COLUMBIA LIBRARIES OFFSm

HEALTH SCIENCES SANDARD

Mynnn4i807

^■'^^^^'^

, '^ ''-.Hi '">■

v##'

•*.j«'^-0 .

% h'

'^^

>if»^'V

f- % >

. 4.:

\ *

V4^

f5 J^.

--' V -li

V V^-^^l-^>

-^ J .>;:;

\.jry^J^:;.

'.-^

r 267 ]

VI. Croonian Lectuhe. — On the Maimaalian Nervous System, its Functions, and
their Localisation determined by an Electrical Method.

By Francis Gotch, Hon. M.A., Oxon., and Victor Horsley, F.R.S., &c.

{From the Physiological Lahoratory of the University of Oxford.)

Received and Read Febmnry 26, 1891.

[Plates 29-35.]
Contents.

Chap. P.ige-

I. Introduction, inclnding the authors' method and plan of research 267

II. Historical 271

III. Detailed description of methods. 1. Anassthesia. 2. Operative procedure.

3. Apparatus. 4. Method and precautions 280

IV. The resting electrical difference in the Mammalian nerve and spinal cord . . 302
V. On the electrical effects evoked in the Spinal Cord and Mixed Nerve by

excitation of the Cortex Cerebri 321

VI. On the electrical effects evoked in the Spinal Cord and Mixed Nerve by

excitation of the Corona Radiata 336

VII. On Bilaterality of representation as evidenced by the electrical changes in the

Spinal Cord and Mixed Nerves 342

VIII. On the electrical effects evoked in the Spinal Coi-d by localised excitation of

its different parts 363

IX. On the electrical effects evoked in the Spinal Cord by excitation of the

Lumbar Nerves 410

X. On the electrical effects evoked in the Lumbar Nerves by localised excitation

of the Spinal Cord 446

XI. On the Functional Activity of the Nerve Centres in the Spinal Cord and their
relations to nerve fibres in the same as determined by the galvanometric
method 478

XII. On the electrical effect evoked in the Spinal Cord and Mixed Nerves by

Absinthe and Strychnia 511

XIII. Summary and Conclusion 515

Description of Plates 526

CHAPTER I.— INTRODUCTION, INCLUDING THE AUTHORS' METHOD AND THE PLAN

OF THE RESEARCH.

In the ' Proceedings of the Royal Society,' vol. 45, 1889, p. 18 (Meeting of November 1,
1888), we published a preliminary account of some of the experiments of wliich the
results are now given in full detail.

MDCCCXCI. — B. 2 M 2 ] 5. 10. 1)1

2(38 MESSRS. F. GOTCH A^M) V lloltSIJlV

III that coinimuiicatiou we stated tliat tlie object of our work tlieii was to
endeavour to ascertain the character of the excitatory processes occurring in nerve
fibres, when, either directly (artificially) excited, or when in that state of
functional activity, which is due to the passage of impulses along tliem from the
central apparatus. The most important way in which such a method could be
applied was obviously one which would involve the investigation of the excitatory
changes occurring in the fibres of the spinal cord when the cortex cerebri is stimu-
lated. We must at once assume that the motor side of the central nervous system is
practically divisible into three elements: — 1. Cortical centres. 2. Efferent (pyramidal
tract) fibres leading down through the internal capsule, corona radiata, and sj)ina,l
cord. 3. Bulbo-spinul centres contained in the medulla and the spinal cord, and
forming the well-known nuclei of the cranial and also of the spinal motor nerves.

It had already been determined, both by direct observation and by the graphic
method (1) that certain areas of the cortex were connected with definite movements
of various parts of the body, and (2) that while the complete discharge of the cortical
apparatus was followed by a very definite and characteristic series of contractions of
the muscles in special relation with the particular point excited, the effectual removal
of the cortical central mechanisn and subsequent excitation of the white fibi'es passing
down through the internal capsule, &c., led to the production of only a portion of the
effect previously obtained from the uninjured brain.

This method of observation in no wise showed \vhat processes were actually
occurring in the spinal and other nerve fibres, and although the ablation of the
cortical centre, to a certain defjree sucrcrested the extent to which the cortex acted,
nevertheless, it did not afford an exact demonstration of the same. Moreover, the
data which the graphic method furnished were precluded, through their being
muscular records, from determining what share, if any, the lower bulbo-spinal central
nerve cells took, either in the production of the chai'acteristic sequence of contractions
or in the modification, whether in quality or in force of the descending nerve impulses
during their transit.

It seemed to us that the only way to approach this subject would be to get, as it
were, between the cortex and the bulbo-spinal system of centres. This would be
accomplished if some means were devised of ascertaining the character of the
excitatory processes occurring in the spinal fibres of the pyramidal tract, when upon
excitation of the cortex, nervous impulses were discharged from the cortical cells and
travelled down the cord.

The question as to the extent to which it is possible to obtain physical evidence of
the actual presence in nerve fibres of excitatory processes and thus to ariive at
reliable data for the comparison of their amounts is one which, up to the present, has
been answered only indirectly, and that in two ways, firstly, by the extension of
Helmholtz's classical experiment of determining the rate of transmission, and
secondly, by observing those variations of electrical states in nerve fibres which

ON THE MAMMALIAN NERVOUS SYSTEM. 269

DU Bois-Rf-ymond discovered to be invariable concomitants of excitatory processes.
As will subsequently be shown in the historical retrospect, it is well known, through
the researches of du Bois-Reymond and others, that the fibres of the spinal cord,
just as nerve fibres in the peripheral trunks, are characterised by show^ing when
unexcited an electrical difference between their longitudinal surface and cross-
sections, and furthermore, that when excited a well-marked diminution of this
restinsr electi-ical state is produced in the fibres of the cord as in those of nerve
trunks.

Now since such excitatory variations in the electrical state are presumably parallel
in time and amount with the presence in the nerve of the unknown processes termed
excitatory, which a, series of stimuli evokes, it Avas reasonable to presume, if the
cortex were discharging a series of nerve-impulses at a certain rate down the
pyramidal tract, that there would be a series of parallel changes in the electrical
condition of the fibres in the cord tract, and that with a suitable apparatus for
responding to such changes these might be both ascertained and recorded.

If this could be done, then the character of the discharge of the cortical centre
into the spinal cord would be, for the first time, definitely ascertained.

As has been said before, the graphic method to some extent suggested the solution
of the problem, but the graphic method could not exclude, as this newer mode of
investigation does, the bulbo-spinal centres. Judging from the rate of contractions
of the muscles convulsed by excitation of the cortex, it was reasonable to expect that
the variations in the electrical condition of the pyramidal tract might intermit fifteen
or twenty times per second. If, therefore, they were to be observed or recorded, it
was obvious that some instrument would have to be used capable of quickly respond-
ing to very minute electrical differences succeeding one another at very short intervals
of time. The only instrument available for this task is Lippmann's electrometer,
and, as we have stated in our previous communication, we had the advantage of the
assistance of Mr. G. F. Burch in obtaining several very sensitive instruments.

This instrument in addition presents the invaluable advantage of its movements
being easily recorded by photography, as originally described by BuRDON Sