronecker, H., anp Nicotarpes, R. 1883 Du Bois-Reymonds Arch.
(quoted from L. Asher ).?
Lupwia anp Cyon 1866 Ber. d. Siichs. ges. d. Wissenschaften (quoted
from Bayliss, Journ. Physiol., 1893, vol. 14, p. 303).
Martin, E. G. 1912 The measurement of induction shocks, p. 34.
Martin, E. G., anp Lacey, W. H. 1914 Amer. Journ. Physiol., vol. 33,
p. 212.
Martin, E. G., anp MenpenHALL, W. L. 1915 Amer. Journ. Physiol.,
vol. 38, p. 98.
Martin, E. G., anp Stizes, P. G. 1914 Amer. Journ. Physiol., vol. 34.
1916 Amer. Journ. Physiol., vol. 40, p. 194.
OswaLp, A. 1916 Arch. f. d. gesammte Physiol., Bd. 164, S. 506-582.
(Physiol. Abstracts).
1916 Arch. f. d. gesammte Physiol., Bd. 166, S. 169-200 (Physiol.
Abstracts).
Porter, W. T. 1908 Boston Med. and Surg. Journ., vol. 158.
1910 Amer. Journ. Physiol., vol. 27, p. 276.
1915 Journ. Physiol., vol. 36, p. 418.
380
42
43
D. OGATA AND SWALE VINCENT
Porter, W. T., AND Storey, T. A., 1907 Amer. Journ. Physiol., vol. 18,
Peale ee AND Turner, A. H. 1916 Amer. Journ. Physiol., vol. 39,
idea en AND QuinBy, W. C. 1908 Amer. Journ. Physiol., vol. 20,
ep AND BILLINGSLEY, P. R. 1916 Amer. Journ. Physiol., vol.
42, p. 16.
1917 Amer. Journ. Physiol., vol. 42, p. 16.
Sotmann, T., ano Piucuer, J. D. 1910 Amer. Journ. Physiol. xxvi,
p. 233.
Sraruine, E. H. 1915 Principles of human physiol., p. 1006.
Stewart, G. N., anp Larrer, W. B. 1913 Arch. Int. Med., vol. 11, p.
365.
Stites, P. G., anp Martin, E. G. 1915 Amer. Journ. Physiol., vol. 37,
p. 102.
TicmrsTeDT, R. 18 3 Lehrbuch der Physiologie des Kreislaufs (quoted
from Asher,? p. 349).
Tur 1898 Hermann’s Jahresb., S. 61 (quoted from Nagel’s Handbuch).
VincEnT, S., AND Cameron, A. T. 1915 Quart. Journ. Exper. Physiol.,
vol. 9, p. 48.
beget Sake
he
a
~
aie, os ‘
PP Pe
ee a vs
Re Clk Sk
(.
ws
et ee
FUL .
Neat
wa
ea tat.
1. “BY apes bs | shy case hs
: Te. tie de
Vy. ‘ Deiat |
ee
" wo gfe Sh aeeley TE “d
BOP ae Pa
tagae
' 7 = > i
basthovarty we sue
wavny ip
{
‘
pate
PLATE 1
EXPLANATION OF FIGURES
Fig. 1 The effect of increased respiratory movements upon vasomotor re-
flexes. Bitch. 9 kilos. 10/5/1918. Ether. The left sciatic nerve was stimu-
lated at intervals, the stimulus increasing from left to right. Upper curve,
respiratory movements. Lower curve, blood-pressure. Base line is that of
zero pressure, with periods of stimulation. The height of the blood-pressure
in mm. Hg. is indicated by the em. measured out and numbered. Time in sec-
onds. For further explanation see text.
Fig. 2 Increased respiratory movements prevented by very deep narcosis.
Same bitch as in figure 1. For explanation see text.
Fig. 3 Effect of increased respiratory movements upon blood-pressure.
Bitch. 14 kilos. 30/5/1918. Ether. Thorax intact. The left sciatic nerve
was stimulated. The result is a marked fall of blood-pressure.
Fig. 4 Effect of increased respiratory movements upon blood-pressure pre-
vented by opening the thorax. Same bitch as in figure 83. Thorax wide open in
the middle line. The same nerve was stimulated with the same strength of
stimulus as in figure 3. The result is a marked rise of blood-pressure.
iS)
~
bo
VASOMOTOR REFLEXES
D. OGATA AND SWALE VINCENT
mt Hi mm ' wits mi i m
fi Ne f “enh matte nt ly , a A Hy a ue m |
HA rN il it
80
ee
bt, Fesnia ' aint sale ahead Ly |
i ie yet Hemi dei Ea
Sala Naty at a NT a A a UU dt Dd i \ va lal g)
NNT iii INCOME TCTNNT ONO TS 11) 9 FATPIFIANITEN TORT AROMAS DT Da
— _@
L
Tint |
bs obisigiclnalasinstlelutiaau cua diw aianiann, SSOSTUETTTAITTGSTUCUOQONUOUUOTECUUOTONUOUUICNTUOUONOTONOUOTOTOCIUTUONUCOTOOTUNNTON OMT TCU MTTT TIM TT
——— a=.
id
PLATE 1
LUELINUATOUFODURONOUUESTIUCUIVIUITON ENON
IPTUUTITIUUTTIUTVOUUTUNNUWUDIVUTUCUUOIISUNOGUEERE
PLATE 2
EXPLANATION OF FIGURES
Fig. 5 A marked fall of blood-pressure which is apparently not due to in-
creased respiratory movements. Dog. 12 kilos. 9/5/1918. Ether. The left
sciatic nerve was stimulated.
Fig. 6 An example of vasomotor reflexes upon weak (left) and strong (right)
stimulations. Same bitch as in figure 3. Thorax was very wide open in the
middle line to eliminate the influence from increased respiratory movements.
A weak stimulation caused a fall and a strong stimulation caused a rise of blood-
pressure.
Fig. 7 Vasomotor reflexes under chloroform. Bitch. 7 kilos. 28/5/1918.
Thorax wide open. A weak stimulation produced a fall (left) and a strong
stimulation produced a rise (right) of blood-pressure.
Fig. 8 Effects of weak and strong stimuli, respectively, under brain compres-
sion. Dog. 18 kilos. 15/8/1918. Brain compression and artificial respira-
tion. Right ulnar nerve stimulated. The fall of blood-pressure increased at
first with the development of the strength of stimulus and then passed over to a
rise crossing a neutral point. ,
Fig. 9 Effect of frequency of stimulation upon vasomotor reflexes. Bitch.
15 kilos. 18/7/1918. Chloroform and curare. Right saphenous nerve was
stimulated. The frequency employed 1, 2, 5, 10, 20, 40, and 80 per second, re-
spectively, from left to right. The one per second stimulation caused a fall,
the two per second stimulation showed practically no effect, and the other stimu-
lations produced a rise. The maximum pressor response was reached at forty
per second stimulation in this case.
Fig. 10 Stimulation of a sensory (saphenous) and a motor (a branch of the
femoral) nerve. Dog. 9 kilos. 10/10/1918. Brain compression and artificial
respiration. Stimulation of a sensory nerve gave a more pronounced fall than
that of a motor nerve.
Fig. 11 Stimulation of nerves of the same category but of different sizes.
Same dog as in figure 10. The stimulation of a larger nerve (sciatic) produced
a more marked response than that of a smaller nerve (peronea!).
386
VASOMOTOR REFLEXES
D. OGATA AND SWALE VINCENT
HUUESUUUUUUUNCUUEUOVODUCOVEUSUNUUCETUUNTIUICUUUS SUSUNCTENUCTTSUTIVUN TT RUST TONTTTTN|] UU UUUUUTUUTUNvEUUOWOSHOUUCLITTNCUTCD [NTT TIE
AAA AAA AV AAA AAA VY VV VAAL
AOTARATANTIVV HURT TR RAD LALLA
cement |
“Hiya
CUTNLAVUVEOUUUUOUUBUDUNDOUUCUDAUOTUSUHUECUOUECUUUVONICUUNNUUOUNNTNUIUSVETUUSTUCUUNUNTITNOUITISHNUUEUEUUTSNTICITY I TUCE EN
: »
iil wl iti
ill
5
IUVOVI OUUUUCUUNUUTU VE CVUVAUUUUUCUNDUDTUTEITIUUUUCUECUUUUCQUUUCNIIUENUUEUOUCNUCOUNUNUNUUSUUURIUCNIIUIUUCUEUUNUNEULONIL,
iN
, AN) ui | ne a aaa : mM \ 4 m
ee cance es cn ll ‘inmgnasl hy iin uty Tl HN Yn eT —" ine
io a
— 4 ———— e. ga ———— 1 — — —o L =
i ‘ ! n i 1 Lili
OOTPTETETTR TTT ERTOT PUTT OTT ITTOTOOTOTENTTTTTT TETTCTT TN NURUTTTRTTOTTITITUPOOTUNTTTTTTTITTTCCOTTN OT OTTTTTTTTTTTOTTTOTOTTTTTTTTTTTRTTeTTaTaTOTTaTTTTTCCTTiTTTTTOTONTTNUCTCTTTTTTTTOT ONT OTTTTTTTTINVTD OTONNTTOITTTICOTTLTOCNCKUTUTTOTITUTTIAKRUN UNUCACOTICUTULNTUDOCOVOURULCLUUORUOUCMCONCOOUITUUUUTUON USE OUNUITTTITOO TON EMnnEN COOITOTITinee® CVUNUORTUIT Tran COUHNCURNTTULULDIUCOUCECDSERONCOOTeLIIULCrOanoR FUOUUTUOVUTTOTONVCUOUNMUNENUTNEDONUUGy
PLATE 3
EXPLANATION OF FIGURES
Fig. 12 Seratching of the skin. Dog. 12kilos. 9/5/1918.' Ether. A marked
fall of blood-pressure. Respiratory movements practically unaffected.
Fig. 13 Application of heat (boiling water) on the skin. Dog. 10 kilos.
25/10/1918. Brain compression and artificial respiration. A fairly marked fall
of blood-pressure.
Fig. 14 Electrical (strong) stimulation of the skin. Dog. 11 kilos.
14/5/1918. Ether. A very marked fall of blood-pressure. Respiratory move-
ments affected very slightly. .
Fig. 15 Scratching of muscle (right sartorius). Bitch. 10 kilos. 22/10/1918.
Ether. a 4)
nite 2 abs ,) aes
’
’ evs Say Be i '
Wire mda
sy
+)
‘ota aT etsy fu
: ay
a TT ak yah v4 ora®
wi Ait Sika i
NAL OF COMP: esuenox0en vou. 30, No. 5
Resumen por el autor, James Stuart Plant.
Instituto Wistar de Anatomia y Biologia.
Factores que influyen en el comportamiento del cerebro de la
rata albina en el liquido de Miiller.
El cerebro de la rata albina sufre un cambio tipico cuando se
‘‘fija”’ en el liquido de Miiller, compuesto de 2 por ciento de bi-
cromato potdsico y 10 por ciento de sulfato sédico. El cerebro
aumenta rapidamente de peso, al que sigue una pérdida de peso
lenta y continua, hasta que al cabo de setenta y cinco dias, al
pesar el cerebro se comprueba que pesa de 20 a 30 por ciento mas
que el cerebro fresco. 1. La edad es la principal condicién que
influye en esta reaccién del liquido de Miller. Los cerebros de
ratas viejas aumentan mds en peso y retienen este aumento
durante los setenta y cinco dias. 2. El peso inicial del cerebro,
0 sea su tamafio, es la condicién de mas importancia después de
la apuntada mas arriba. Del mismo modo que con los cerebros
de la misma edad, los cerebros mds ligeros aumentan mds de
peso durante la primera parte de su permanencia en el liquido
de Miller. Esta diferencia disminuye gradualmente y desa-
parece al cabo de los setenta y cinco dias. 3. La edad pr6xi-
mamente semejante (entre ciertos limites) tiene casi el mismo
valor determinante que la igualdad de edad. 4. El sexo es un
factor sin importancia, del mismo modo que la composicién
hereditaria (relacién dentro de un tronco determinado).
Translation by José F. Nonidez
Carnegie Institution of Washington
AUTHOR'S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, JUNE 30
FACTORS INFLUENCING THE BEHAVIOR OF THE
BRAIN OF THE ALBINO RAT IN MULLER’S
FLUID -
JAMES STUART PLANT
Neurological Laboratory of The Wistar Institute of Anatomy and Biology
The brain of the albino rat, placed for a period in Miiller’s
fluid, exhibits a typical change. In the course of time it not
only hardens, but also markedly increases in weight. There is a
rapid increase to a maximum in about one week’s time, after
which there is a slow, steady loss until the seventy-five day
weighing, at which time the brain weighs 20 to 30 per cent more
than when fresh. It was thought that changes in this typical
curve might be induced in the brains of rats previously anes-
thetized for prolonged periods, and it was hoped that this cri-
terion would be more delicate than the microscopic or analytic
tests which had, so far, failed to demonstrate a change. The
work was done at The Wistar Institute of Anatomy during the
academic year 1913-1914.
PLAN
The main question of the effect of the anesthetic on the typ-
ical curve remains unanswered. From the start, however, it
was recognized that various factors influenced the reaction of
‘control’ brains to Miiller’s fluid—factors which are inherent in
the material. It is these factors and their influence on the reac-
tion with which the present paper deals.
PROCEDURE
The. brains studied belonged to ‘stock’ albino rats. The
animals were killed with ether and the brains quickly removed
with every care not to damage them. They were immediately
weighed and then suspended in 50 ec. of Miiller’s fluid. The
411
412 JAMES STUART PLANT
whole was kept in a black cardboard case in a dark closet.
Subsequent weighings were as follows. The brain was removed
from the solution and placed for about ten seconds on a dry
piece of filter-paper. The string by which it had been suspended
was during this time removed. The brain was then placed on
a watch-glass and immediately weighed. It was returned to the
Miiller’s fluid as quickly as possible. The watch-glass was then
weighed. Reweighings were carried out at 24 hours, and at 7,
14, 30, and 75 days after killing the rat. On completion of the
weighings the percentage of water in the brain at the final weigh-
mg was determined. This procedure involved placing the
brain, immediately after its last weighing, in a small glass vial
of known weight. This was kept in a drying oven (temperature,
97°) for one week. On removal, the vial was cooled in a desic-
cator at room temperature and weighed.
The Miiller’s fluid used was made up in 1000 ce. lots. To
25 grams of potassium bichromate c. p. and 10 grams of sodium
sulphate c. p. was added 1000 cc. of distilled water. Time was
given for dissolving the salts and, after thorough agitation, the
solution was divided into two-500 ec. lots and kept for one month
before being used. In every instance ‘pairs’ of brains were
fixed in fluid from the same bottle. The Miiller’s fluid was
always kept in a dark closet.
No attempt was made to control the temperature during the
reaction of fixation other than that all specimens were kept in the
same dark closet at room temperature. Thus the results may be
considered as comparable.
Necessarily our original results are in fechas of absolute weights
and absolute gains. In the presence of so diverse initial weights
it seemed, however, best to state all gains in weight as percent-
ages of the original weight. This makes the data comparable.
Corresponding to this, all statements of the relation of one
brain’s gain to that of another are in terms of a percentage of
the percentages of gain of the heavier brain. This leads to
higher figures, in the relations of the gains, than would be the
case were the actual differences between the absolute gains
stated.
ALBINO RAT BRAIN IN MULLER’S FLUID ALS
OBSERVATIONS
If we consider the brains of pairs (rats of the same age, sex,
and litter, i.e., as similar as possible), there appears a very dis-
tinct tendency for the brains of older rats to gain more in Miiller’s
fluid than do those of the younger rats. Table 1 presents fifteen
pairs arranged on the basis of their age. In ten of the fourteen
possible comparisons the brains of older rats gain more in twenty-
four hours. Also in ten of the comparisons the older brains
TABLE 1
Percentages of gain of pairs of albino rat brains arranged according to age and
weighed at intervals from twenty-four hours to seventy-five days
AVERAGE PERCENTAGES OF GAIN OF BOTH BRAINS
ese eh IN MULLERS FLUID
SEX AGE
1 2 24 hours 7 days 14 days 30 days 75 days
days grams grams
°) 52 1.465 1.368 19.0 28.2 26.2 23.5 22.9
°) 55 1.615 1.578 20.8 o2.1 28.1 26.8 25.4
of 57 1757 1.671 20.6 31.2 26.0 25.9 25.7
of 59 1.635 1.519 21.2 Bout 29.9 27.0 26.7
eh 61 1.834 1D 18.0 32.2 28.0 25.9 25.0
g 61 1.699 1.581 20.0 29.5 21.0 24.1 23.0
fof 62 1.490 1.477 18.6 31.8 28.3 25.3 24.4
fof 62 1.662 1.656 19.3 28.2 24.6 23.1 22.4
g 62 1.497 1.496 20.5 31.7 29.1 25.7 25.3
of 62 1.831 1.699 20.8 32.8 30.8 28.0 7.8
°) 64 1.677 1.606 20.9 32.0 28.6 26.2 250
fof 67 1.651 1.587 21.7 32.6 29.7 26.6 26.4
°) 72 ‘1.610 1.492 27.6 33.6 30.7 27.9 26.7
of 160 1.791 1.752 25.1 37.2 35.8 32.8 32.2
ey 218 2.008 1.824 28.3 40.0 38 .6 315) 47/ 34.6
gain more in seventy-five days, though this does not in every
case involve the same comparisons as were favorable to the
older rats at the twenty-four hour weighing. Table 2 presents
a summary of the data of table 1. The averaged figures for
the youngest three pairs and for the oldest three pairs are given.
The data show that age is a very important factor in the reac-
tion of the brain of the albino rat to Miiller’s fluid.
Brain weight increases as a function of age, and there exists
between these two characters a very high coefficient of correla-
414 JAMES STUART PLANT
TABLE 2
Percentages of gain of averaged entries of table 1
AVERAGE PERCENTAGES OF GAIN IN
INITIAL WEIGHT M@LLERS FLUID
PAIRS AVERAGED AGE
24 7 14 30 75
5
1 < hours | days days days days
days .| grams | grams
First three entries......... 55 | 1.612) 1.539) 20.1 |) 30.5" | 2628) | 25 4a
Last three entries........- 150" |* £803) 17689) 27-0") 3720 35.07) Sane
tion. If we make a comparison of the individual brains of the
respective pairs involved in table 1, however, we may study the
effect of initial brain weight in animals of like age. The data
are given in table 3. In place of the percentage of gain of the
lighter brain there is entered at the several columns marked
‘Per cent deviation of 2’—under ‘Time in Miller’s fluid’—only
the relation of that percentage to the percentage of gain of the
heavier brain. Of the fifteen pairs involved, it will be noted
that in twelve the lighter brain gains more in the first twenty-
four hours (represented by a + in the second column). Also in
twelve of the fifteen pairs, thé heavier brain later ‘catches up’
—that is, the relative gain of the heavier brain is greater at
seventy-five days than at one day. ‘This phenomenon is clearly
demonstrated in the ‘averages’ at the bottom of the table. The
results may be summarized as follows:
1 DAY 7 pars | 14 pars | 30 pays | 75 pays
Average difference.................. +3.9} 41.5] +1.3) 41.2} +0.2
Standard) deviation.....-...-2 one +6.1 =—-3-9)| = 50),| 522) Soe
Ul
Thus it appears that the lighter brain gains more in the early
part of the stay in Miiller’s fluid, but that this difference prac-
tically disappears at the seventy-five-day weighing. It is to be
noted that the standard deviations are lowest at the seven-day
weighing. This seems to represent the period of least indi-
vidual variation. As this represents the time of maximum
increase, we may consider that as the more stable period in the
a
ALBINO RAT BRAIN IN MULLER’S FLUID
415
curves and think of the rise and fall in percentage of increase as
periods more subject to individual variation.
While increasing age presupposes increase in brain weight, it is
apparent that these two
age and brain weight—act as oppos-
ing factors in the determination of the reaction of the brain to
Miiller’s fluid. That is, the older brains (these are heavier)
TABLE 3
The effect of initial brain weight—albino rat—on the percentage of gain of paired
brains.
INITIAL
WEIGHT
Pairs arranged according to age.
Deviation of the lighter brain
TIME IN MULLERS FLUID. PERCENTAGES OF GAIN
eg Reem 24 hours 7 days 14 days 30 days 75 days
1 2 Per Per Per Per Per
cent de- cent de- cent de- cent de- cent de-
viation viation viation viation viation
of 2 of 2 of 2 of 2 of 2
days | grams | grams
Q 52 |1.465)1.368) 19.3)— 2.8/27.9] +2.1/26.8) —4.2/24.1!|— 4.9/23.5|— 5.7
Q 55 |1.615]/1.578} 20.4)+- 3.6)31.7| +2.5/27.7| +2.6125.8)/4 8.2/25.1/4+ 2.1
rot 57 |1.757/1.671) 19.5)4+11.1/80.5) +4.1/25.5} +4.1/25.2/+ 5.1/24.8/4+ 6.8
ot 59 |1.635)1.519) 20.5)+ 6.5/33.0) +0.2/30.3) —2.2/27.4;— 2.4/27.2|— 4.0
Q 61 |1.699/1.581) 18.8)4+13.3)/28.7| +6.0/26.2) +8.8)23.3}/+ 7.0121.6/+12.9
fof 61 |1.834!1.715) 17.9|+ 1.3/31.5| +4.2/27.5| +4.1125.4/+ 3.8/24.7/4+ 1.8
of 62 |1.490/1.477| 18.8)}— 1.8)382.4| —3.4/29.3) —7.2/26.3/— 7.3/25.0/— 4.8
Q 62 |1.497|1.496) 20.5/+ 0.3/31.4) +1.7/28.7| +2.1/25.9)— 0.7/25.1/4+ 2.1
of 62 |1.662/1.656} 19.2)+ 0.9/27.4) +6.0/23.7| +7.2/22.6)/+ 4.4/22.2)4+ 4.5
of 62 {1.831}1.699) 18.9/+19.6)31.5) +8.4/29.4) +9.8)26.7/+10.3)26.3/+11.4
2 64 |1.677/1.606) 21.1)— 1.2/32.6| —3.5/29.4| —4.9/26.8|/— 4.3]26.2|/— 4.3
rot 67 |1.651/1.587| 21.6/+ 0.8/33.3) —4.6)80.5) —5.1/27.3}— 4.7/28.0)/—11.2
2 72 |1.610/1.492} 26.9)/+ 5.2/33.3} +1.3/80.4| +1.7/28.3]— 2.3/27.4/— 5.2
ot 160 |1.791/1.752| 25.0/+ 0.3/38.0| —4.0|/85.3| +2.4/32.6/+ 1.1/32.5|/— 2.2.
rot 218 |2.008]1.824) 28.0/+ 2.0/40.1) +1.2/38.5| +0.9/34.8/4+ 4.9/84.7)/— 0.8
VETAD Cote sod Ws + 3.9 +1.5 +1.3 + 1.2 + 0.2
Standard deviation....... + 6.1 +3.9 +5.0 + 5.2 = 6.5
gain more than do the younger ones; yet the lighter brains gain
more than do the heavier ones if we can eliminate age as a fac-
tor, as was done in table 3. The curve of increase in weight
may be considered as capable of solution into at least two curves
—expressing these two factors. Age appears to be by far the
more potent factor.
416 JAMES STUART PLANT
A further study was made of fifty-nine brains arranged accord-
ing to increasing brain weight but without regard to sex or litter.
In table 4 these are arranged according to initial brain weight in
three age groups. Within these grouns two phenomena are
apparent (shown in the averages under the vertical column,
‘Percentage difference from the following value’). These are the
early greater gains for the lighter brains (this does not hold
clearly for the ten brains of the youngest group where there is
practically no difference); and the fact that at the seventy-five-
day weighing the lighter brains show relatively a less percentage
of increase than they do at the twenty-four-hour weighing.
Since these facts are just those which determine the curve when
brains of rats of the same age, sex, and litter are compared, we
may conclude that in the reaction of the brain to Miiller’s fluid:
1. Sex is negligible.
2. Inherited composition is negligible.
3. Approximate similarity of ages (the range being limited)
may be considered as having the same effect as though the ages
were identical.
The data on the percentage of water—in the last column of
table 4—will be discussed later.
A group of four brains—all belonging to young rats—was sub-
jected to an additional procedure. The brains, immediately
upon removal, were separated into cerebrum, cerebellum, stem,
and olfactory bulbs. Each part was then treated as were the
‘whole brains of the other series. The data are given in table 5
in this way, that that percentage of the whole brain weight
represented by the weight of each part at each weighing is re-
corded. The figures for the four brains show but slight varia-
tion, and table 5 therefore presents only the averages of the four.
The relative weights of the various parts undergo considerable
change in Miiller’s fluid, but this change is mainly consummated
in the first twenty-four hours. Thus we may assume from this
study that, while the relations of the various parts are altered in
the fixing solution, the length of time, after the first twenty-four
hours, during which the parts are subjected to this treatment, is |
a matter of minor import.
ALBINO RAT BRAIN IN MULLER’S FLUID 417
TABLE 4
The effect of initial brain weight—albino rat—on the percentages of gain of brains
arranged in age groups, regardless of sex or litter
. 38 AVERAGE PERCENTAGES OF GAIN
ak —
NuM- | AVER as Ses 3
INITIAL WEIGHT a “| 24 Qo & = mn
(crams) — /PCxses | ace | FOURS ace Pal || 30 75 [8 oe 38
ges days | days | days | days Bae z Ee
tale BSS$8| 53
4 4 ow
50 to 60 days
1.35-1.40 1 52 | 18.7 |— 2.8] 28.5 | 25.7 | 22.9 | 22.2 |— 5.6] 79.3
1.40-1.50 1 52 | 19.3 |—11.1] 27.9 | 26.8 | 24.1 | 23.5 |—11.2] 79.3
1.50-1.60 3 53 | 21.7 |+ 5.8} 33.2 | 29.2 | 27.5 | 26.5 |+ 1.2] 80.1
1-.60—1.65 2 o7 | 20.5 |— 2.2) 32.3 | 29.0 | 26.6 | 26.2 |— 3.0} 79.9
1.65-1.70 2 58 | 20.9 |+ 3.2) 32.8 | 28.8 | 27.7 | 27.0 |+ 8.7] 80.3
1.70-1.80 il 57 | 19.5 SOR) | 2020) | 2oeen eas 80.3
Average.1.58 10 55 | 20.5 |— 0.9 31.7 | 28.1 | 26.3 | 25.7 |— 2.1) 80.0
60 to 70 days
1.40-1.50 7 64 | 21.7 |+ 1.9) 32.9 | 29.8 | 26.6 | 25.7 |+ 0.5] 79.7
1.50-1 .60 5 65 | 21.3 |— 8.6) 32.1 | 29.1 | 26.2 | 25.6 |— 9.3] 79.3
1.60-1.65 6 69 | 23.3 |+12.3) 34.1 | 31.5 | 28.8 | 28.2 |+10.6} 80.2
1.65-1.70 |} 11 68 | 20.7 |+ 2.9) 31.5 | 28:5 | 26.0 | 25.5 |4+ 3.1] 79.7
1.70-1.80 5 65 | 20.1 |+ 8.2) 31.6 | 28.2 | 26.4 | 24.7 |— 1.8} 80.1
1.80-1.90 3 63 | 18.6 Se e209: leave tala 2 80.7
Average.1.65 | 37 66 | 21.1 |+ 3.9} 32.3 | 29.2 | 26.6 | 25.9 |4+ 1.8] 79.8
180 to 240 days
1.65-1.70 1 189 | 28.7 |+ 8.5} 42.4 | 39.2 | 36.1 | 35.4 |+ 9.3) 81.6
1.70—1.80 4 191 | 26.5 |+ 2.8) 38.2 | 36.2 | 33.2 | 32.4 |+ 3.2) 80.5
1S 5 36.4 | 32.4 | 31.5 |+ 6.1} 80.5
1.9 2 33.0 | 30.4 | 29.7 78.4
Average.1.83 | 12 | 213 | 25.9 |+ 6.1) 37.9 | 36.0 | 32.7 | 31.8 |+ 5.6} 80.2
*The percentages were obtained originally by the use of values carried to
three places. These have now been reduced to one-place numbers and there
are therefore some apparent discrepancies in the percentage-difference columns.
These differences, however, are not significant.
418 JAMES STUART PLANT
TABLE 5
Averaged percentage weight relations of the parts of four albino rat brains during
the course of their reaction to Miiller’s fluid
WEIGHT OF PERCENTAGE| PERCENTAGE! PERCENTAGE SL SEE NEE
REPRE- TIME IN
AGE emer a pet ace x een REPRE | sENTED BY | MULLER’S
SrsAGe OF (| gmepnn a | smn) E097 | pues ee
a2, 1.610 Ga. 17 17.40 13.80 Sialos" Initial
QV27 63.10 17.68 15.08 4.14 24 hours
PAU 64.82 IZ PALL 14.33 3.63 7 days
2.094 63.77 ieaiG 14.72 39.5 30 days
Difference between per-
centages at initial and
24-hour weighing.......| —2.07 +0.28 +1.28 +0.51
Difference between per-
centages at 24-hour
and 30-day weighing....| +0.67 +0.08 —0.36 —0.39
WATER RELATIONS
We have studied the water relations after seventy-five days in
fifty-nine whole brains and after thirty days in the parts of three
brains. There is evidently, in’the reaction of the brain to the
Miiller’s fluid, a deposition of salts in the brain tissue. This is
shown in table 6. Part A deals with the fifty-nine whole brains;
Part B with the parts of three brains (all belonging to young
females). The deposition of salts at the end of seventy-five
days in the whole brain is from 3.9 per cent to 4.3 per cent of
the total water of the brain despite the fact that the salts in
Miiller’s fluid are present in a concentration of but 3.5 per cent.
This shows a deposition of salts in the tissues. If, in addition,
there is some diffusion of solids from the brain to Miiller’s fluid—
and, from inspection, this appears to be the case—the percent-
age of salts deposited must be even higher than that indicated
by the figures given.
The final percentage of water-in a given brain at the seventy-
five-day weighing is only slightly greater than that of a fresh
brain belonging to a rat of the same age, sex, and litter. In
view of the 20 to 30 per cent net increase in weight in Miller’s
eS
ALBINO RAT BRAIN IN MULLER’S FLUID 419
TABLE 6
Part-A. Water relations at the seventy-five day weighing of fifty-nine whole brains
(see table 4)
UNDER 60 | 6070 120 | over 120
DAYS DAYS DAYS
Average fresh brain weight.......00. 0.0. uss. see ek 1.582 | 1.651 | 1.826
Percentage of water (from Donaldson, ’16)..... etacs. 79.1% | 78.9% | 78.1%
Calculated amount of water represented in fresh
LOLS, ae pce ale le ARS Aaa aia th ls Ph, 9 atta 2 ad 0 ge 1.252'| 1.303 | 1.426
Hime bean meigha? /). LILES A Oe A 1.988 | 2.078 | 2.407
Buel amount, of water.<..°.:lc.se.hand.d poueseer 1.590 | 1.659 | 1.932
Percentage of water—observed....................... 80.0% | 79.8% | 80.2%
Increase in weight due to water, gms................. 0.338 | 0.356 | 0.506
Increase in weight due to salts, gms.................. 0.068 | 0.071 | 0.075
Percentage of salts in total increase in weight.........| 16.6% | 16.7% | 13.0%
Percentage of salts in the total water in brain........ 4.29, \ 4.3% | 3.9%
Part B. Water relations at the thirty-day weighing of parts of three brains
: OLFAC-
‘ CEREBRUM STEM dene ria
Average fresh brain weight................. 1.097 | 0.292 | 0.235 | 0.058
Percentage of water (from Donaldson, ’16)...| 80.0% | 76.1% | 79.7% | 82.3%
Calculated amount of water represented in
UTE SUnS Ter PS 6) RS ae ee ne 2 ea 0.878 | 0.222] 0.188} 0.048
RRC Ges. ote hae he eet Atos oo ak 1.403 | 0.386 | 0.329] 0.078
Final amount of water....................... 1.139 | 0.301 | 0.266 | 0.067
Percentage of water—observed...............| 81.2% | 77.9% | 80.8% | 85.6%
Increase in weight due to water, gms........ 0.261 | 0.079 | 0.078 | 0.019
Increase in weight due to salts, gms.......... 0.044 | 0.016 | 0.015] 0.001
Percentage of salts in total increase in weight
(for totaljbrain:14)8%)). si... edn. ad. 2 14.5% | 16.4% | 16.4% | 4.9%
Percentage of salts in total water in the part
RRR VGA ORDA 5 V5 aan. o>, paminne Be ele nd «oes 3.9% | 5.0% |. 5.8%. | 125%
fluid, this seems a striking fact, though where the initial per-
centage of water is so high, it is evident that it takes a relatively
large difference in the absolute water content to markedly
affect the percentage value.
In the three brains divided into their parts the salts are de-
posited in the following percentages after thirty days:
420 JAMES STUART PLANT
Cerebelltm ss (275-24 a tae ie ee ee ee te is os nee ee 5.8 per cent
Stemmid aac. Ai LB ER oF ks 5.0 per cent
Ceréebruim::coic os, Lean et et AR ee Be Ia ee 3.9 per cent
OLRCHOTY DULDS. 2 aide Ie Sth se sian aie eeie ols Maen peee 1.5 per cent
Average cor whole Braiienrts seks: sacs: elt eee ++. 4.3 per cent
With the exception of the cerebellum, the percentage of salts
deposited is in direct relation with the proportion of myelin in
the part involved. As none of the parts were washed, it may
well be that the interstices of the cerebellum were the site of
large deposits of salts, a physical factor which may account for
this anomalous result.
CONCLUSIONS
General reaction of the brain to Miiller’s fluid, or type curves.
The brain of the albino rat undergoes a typical change when
‘fixed’ in Miiller’s fluid (2.5 per cent potassium bichromate and
1 per cent sodium sulphate). There is a rapid increase in
weight followed by a slow, steady loss until at the seventy-five-
day weighing the brain weighs 20 to 30 per cent more than when
fresh. ,
Factors affecting this reaction, or components of the type
curve.
1. Age is the main condition controlling this reaction to
Miiller’s fluid. The brains of older rats gain more and retain
this higher relative gain throughout the seventy-five days.
2. Initial brain weight, or size, is the condition of next impor-
tance. As between brains of like age the lighter brains gain
more during the earlier part of their stay in Miiller’s fluid. This
difference is gradually lessened, and it disappears at the seventy-
five-day weighing.
Thus, while age and initial brain weight are highly correlated,
they constitute factors which, when taken alone, influence in
opposite ways the reaction of the brain to Miiller’s fluid. The
early greater increase of the smaller brain of two rats of the
same age may be due to one or both of the following factors:
a. If we consider the brain as a sphere and the fluids as pene-
trating at a fixed rate, then in the smaller brain a slightly greater
ALBINO RAT BRAIN IN MULLER’S FLUID 421
proportion of the brain will be penetrated—that is, swollen by
the fixing fluid—at any given instant in the early part of the
reaction. This would give a more rapid enlargement in the
smaller brain.
b. The smaller brain has a higher percentage of water which
might make diffusion more rapid. If this is a controlling factor,
the matter of a higher percentage of water must be of more
immediate importance than is that of a lesser percentage of
myelin.
3. Approximately similar age (the range being limited) has
nearly the same determining value as equality of age.
4, Sex is a negligible factor, as is also inherited composition
(relationship within a given strain).
FINAL WATER RELATIONS
The increase in weight is due mainly to the taking up of water,
but the percentage of salts deposited in the fixed tissues is much
greater than that in the fixing fluid. With the exception of the
- cerebellum, the deposition of salts is proportional to the myelin
present in the part of the brain.
RELATION OF RESULTS
The curve of reaction of ‘control’ brains to Miiller’s fluid is
evidently of such constancy in character as to make it a satis-
factory criterion for a judgment as to alterations produced in
the brain by various experimental procedures. It seems that
in problems involving changes not available to such microscopic
or analytical tests as we have, this reaction might furnish a
valuable means of study. It appears from the foregoing results
that when tests are made for the experimental modification of
the response of the brain to Miiller’s fluid, it is necessary to have
the test and control brains of the same age and, in those cases in
which the brains differ in weight, to allow sufficient time for the
compensation of this difference.
422 JAMES STUART PLANT
LITERATURE
While various observations bearing upon this problem have
been made, none have employed exactly our experimental pro-
cedure. Various solutions of potassium bichromate have been
used—Donaldson (94), Fish (’93), and King (’10)—but a study
in the changes in the reaction to Miiller’s fluid has not been
made. A similar study was carried out by Hrdlicka (’06) in
which various formalin preparations were used as fixing reagents.
The effect of age upon this reaction has been discussed in a gen-
eral way—Donaldson (94) and King (’10)—but it seems that
the extent of its control over the reaction has not been previously
stated.
BIBLIOGRAPHY
Donatpson, H. H. 1894 Preliminary observations on some changes caused in
the nervous tissues by reagents commonly employed to harden them.
Jour, Morph., vol. 9.
1916 A revision of the percentage of water in the brain and in the
spinal cord of the albino rat. Jour. Comp. Neur., vol. 27.
Fisn, P. A. 1893 Brain preservation with a résumé of some old and new
methods. Wilder Quarter-Century Book, Ithaca.
Hropuickxa, A. 1906 Brains and brain preservatives. Proc. U. S. Nat. Mu-
seum, vol. 30.
Kine, H. D. 1910 The effects of various fixatives on the brain of the albino
rat, with an account of a method of preparing this material for a study
of the cells in the cortex. Anat. Rec., vol. 4.
FS ei i 9s , |
betapich Vitel ae . PO HI TT 44.
Waray. ' ait ths vi rds fae
v4 | a
1 .
. ¥ » .
aa TG) fa) hnareaiz ns meen te hee. eae
: j
ry L ‘ j
ay. t Pi
koe ites aii) ai “aN Stage MAE sbi alivats any ‘ip et Rosa as
at Toh Birtley MA a Lier tho: SORE: Pray a ay gat. He :
Me aby ios. SEK bof onl Niel ip) ete ee % oe anigodta ait) ie
| (CP sea a DIYARG deb Aomericda 6 PE Re arsesley LNg bak
hora toon toh aaheepigiaige eile, aah ep icaNE Shag |
bi aiahirsbtion “f nay ba ie LEE Ss oie +.) ae dost ¥e oe a is
i bet rt “ay vila gE ACs RG Gea te bis y plaka oe
| ya i)
so potas i oti Rin eae cha ry rp Aa arial aang NF se
oe anti Lehn Hs iaygeyeal ep rd Halal i He on
Satine dyif; xan fa ane dress). vel Yh at ahaa Hen Bee
coal ipod: selsb-Aattinoserey MRA’ Faxkdeame. a ale
nye eines ip FD MO ee oha't, iia fay Au teal wveaglbnt,
EE ERE IE ant aati gihent AiGhoy age
es al i, ad (is dah |
LMA to Pinta fac) Vigne S
Ky
Resumen por el autor, Olof Larsell.
Universidad de Wisconsin.
Estudios sobre el nervio terminal: la tortuga.
El autor describe las relaciones periféricas y la distribuci6én
del nervio terminal en el embrién de la tortuga. Un plexo del
nervio, situado en el tabique nasal, esta’ intimamente mezclado
con un plexo de la rama oftdlmica del nervio trigémino. El
autor compara las células ganglionares del nervio terminal, en
lo referente al tamafio y caracteres morfoloégicos generales, con
las células de los ganglios sensorio y del simpdtico de los mismos
embriones. Las células del nervio terminal presentan una seme-
janza sorprendente, tanto en,el tamafio como en la forma, con
las células halladas en los ganglios del simpatico, especialmente
las de los situados en la regién cefdlica. Las células de los
racimos ganglionares del nervio terminal no pueden diferenciarse
de células emigrantes que se presentan a lo largo de la rama
oftalmica del trigémino. Todo esto indica que el nervio terminal
esta relacionado con el sistema del simpatico.
Translation by José F. Nonidez
Carnegie Institution of Washington
AUTHOR’S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE, JUNE 30
STUDIES ON THE NERVUS TERMINALIS: TURTLE!
O. LARSELL
Department of Anatomy, University of Wisconsin
SIXTEEN FIGURES
The present contribution was begun as part of a comparative
study of the nervus terminalis in several groups of vertebrates.
The work had not progressed far before it was found advisable to
confine attention to one group at a time, so that the greater
portion of the present report embraces the results of observa-
tions made since the studies on the nerve in mammals by the
author (718) was published.
The somewhat extensive literature of the nervus terminalis
was reviewed in the previous article, and only two papers which
have appeared on the subject during the past year will be men-
tioned briefly. hese papers are by Van Wijhe (’18) and
Ayers (719).
Van Wijhe’s paper reviews much of the literature of the nervus
terminalis in the various groups of vertebrates briefly and homol-
ogizes the nerve with one he noted a number of years ago (’94)
in Amphioxus, which he termed at that time the ‘ner'vus apicis.’
He states: “‘Before the homologue of the profound ophthalmicus
there is in Amphioxus still another nerve which supplies the
utmost point of the snout. On account of this and because it
arises from the morphological fore-end of the cerebral ventricle I
called it the nervus apicis.”’
Ayers (719), in continuing his studies of Cephalogenesis, begun
long ago, has found the nervus terminalis (Van Wijhe’s ‘nervus
apicis’) in Amphioxus, and calls attention to its large size as
1 Contribution from the Zoological Laboratory of Northwestern University,
William A. Locy, Director, and from the Anatomical Laboratory, University of
Wisconsin.
423
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, No. 5
424 O. LARSELL
compared with the olfactory nerve in that form. He finds it
also in Cyclostomes and states that in Bdellostoma the nerve
presents an intermediate stage, as respects its size and relations,
between Amphioxus and the selachians. He calls attention to
the distinction between the vomeronasal nerve, which he terms
the ‘nervus septalis’ and the olfactory nerve, and suggests a new
classification of the cranial nerves in which the nervus terminalis
would be number J, and, with the ‘nervus septalis’, would be
added to the list of twelve cranial nerves usually recognized.
The nervus terminalis is considered a sensory nerve which has to
do with a group of chemical sense organs, and is related physi-
ologically to the vomeronasal (his septal) nerve. The conclusions
are reached from a study of Amphioxus and cyclostomes. Doctor
Ayers believes that in higher forms the nerve has undergone con-
siderable modification due to changes in head structure.
I am under great obligation to Doctor Ayers for opportunity
to read his manuscript prior to publication and for permission to
make use of his observations. He also afforded me opportunity
to read Van Wijhe’s paper, which I had not previously seen. It
is a pleasure to express here my sense of indebtedness to him.
My acknowledgments are also due Prof. William A. Locy, of
Northwestern University, under whose direction the general
problem was originally begun and who has since continued his
interest.
MATERIAL AND METHODS
Embryos of the painted turtle (Chrysemys marginata) were
used. Most of these had been fixed in a formol-bichromate-
acetic fluid, some in formalin of 10 per cent, others in Tellyes-
niczky’s fluid, and a few living embryos were obtained and pre-
pared by the Cajal and the Vom Rath methods.
Stages beyond 10- to 1l-mm. carapace length had become
chitinized to such an extent in the rostral region that intact
serial sections could not be obtained. Chiefly for this reason,
the present contribution is confined to a description of the nervus
terminalis in embryos up to 11-mm. carapace length (about 17
mm. greatest length.
at a
NERVUS TERMINALIS: TURTLE 425
Numerous dissections of embryos and of newly hatched turtles
were made with the aid of the binocular microscope. The head
was split slightly to one side of the midsagittal plane, and the soft
parts were then sufficiently removed to expose the nerve and
its adjacent structures.
The embryos sectioned were cut in the sagittal plane or trans-
versely, and were stained by various methods. The most
generally satisfactory stain for older stages was found to be
iron-hematoxylin, but some of the most instructive series were ob-
tained by overstaining with Delafield’s hematoxylin, followed by
a counterstain of saturated aqueous orange G to which two drops
of glacial acetic acid were added for each 50 cc. of stain.
The serial sections studied were-as follows:
1 series 6-mm. embryo, sagittal, stained with iron-hematoxylin.
2 series 6.3-mm. embryo, transverse, stained with hematoxylin and Congo red.
3 series 7.5-mm. embryo, transverse, stained with hematoxylin and Congo red.
1 series 8-mm. embryo treated by the Cajal method, cut sagittally.
1 series 9-mm. embryo, sagittal, stained with hematoxylin and Congo red.
1 series 9-mm. embryo, sagittal, treated by Vom Rath method.
1 series 9.5-mm.-carapace-length embryo, stained with hematoxylin and Congo
red, sagittal plane.
2 series 10-mm.-carapace-length embryos, sagittal, stained with iron-hema-
toxylin.
1 series 10.5-mm.-carapace-length embryo, sagittal, stained with hematoxylin
and erythrosin. _
1 series 11-mm.-carapace-length embryo, sagittal, stained with hematoxylin and
Van Gieson’s stain.
1 series 1l-mm.-carapace embryo, sagittal, stained with hematoxylin and
orange G.
‘DESCRIPTIVE
The nervus terminalis in the turtle has its origin by several
small roots from the ventromesial surface of the forebrain, just
caudad to the olfactory bulb. It can be demonstrated by dis-
section in suitably prepared material. Figure 1 represents a
dissection of an embryo of 11-mm. carapace length, showing the
left nervus terminalis and its relation to neighboring structures.
In the specimen figured the rootlets were not evident until some
of the overlying brain tissue had been removed by brushing. By
this process three roots were demonstrated. In other dissections
but two roots were brought to light, usually after some brushing.
O. LARSELL
426
NERVUS TERMINALIS: TURTLE 427
In sections of corresponding stages, cut in the sagittal plane, two
or three roots were observed to enter the brain substance, but
their fibers could be followed within the brain for only a short
distance (fig. 2).
On following these roots distally from the brain, they are seen
to unite (figs. 1 and 2) and a ganglionic swelling was invariably
found just beyond their point of union. In the dissection figured
it will be noted that the two more dorsal roots unite to form a
single short trunk before entering the ganglion, while the ventral
root enters the ganglion directly. In sections a number of
ganglion cells (figs. 2 and 3) may be observed in this mass, but
in none of the embryos examined did this ganglion appear to have
as many cells as others located more rostrally, especially the one
marked gn. (fig. 3).
Rostrally from the ganglionic mass which lies at the junction
of the rootlets, the nerve continues as a compact trunk as far as
the most anterior part of the olfactory bulb, midway between the
Fig. 1 Dissection of head of turtle embryo of 11-mm. carapace length to
show the nervus terminalis and its relation to neighboring structures. The
left lateral half of the brain is viewed from the mesial aspect. XX ca. 15.
Fig. 2 Central roots of the nervus terminalis at their point of junction with
the brain. Turtle embryo of 1l-mm. carapace length. Hematoxylin and
orange G stain. X 410.
Fig. 3 Reconstruction of the right nervus terminalis of turtle embryo of
1l-mm. carapace length, stained with hematoxylin and orange G. The nervus
terminalis is represented as lying on the medial surface of the vomeronasal
nerve for all of that part of its course which is parallel to the latter. Part of
its course, however, as shown in figure 1, is lateral to the vomeronasal nerve.
X ca. 16.
ABBREVIATIONS
bl.v., blood-vessel m.ob.ven., ventral oblique muscle
bu.olf., olfactory bulb m.r.ant., anterior rectus muscle
cer.hem., cerebral hemisphere n.olf., olfactory nerve bundles
cr.cav., cranial cavity n.oph., ophthalmic branch of V nerve
gn., main ganglionic mass (ganglion 7.ter., nervus terminalis
terminale) of nervus terminalis n.vom., nervus vomeronasalis
gn.’, accessory ganglion terminale na.ant., anterior naris
gn.c., ganglion cells ’ na.po., posterior naris
gn.cl., ganglionic clusters olf.epith., area of olfactory epithelium
mes., mesencephalon op.chi., optic chiasma
m.ob.do., dorsal oblique muscle ret., retina
428 O. LARSELL
dorsal (vomeronasal) bundle and the main bundle of olfactory
fibers proper, as shown in figure 1. At this point it turns to
follow these bundles, passing between them in such a manner’
that it could not be traced further by the method of dissection.
Sections, however, reveal the further course of the main bundle
of the nerve and indicate the presence of numerous ganglionic
cells scattered along its trunk as it passes over the mesial surface
of the olfactory bulb. These cells form clusters of various sizes.
One of the largest of these clusters is no doubt indicated by the
swelling (fig. 1, gn.’) shown in the dissection. Another and larger
ganglionic mass is shown (fig. 3, gn.) at the point where the ter-
minalis passes lateral to the vomeronasal bundle. This corre-
sponds to the position usually occupied by the largest cluster of
cells in the majority of the embryos which were sectioned.
From the position of this ganglion distally the nerve could not
be followed further by the method of dissection, because its
strands became too intimately mingled with those of the olfactory
and vomeronasal nerves. Fortunately, however, in some of the
series of the older stages studied, a differential stain was obtained
by the method previously described, so that the terminalis bundles
could be distinguished from those of the other two nerves and,
further rostrad, from the fibers of the ophthalmic branch of the
V nerve. This differentiation was aided by the fact that the
olfactory and vomeronasal nerves appear as compact bundles of
wavy fibers with few nuclei scattered among them. The strands
of the nervus terminalis are smaller and much less compact and
present relatively numerous sheath nuclei as well as larger gan-
glionic cells. Where the strands of the trigeminus were inter-
mingled on the septum, they also had a characteristic appearance, °
apparently due to the process of myelination, as well as to dif-
ferential staining. ‘These characteristics, however, could not be
noted with any degree of certainty in the smaller tracts, so that
the reconstruction represented in figure 3 indicates only the
larger bundles and their grosser ramifications.
In some of the preparations the nerve was seen to divide in-
tracranially at the ganglion gn. into two strands, which, however,
reunited to form a compact trunk before the nerve left the brain
NERVUS TERMINALIS: TURTLE 429
cavity. This condition is illustrated in figure 7. There were
some indications of much finer strands also in this region, but
they were not sharply enough differentiated from the olfactory
strands to justify inclusion in the figure as part of an intracranial
plexus of the terminalis.
After emerging from the cranial cavity, the nervus terminalis
is composed of several well-marked strands which continue par-
allel with the vomeronasal bundles, mesial and in part dorsal, to
the latter. A short distance from the point of emergence of the
nerve, its strands begin to form a plexus over the nasal septum.
Lack of silver preparations of the older stages made it impossible
to follow this plexus for any considerable distance, especially in
its more rostral part, where it becomes more complex due to
entrance of fibers from the trigeminus. It was very evident that
both the nervus terminalis and rami from the ophthalmic branch
of the V nerve take part in the formation of a plexus on the nasal
septum.
Clusters of ganglionic cells (fig. 3) were scattered throughout
this plexus. Along the vomeronasal nerve there was a nearly
continuous mass of cells from the point where the nerve turned
ventrorostrally at the bulbus olfactorius, to the point where the
more profuse spreading out of the septal plexus began. A some-
what similar arrangement of cells along the vomeronasal nerve
was found by Johnston (’13) in Emys, but apparently the cells
were not so numerous as in Chrysemys.
A fortunate Cajal preparation of an 8-mm.-total-length embryo
gave a very clear demonstration that the trigeminus forms a more
important portion of this septal plexus than might have been an- —
ticipated. As shown in figure 4, which represents a reconstruc-
tion from fourteen serial sections cut in the sagittal plane, the
trigeminal portion of the plexus is formed by the ramifications of
the ophthalmic nerve. The fibers were stained quite uniformly
brown or black. They could be followed to the gasserian gan-
glion, with the beautifully stained cells of which they united.
The nervus terminalis in this preparation is represented by a
few clusters of cells and some yellowish fibers. The cells could
not with certainty be distinguished from the mesenchymal cells,
430
O. LARSELL
NERVUS TERMINALIS: TURTLE 431
but sections of embryos of approximately the same stage which
were stained by other methods, indicate that the clusters are
composed of ganglionic cells, which from their position no doubt
belong to the nervus terminalis.
Some details of this plexus of the ophthalmic nerve are illus-
trated in figures 5 and 6. As shown in the figures, relatively
small strands of fibers meet at nodal points, from which the indi-
vidual fibers are redistributed to bundles diverging at various
angles. No ganglionic cells could be observed about these nodal
points at this early stage. In older stages, however, cell clusters
of various sizes are numerous at the points where the ophthalmic
nerve ramifies on the septum, and are found as far forward (figs.
3 and 7) as the most rostral part of the septum. It seems likely
that the more rostral of these cells correspond to clusters of sym-
pathetic cells described and figured by Willard (’15) in Anolis.
The observation of Rubaschin (’03) of a ‘ganglion olfactorii nervi
trigemini’ on one of the branches of the ophthalmic nerve in the
chick appears also to be related.
At various points the branches of the trigeminus and of the
terminalis become so intimately related that the two cannot be
told apart, and the smaller strands of the plexus which continue
from these points, appear to contain fibers from both nerves. As
shown in figure 7, which represents a reconstruction from nine
serial sections, from which the finer strands of the plexus are
omitted, the nervus terminalis anastomoses with one of the
larger branches ‘of the ophthalmic nerve just dorsal to the vome-
ronasal nerve. At the point of anastomosis is a large ganglionic
cluster. The two nerves had the characteristic different appear-
ance, previously noted, proximal to their point of union, but their
Fig. 4 Reconstruction of the septal plexus of the ophthalmic branch of the
trigeminal nerve of a turtle embryo of 8 mm. greatest length, prepared by the
Cajal method. The figure was reconstructed from sections 79 to 92 of the series.
The numerals indicate the sections from which the adjacent structures were
projected. For the sake of simplicity, section 79 is indicated by the numerals 1,
section 92 by 14, and intervening sections accordingly. X 80.
Fig. 5 Portion of the plexus from section 81, at point marked 3* in the pre-
vious figure, to show details of structure. > 385.
Fig. 6 Portion of the plexus from section 88 of the series, at point marked
5* in the reconstruction, to show detail. XX 385.
: 0. LARSELL
~,
olf epith:
NERVUS TERMINALIS: TURTLE 433
strands could not be told apart distal to this point. The relation
of the two nerves and something of their appearance are repre-
sented more highly magnified in figure 8, but this does not show
the slight although clearly apparent differentiation of staining
which the preparations themselves reveal in the larger bundles.
The trigeminal fibers appear coarser and more compactly collected
into bundles than do those of the terminalis, but in the smaller
strands these characteristics are not apparent, probably because
of the small number of fibers composing them.
The olfactory and vomeronasal bundles, which are also rep-
resented in the figure, resemble each other in being composed of
very delicate fibers with a characteristic wavy appearance which
could not be well represented in the drawing.
Huber and Guild (713) in the rabbit and Brookover (717) in
the human found indications of such an anastomosis of terminalis
and trigeminal strands on the nasal septum, and the pres-
ent writer (’18) partially demonstrated it in the cat. It is
rather striking that three nerves, the terminalis, the olfactory
(Read, ’08), and the trigeminus, should each form a plexus on -
the nasal septum. Two of these plexuses overlap to a marked
degree. The vomeronasal nerve is not plexiform, except as the
large bundles composing it branch and reunite to a slight extent
in their course to the vomeronasal organ. In this connection a
statement from the previously cited paper of Ayers is of interest.
He states: ‘‘The nasal chamber in man therefore contains the
surface distribution of these three (terminalis, septalis, olfac-
torius) cranial nerves as well as the surface terminations of in-
vading branches of a fourth and more recent cranial nerve, the
trigeminus.”’
Fig. 7 Reconstruction from ten sections (175 to 185) of the series from which
figure 3 was reconstructed, to show anastomosis between rami of the nervus
terminalis and of the ophthalmic branch of the trigeminal nerve. 44.
Fig. 8 Reconstruction from four of the sections (sections 182 to 185) in-
cluded in figure 7, to show at higher magnification the anastomosis of one of the
rami of the ophthalmic plexus with a bundle of the nervus terminalis. This
figure also gives some idea of the appearance of the fiber bundles. X 180.
434 O. LARSELL
GANGLION CELLS
An effort to reach some conclusion as to the character of the
ganglion cells of the terminalis in the turtle embryos was made by
comparing them with cells of other ganglia, cranial, spinal and
sympathetic. The gasserian, sphenopalatine, and ciliary ganglia
were studied in the head region. The spinal ganglia, beginning
with the first thoracic, and the sympathetic chain ganglia were
studied in the body region. Measurements were made of the
nuclei with the aid of an ocular micrometer, the results of which
are indicated in table 1. The cells measured were taken at
random. The only selection exercised was in measuring nuclei
TABLE 1
AVER-
NUM- | SIZE OF|SIZEOF| AGE
u BER | LARG- | SMALL-| SIZE OF
POSITION OF CELLS “anes (Gea Rea! SS REMARKS
URED | CLEUS | CLEUS | NUM-
BER
Periphery of spinal | 48 | 12.3 | 8.2 | 9.8 | Many unipolar.
gangha
Central part of same | 52 8.2 | 5.3] 7.3 | Bipolar, approaching unipolar
ganglia condition
Above combined 1005) 223. 9|\ 523
Gasserian ganglion 52 | 11.4! 7.0] 9.0 | Two sizes present, but small
cells not many
Ciliary ganglion, 50 | 12.4 | 7.0|9.0 | Nuclei large in proportion to
large cells entire cell
Ciliary ganglion, 22 7.9| 5.3 | 6.4
small cells
Above two combined | 72 | 12.4} 5.3 | 8.3
Sympathetic chain | 50 8.1 | 5.3 |6.2 | Five nuclei larger than 6.7 u
ganglia
Sphenopalatine 50 8.115.838 | 6.74
ganglion
N. terminalis, periph- | 50 8.8 | 5.3 | 6.72 | Some of these cells probably
eral clusters belonged to clusters related
to the trigeminus
N. terminalis, central | 30 8.4 | 5.3 |6.7 | Of the 80 nuclei measured in
root clusters . both peripheral and cen-
tral clusters three were
larger than 8.1 » and three
were smaller than 5.5 xz.
NERVUS TERMINALIS: TURTLE 435
which were spherical or nearly so. In the case of some cells,
especially many in the terminalis ganglia, the nuclei were so
elongated that it was necessary to measure the greater and the
lesser diameters and take the mean of the two.
All of the measurements and the drawings of the ganglion cells
represented were made from a single embryo of 10-mm. carapace
length, stained with iron-hematoxylin. This was done for the
sake of uniformity, although the statements hold in general for
all the embryos examined which were sufficiently advanced to
show any pronounced differentiation of the various types of nerve
cells. In drawing the figures the outlines of the cells and nuclei
were traced with the aid of the camera lucida, and the same com-
bination of lenses was employed in each case, so that the figures
represent directly the variations in form and size of the various
types.
Comparison of embryos at different stages of development in-
dicated that in embryos of 10-mm. total length, the spinal gan-
glion cells were on the average somewhat larger than those of
the sympathetic chain ganglia. There was, however, but slight
difference in the size of the individual cells within the spinal
ganglia. The sympathetic chain ganglion cells had still much
the appearance of indifferent cells. In the gasserian ganglion of
embryos at this stage of development, many of the cells were
larger than the spinal ganglion cells of the same embryo.
In embryos of 9.5 to 11l-mm. carapace (15.5 to 17 mm. greatest
length) there is a marked difference between the size of the
largest sensory ganglion cells and those of the sympathetic
chain ganglia. The latter (fig. 12) are of pretty uniform size,
but the spinal ganglia and, in less marked degree, the cranial
ganglia, showed two fairly distinct sizes of cells. The larger
type in the spinal ganglia (fig. 9) was found near the periphery
of the ganglion. Many had already reached the unipolar con-
dition, but others showed various transitional stages from the
primitive bipolar cells.
Of one hundred cells measured from three different ganglia,
one in the thoracic region, one in the lumbar, and one in the
sacral, the largest nucleus had a diameter of 12.3 u and the
O. LARSELL
436
NERVUS TERMINALIS: TURTLE 437
smallest was 5.3 » in diameter. The average size of the entire
one hundred nuclei was 8.5 uw. These cells were divided into
two groups, those with a nuclear diameter greater than 8.2 u
and those whose nuclear diameter was less than this, down to
5.3 » which was the smallest found. While this division was
somewhat arbitrary, there was sufficient ground for it in the
character of the cells, aside from their size, to make it appear
justifiable. The smallest cells (fig. 10) had considerably less
cytoplasm surrounding the nucleus, both relatively and ac-
tually. Only various stages of the bipolar condition were ob-
served in these smaller cells. They were found closely packed
together in the central portion of the ganglion, while the larger
cells were nearer the periphery.
As indicated in the table, the fifty-two cells which showed a
nuclear diameter less than 8.2 u had an average diameter of
7.3 uw. The forty-eight cells whose nuclear diameter was greater
than 8.2 « were found to have an average nuclear diameter of
9.8 yu.
The smaller cells (fig. 10) may correspond to the small ganglion
cells found in the spinal ganglia of mammals by Dogiel (’08),
Ranson (712), and other workers, but it seems likely that some
of them at least represent cells of the larger type which have not
Fig. 9 Cells from the peripheral portion of the first thoracic spinal ganglion
of a turtle embryo of 10-mm. carapace length. Iron-hematoxylin stain. > 500.
Fig. 10 Cells from the central portion of the same ganglion from which figure
9 was drawn. Turtle embryo 10-mm. carapace length, iron-hematoxylin stain.
x 500.
Fig. 11 Cells from the gasserian ganglion of turtle embryo of 10-mm. cara-
pace length. Iron-hematoxylin. X 500.
Fig. 12 Cells from the third thoracic sympathetic chain ganglion of turtle
embryo of 10-mm. carapace length. Iron-hematoxylin. X 500.
Fig. 13 Cells from the ciliary ganglion of turtle embryo of 10-mm. carapace
length. Cells of smaller type indicated at A, and the larger type at B. Iron-
hematoxylin. X 500.
Fig. 14 Cells from the sphenopalatine ganglion of turtle embryo of 10-mm.
carapace length. Iron-hematoxylin. X 500.
Fig. 15 Cells from the peripheral cell clusters of the nervus terminalis of a
turtle embryo of 10-mm. carapace length. Iron-hematoxylin. X 500.
Fig. 16 Cells from the central root clusters of the nervus terminalis of a turtle
embryo of 10-mm. carapace length. Iron-hematoxylin. > 500.
438 O. LARSELL
yet reached the same degree of development. This view appears
to be favored by the extremely crowded condition of these cells
within the ganglia. This would appear to result in a reduction
both of the amount of space and of nourishment which the indi-
vidual cell may obtain, thus retarding its growth. The fact
that the larger cells were found near the periphery of the ganglia,
where there would appear to be space for greater expansion of
the cells during growth as well as more abundant nourishment,
may account for their larger size at this stage of development.
There were also cells of intermediate size between the two
groups, but these were not so numerous as the cells of either
group.
In the gasserian ganglion (fig. 11) of the older stages the number
of small cells was not so great as in the spinal ganglia, but here
also some were present, as the figure indicates.
The sympathetic chain ganglia of the older stages, as in the
younger, contained cells of rather uniform size and appearance (fig.
12). Two or three processes were observed on most of the cells
which were examined on this point. The nuclei were smaller than
those of the small spinal ganglion cells, showing an average size
for fifty cells of 6.2 u, as compared with 7.3 » for the latter type.
The relative amount of cytoplasm surrounding the nucleus ap-
peared to be about the same in the two types. No difference in
the size of the peripherally located cells, as compared with those
situated nearer the centers of the sympathetic ganglia, was
observed.
The ciliary ganglia showed two groups of strongly contrasting
cells. Without entering into a detailed description of this
ganglion or attempting to review the large amount of litera-
ture which has accumulated concerning it, the present pur-
pose will be served by calling attention to Carpenter’s (’06)
excellent study of it in the chick and adult fowl. He finds two
distinct sizes of cells. The smaller cells were arranged in a
definite group on the dorsal side of the ganglion. This group
composed about one-third of the entire mass.
A similar group of small cells (fig. 13, 4) was present in many
of the turtle embryos, but not in all, studied by the present
NERVUS TERMINALIS: TURTLE 439
writer. In all cases, however, whether or not arranged in a
distinct group, small ganglionic cells were found in the ganglion.
These differed not only in size, but also in form, from the larger
more typical cells. Carpenter considers the large cells in the
chick to be derived from the midbrain and to have migrated
along the oculomotor nerve to the ciliary ganglion. The small
cells he believes to be sympathetic cells which have migrated
forward along the ophthalmic nerve.
Fifty of the large cells were measured. The largest had a
nuclear diameter of 12.4 yu, the smallest of 7 u. The average
nuclear diameter of the fifty cells was 9 ». While the size of
these nuclei approached that of the large spinal ganglion cell
nuclei, and the largest found in the ciliary ganglion even ex-
ceeded the largest observed in the spinal ganglia, the amount of
cytoplasm surrounding the nuclei of the ciliary ganglion cells was
considerably less, as may be seen by comparing figures 9 and 13,
B. The actual size, therefore, of the ciliary ganglion cells was
somewhat less than that of the large spinal ganglion cells.
Twenty-two cells of the smaller type were measured. These
were found in several adjacent sections, forming a distinct mass
in the embryo on which the measurements were made. The
largest of these cells had a nuclear diameter of 7.9 uw. The
smallest was 5.3 », and the average of the twenty-two was 6.4 u.
As figure 13 (A) indicates, there were also differences in the form
of the cells and in the amount of cytoplasm surrounding the
nucleus, as compared with the large cells. .
There remain the cells of the sphenopalatine ganglion, which
form a relatively small cluster. These cells are of quite uniform
size and structure (fig. 14) with spheroidal, eccentrically located
nuclei. Most of them were in the primitive bipolar stage of dif-
ferentiation. Except that there was less individual variation
among these cells than among those found in the clusters of the
nervus terminalis, to be described, the sphenopalatine cells may
be said to resemble the terminalis cells more closely than any of
the other ganglionic cells studied. Of fifty sphenopalatine cells
which were measured, the largest had a nuclear diameter of
8.1 ». The smallest was 5.3 », and the average for the fifty
nuclei was 6.74 yu.
THE JOURNAL OF COMPARATIVE NEUROLOGY, VOL. 30, NO. 5
440 O. LARSELL
With these measurements as criteria, a comparison may be at-
tempted between the ganglionic cells of the nervus terminalis and
these other cells of well-recognized function, with the purpose in
mind of throwing more light on the relationship of the nervus
terminalis.
Many of the terminalis cells (figs. 15 and 16) have a peculiar
elongated form not met with elsewhere in the ganglia which were
subjected to observation. This is well illustrated in some of the
cells represented in figure 15. Only two processes were observed
in any of these cells, and these were in most cases continuous
with the longer axis of the cell. In view of the findings of
McKibben (714) in the dogfish and of the various authors who
have made observations on the ganglion cells of the terminalis
in the mammals, it seems likely that this bipolar condition is
developmental in the turtle.
- The nuclei were elongated in many cases and many were also
considerably distorted otherwise, as the figures indicate. This
was more frequently the case in the peripheral clusters than in
those within the cranial cavity. These conditions made the
terminalis cells more difficult to measure, so that the results
obtained represent the mean of several measurements on many
of the nuclei. In order to determine the difference in size, if
any, between the cells which were located in the ganglionic
swelling near the central roots of the nerve and those along the
olfactory and vomeronasal nerves outside the cranial cavity,
the measurements were tabulated separately for the two groups.
Thirty cells from the centrally located ganglion (fig. 16) indicated
an average nuclear diameter of 6.72 uw, the largest being 8.4 u,
and the smallest 5.3 « (table 1). The fifty cells of the peripheral
clusters (fig. 15) which were measured indicated an average
nuclear diameter for the entire number of 6.7 , practically the
same as that of the centrally located cells. The largest nucleus
of the peripheral clusters was 8.8 » in diameter, the smallest
was 5.3 uw. alek.ot | 255.2 310.3
0.178 154765) 29. 15S 2505. 228.9), Gb.05 73.67
0.000 0.000 | 0.169 1.590 | 2.266 13.479 | 27.207
0.027 0.321 W017 | Selssul 2744 19.718 19.621
0.099 1.154 | 7.453 | 9.906} 5.666 | 40.058 | 50.604
0.304 2.951 | 17.774 | 25.896 | 25.8638 | 138.23 171.02
0.644 4,736 | 21.439 | 31.973 | 30.730 | 91.3805 103.65
0.263 2.106 | 11.116 | 13.485 | 9.811 13.892 | 24.363
0.171 1.440 | 5.362 | 6.228} 5.180 11.633 11.431
0.434 3.546 | 16.478 | 19.663 | 14.991 25.525 | 35.794
0.0005) 0.0060} 0.0207; 0.0627} 0.0620) 0.3870) 0.3880
0.0044} 0.0333) 0.1800} 0.2230} 0.3310) 0.6051) 0.4973
0.0026) 0.0214) 0.1007} 0.1456) 0.0814; 0.1898} 0.2623
0.0002) 0.0024) 0.0192} 0.0270) 0.0514; 0.0236) 0.1011
0.0077} 0.0631); 0.2706} 0.4583) 0.5248 1 555 1.2487
483
ADULT
67 YEARS
1297.9
1020.0
278.0
81.93
21.081
17.296
34.120
154.27
98.00
12.639
13.018
25.657
0.3453
0.7938
0.1767
0.0365
1.3525
484 C. G. MACARTHUR AND E. A. DOISY
TABLE 12—Continued
FETUS FETUS CHILD CHILD CHILD ADULT ADULT ADULT
3 7 1 3 8 21 35 67
MONTHS MONTHS MONTH MONTHS MONTHS YEARS YEARS YEARS
Lipin
phos-
phorus .| 0.0075} 0.0643) 0.3747) 0.5020) 0.6206) 2.0980} 3.2550) 3.2360
Protein
phos-
phorus .| 0.0043} 0.0102) 0.0319} 0.0533) 0.0591} 0.1775) 0.1871} 0.1782
Organic
phos-
phorus .| 0.0044) 0.0333} 0.1639) 0.2969} 0.1600) 0.1489} 0.6054) 0.1262
Inorganic
phos-
phorus .| 0.0096} 0.0738} 0.3758} 0.5415) 0.2926) 0.7241) 0.6356) 0.7691
Total phos-
phorus. ..| 0.0258] 0.1821) 0.9463) 1.2836) 1.1323} 3.9585} 4.6831) 4.3095
Figure 1 was plotted from the data in this table relating to the earlier period of
growth.
CHEMICAL CHANGES IN HUMAN BRAIN 485
TABLE 13
Whole brain: Milligrams added per day
eae | Cate | Seoees | spine | Pacis | ore
3-MONTH | 7_yonTH | 1-monTH | 3-MONTH | 8-MONTH we
REIS FETUS CHILD CHILD cuitp | 2/-YEAR
Whole brain...............| 190.0 | 848.0 | 3764.0 | 2127.0 | 501.6 | 131.2
RRP on ca eos, con's al. 174.0 | 766.0 | 3270.0 | 1763.0 | 417.0 | 113.0
PA, ees eis sd = axa ciao oa 15.3 82.3 494.0 | 364.0] 84.6 18.2
PHOspHatids,. ..... 3... 2 1.98 | 10.80 85.3 34.9 | 26.3 5.4
WEreDrOsIdes... 20.5... i 6. 0.0 0.0 1.88 23.7 4.07 1.4
BON SIIOS, 424.005 06-0}. O90 2.45 7.73 35.4 2.99 2.2
REO RUEE Ob.0: 5. $6.0 Sun s.0 oe « 1.10 8.79 70.0 40.9 0.74 4.4
Marana cs cc) eee | a 162 er ab Tae ts 4
Watal proveins. 2.0.5.3 6..2) U7216 1 9Sae 185.6 | 175.6] 38.5 5.1
Organic extractives......| 2.81 | 15.4 100.1 38.7 7.2 | —0.6
Inorganic extractives.... 1.90 | 10.6 43.6 14.4 5.2 | —0.3
Total extractives........| 4.81 | 26.0 143.7 53.1 | 11.4 | —0.3
Lapin.sulphur.........<...| 0,006) 0,05 0.16 0.70} 0.08 0.04
Protein sulphur. ....3..: 0.049) 0.24 1.07 1.55} 0.61 0.0
Neutral sulphur.......... 0.029) 0.16 0.88 0.75} 0.01 0.0
Inorganic sulphur........| 0.002} 0.02 0.19 0.13) 0.11 | —0.01
Doral sulphur: ....22.......)" 0.086) 0:47 2.30 3.13} 0.81 0.03
Lipin phosphorus........ 0.083} 0.47 3.45 2.12} 1.01 0.27
Protein phosphorus...... 0.048) 0.081 0.24 0.36; 0.09 0.01
Organic phosphorus...... 0.049) 0.24 1.45 2.221 0.0 | —0.02
Inorganic phosphorus....| 0.107) 0.54 3.36 0.93) 0.17 0.03
Total phosphorus.......... 0.287; 1.30 8.50 5.62) 1.27 0.29
A part of these data are plotted in graph 2.
TABLE 14
Whole brain: Average percentage increase per day
3-MONTH | 7-MONTH
FETUS FETUS 1-MoNnTH | 3:
MONTH | 8-MONTH
Wihole brainer (4 tascey- ces erick ores
Dao igs 0.88 0.26 0.028
Witenes cere eee oer emcee ae UAT 0.88 0.23 0.04
Solids 4:2 Satie scien ae eee ane 2.4 her 0 0.36 0.046
Phosphatids) ae OR Pp eae (LC | 1.59 0.62 0.093 | 0.014
Proteins. . RA Rees A Ae ES he ee a 1.42 0.66 0.087 | 0.007
Organic SEU Teh eee lee 1.52 0.32 0.046 | 0.0
Inorganic Pein Verne ROTI ieearie hig! sa? 1.28 0.25 0.067 | 0.0
WotalextTactivieszecns. ch cece cecere|) NuoO 1.44 0.29 0.048 | 0.0
Calculated from data in table 11. The average number of milligrams added _
per day was divided by the average weight for the given period, instead of the
weight at the beginning of the period, as is usually done. When there are rapid
changes in weight, this method is not as accurate as that used in table 12. It
is believed that the temporary rise in growth at about the seventh month of
fetal life is due to the method of calculation.
The curve marked extractives (c) in graph 3 was plotted from the shove
data.
486
SUBJECT AND AUTHOR INDEX
CTIVITY of the nervous system. IIT.
On the amount of non-protein nitrogen
in the brain of albino rats during
twenty-four hours after feeding. Meta-
DO Lee eabinee tes ree aia eile wots: Siesta sila n'occ oc elevoe 397
Albino mouse. The nervus facialis of the.... 81
——— rat in Miiller’s fluid. Factors influ-
encing the behavior of the brain of the... 411
——— rats during twenty-four hours after
feeding. Metabolic activity of the nerv-
ous system. III. On the amount of
non-protein nitrogen in the brain of...... 397
Auten, Witu1AM F. Application of the Mar-
chi method to the study of the radix
mesencephalica trigemini in the guinea-
PARE MMOOPEE RE che Pek cicl atc ne'a sie p cuatlacs fapanione sieges 69
Axis, Epwarp Pueps, Jr. The ophthal-
mic nerves of the gnathostome fishes..... 69
Arry, Lresytiz B. A retinal mechanism of
iT TAT a CV) OS eee eee ee eee 343
Ayers, Howarp. Vertebrate cephalogenesis.
V. Transformation of the anterior end of
the head, resulting in the formation of the
“IEEE = SoReal AO ee Oe ae neers 323
EHAVIOR of the brain of the albino rat
in Miiller’s fluid. Factors influencing —
RUA st tint gl oaiels sels serrate sta ry lieinels 411
Brain during growth. Quantitative chemical
Charnes AN The MUMIAN: oo oso a,< cies eas oe les 445
——— of albino rats during twenty-four hours
after feeding. Metabolic activity of the’
nervous system. III. On the amount of
non-protein nitrogen in the.............. 397
—— of thealbinoratin Miiller’s fluid. Fac-
tors influencing the behavior of the...... 411
ELLS in normal, subnormal, and senes-
cent human cerebella, with some notes
on functional localization. A prelimi-
nary quantitative study of the Purkinje. 229
——— tunnel space, and Nuel’s spaces in the
organ of Corti. The development of the
PUNT: ARG E cS Be RD ie Ae eR 283
Cell with especial consideration of the ‘ Golgi-
net’ of Bethe, nervous terminal feet and
the ‘nervous pericellular terminal net’ of
Held. On the finerstructure of the syn-
apse of thesMawphner, ccc nes esate seen 127
Cephalogenesis. IV. Transformation of the
anterior end of the head, resulting in the
formation of the ‘nose.’ Vertebrate...... 323
Cerebella, with some notes on functional lo-
ealization. A preliminary quantitative
study of the Purkinje cells in normal, sub-
normal, and senescent human............ 229
Changes in the human brain during growth.
Quantitative chemical..................-.- 445
Chemical changes in the human brain during
ptowthn. . Quantitativesss. ec. tere. 445
Corti. The development of the pillar cells,
tunnel space, and Nuel’s spaces in the
OTE OL es are olatarsraisteie a UNtay e Oteele sete 283
| Be aes rec none of the pillar cells, tun-
nel space, and Nuel’s spaces in the
Organon Corti, Thess ives. o le 283
Doisy, E.A., MacArruur, C.G.,and. Quan-
titative chemical changes i in the human
brain during @rowth. i046 ).02i5..<0 0 foes ts 445
FFICIENT vision. A retinal mechanism
OLR shea ask, PUES ee Te Oe aL 343
Evuis,.Rosert §. A preliminary quantita-
tive study of the Purkinje cells in normal,
subnormal, and senescent human cere-
bella, with some notes on functional lo-
CHlazatOM etc. css chemother eee 229
ACIALIS of the albino mouse. The ner-
WLI Ss een Rite ene fad cetepete ate nters ik oriore ie ane 81
Factors influencing the behavior of the brain
of the albino rat in Miiller’s fluid......... 411
Fiber in teleosts. Concerning Reissner’s..... 217
Fishes. Theophthalmic nerves of the gnatho-
SUOWMG Hs alae seve insa piesa velpia'a ecaraverg oe eine ala aiata ates
Fluid. Factors influencing the behavior of
the brain of the albino rat in Miiller’s.... 411
Formation of the ‘nose.’ Vertebrate cepha-
logenesis. IV. Transformation of the
anterior end of the head, resulting in the.. 328
Functional localization. A preliminary quan-
titative study of the Purkinje cells in nor-
mal, subnormal, and senescent human
cerebella, with some notes on............. 229
(cena fishes. The ophthal-
NIG TOE VERIOL HE <2 5 aero ccc nat eens sears 69
Golgi. Frontispiece. Portrait of Professor
Canalo sine oki Peete t sei e niece aes Pore ee ree 168
‘Golgi-net’ of Bethe, nervous terminal feet
and the ‘nervous pericellular terminal net’
of Held. On the finer structure of the
synapse of the Mauthner cell with es-
pecial consideration of the................ 127
Growth. Quantitative chemical changes in
the human: Pram Gurney. «. sate vinias eee 445
Guinea-pig. Application of the Marchi
method to the study of the radix mesen-
cephalica trigemini in the................ 169
EAD, resulting in the formation of the
‘nose.’ Vertebrate cephalogenesis. IV.
erppetormation of the anterior end of
Held. sym the finer structure of the synapse
of the Mauthner cell with especial con-
sideration of the ‘Golgi-net’ of Bethe,
nervous terminal feet and the ‘nervous
pericellular terminal net’ of.............. 127
Human cerebella, with some notes on func-
tional localization. A preliminary quan-
titative study of the Purkinje cells in
normal, subnormal, and senescent........ 229
ORDAN, Hovey. Concerning Reissner’s
GHEMUBIEEIGOSER hes :s Sou casa sy leccmien a ie 217
OMINE, Suicryuxt. Metabolic activity
of the nervous system. III. On the
amount of non-protein nitrogen in the
brain of albino rats during twenty-four
INOUTH/AIter LEGGING: |... .. ss. secs eelerns =e 397
ARSELL, Otor. Studies on the nervus
terminalis: Mammals...................
Studies on the nervus terminalis:
GUTTAG Ra ere a ei ac Giclee Sayeiace Ws te etal 423
487
488
Localization. A preliminary quantitative
study of the Purkinje cells in normal, sub-
normal, and senescent human cerebella,
with some notes on functional............ 229
acARTHUR, C. G., and Dowy, E. A.
Quantitative chemical changes in the
human brain during growth.......... oo ET
Mammals. Studies on the nervus terminalis. 1
Marchi method to the study of the radix mes-
encephalica trigemini in the guinea-pig.
Applicationiofthe.2. icc. ce se asee-- 2 169
Maruti, Kryoyasu. On the finer structure of
the synapse of the Mauthner cell with es-
pecial consideration of the ‘Golgi-net’ of
Bethe, nervous terminal feet and the
‘nervous pericellular terminal net’ of Held. 127
——— The effect of over-activity on the mor-
phological structure of the synapse....... 253
Mauthner cell with especial consideration of
the ‘Golgi-net’ of Bethe, nervous termi-
nal feet and the ‘nervous pericellular termi-
nal net’ of Held. On the further struc-
ture of the synapse of the.............. vas, 127
Mesencephalica trigemini in the guinea-pig.
Application of the Marchi method to the
Bud yO the TAGIX.. 50. <2a5 +e eos 169
Metabolic activity of the nervous system.
III. On the amount of non-protein ni-
trogen in the brain of albinorats during
twenty-four hours after feeding........... 397
Morphological structure of the synapse. The
effect of over-activity on the............. 253
Mouse. The nervus facialis of the albino..... 81
Miiller’s fluid. Factors influencing the be-
havior of the brain of the albino rat in... 411
ERVES of the gnathostome fishes.
Ophthalmic vecses15 occ ve tee tee a eer 69
Nervoussystem. III. Ontheamount of non-
protein nitrogen in the brain of albino
rats during twenty-four hours after feed-
ing. Metabolic activity of the........... 397
——— terminal feet and the ‘nervous peri-
cellular terminal net’ of Held. On the
finer structure of the synapse of the
Mauthner cell with especial consideration
of the ‘Golgi-net’ of Bethe............... 127
Nervus facialis of the albino mouse. The.... 81
terminalis: Mammals. Studiesonthe 1
—— turtle. Studies on the................ 423
Net’ of Held. On the finer structure of the
synapse of the Mauthner cell with especial
consideration of the ‘Golgi-net’ of Bethe,
nervous terminal feet and the ‘nervous
pericellular terminal....................+. 127
Nitrogen in the brain of albino rats during
twenty-four hours after feeding. Meta-
bolic activity of the nervous system. IIT.
On the amount of non-protein............ 397
Non-protein nitrogen in the brain of albino
rats during twenty-four hours after feed-
ing. Metabolic activity of the nervous
system. III. On the amount of.......... 397
‘Nose.’ Vertebrate cephalogenesis. IV.
Transformation of the anterior end of the
head, resulting in the formation of the... 323
Nuel’s spaces in the organ of Corti. The de-
velopment of the pillar cells, tunnel
space, and
GATA, D., and Vincent, SwALe. A con-
tribution to the study of vasomotor
reflexes. 21, 5. o552 sane eeyl eee Peek ted 355
Ophthalmic nerves of the gnathostome fishes.
ests 1S ie Re ek = see nee 69
Organ of Corti. The ge pele of the
pillar cells, tunnel space, and Nuel’s
BDACEBIN CUE. : hot te.s eee tre ue ah 283
Over-activity on the morphological structure
of thesynapse. The effect of............ 253
INDEX
ILLAR cells, tunnel space, and Nuel’s
spaces in the organ of Corti. The de-
velopment of the, .......25<-2:s)e+- oon . 283
PLANT, JAMES Stuart. Factors influencing
the behavior of the brain of the albino
ratin Miller's fuid.:.. 2s,.-+- 7a 411
Purkinje cells in normal, subnormal, and
senescent human cerebella, with some
notes on functional localization. A pre-
liminary quantitative study of the....... 229
R225 mesencephalica trigemini in the
guinea-pig. Application of the Marchi
_ Inethod to the study of the............. 169
Ratin Miiller’s fluid. Factors influencing the
behavior of the brain of the albino....... 411
Rats during twenty-four hours after feeding. +
Metabolic activity of the nervous system.
III. On the amount of non-protein ni-
trogen in the brain of albino............. 397
Reflexes. A contribution to the study of
VASOMOLOL:... asks csen ashe +s cee eee 355
Reissner’s fiber in teleosts. Concerning...... 217
Retinal mechanism of efficient viaion. A.... 343
RuineHart, D. A. The nervous facialis of
theialbino MOuses..4. hi >- seen eee
PACE, and Neul’s spaces in the organ of
Corti. The development of the pillar
cells; tunnel e710 250 ee oe eee 283
Spaces in the organ of Corti. The develop-
ment of the pillar cells, tunnel space, and
Nuelist. cece ou bee mak wean eee 283
Structure of the synapse of the Mauthner cell
with especial consideration of the ‘ Golgi-
net’ of Bethe, nervous terminal feet and
the ‘nervous pericellular terminal net’ of
Held: On the: finer. s3.b-<-ee tee eee 127
Structure of the synapse. The effect of over-
activity on the morphological............ 253
Synapse of the Mauthner cell with especial
consideration of the ‘ Golgi-net’ of Bethe,
«nervous terminal feet and the ‘nervous
pericellular terminal net’ of Held. On
the finer structureiof the) 7.2 /2-.- oes 127
The effect of over-activity on the mor-
phological structure of the................ 253
System. III. On the amount of non-protein
nitrogen in the brain of albino rats during
twenty-four hours after feeding. Meta-
bolic activity of the nervous.............. 397
ELEOSTS. Concerning Reissner’s fiber
LTE 2 . teae = folk. aac: ee 217
Terminalis: Mammals. Studiesonthenervus 1
turtle. Studies on the nervus......... 423
Terminal net’ of Held. On the finer struc-.
ture of the synapse of the Mauthner cell
with especial consideration of the ‘Golgi-
net’ of Bethe, nervous terminal feet and
the ‘nervous pericellular..............-.20 127
Transformation of the anterior end of the
head, resulting in the formation of the
‘nose.’ Vertebrate cephalogenesis. IV.. 323
Trigemini in the guinea-pig. Application of
the Marchi method to the study of the
radix mesencephalica..-......c.. 1c. ocetee 169
Tunnelspace;and Nuel’sspacesin the organ of
Corti. The development ofthe pillarcells 283
Turtle. Studies on the nervusterminalis.... 423
AN DER STRICHT, O. The develop-
ment of the pillar cells, tunnel space,
and Nuel’s spaces in the organ of Corti. 283
Vasomotor reflexes. A contribution to the
SHUdVIO! So. cu ol Ss he ees La ee 355
Vertebrate cephalogenesis. IV. Transforma-
tion of the anterior end of the head, re-
sulting in the formation of the ‘nose’:... 323
Vincent, SWALE, Ocata, D., and.
*
48
+
K)
.
*,
4
stats
4
*
i
.
<
ys -
ot «
ts aes”
eat. 4 2
Pe 55 i
state
Satitete
at
‘%
4
re ae
" Rot Oe © ee ee
ARS S
ey
AA
?
¥
os 56
~
4s
riers tate.
a
ae
+s *
#5
a
a
$a°,*
Ay
24
»~
&
be
Perry
toes,
‘
.
ee a
ta .*.
PX)
$$ 2-4
1 .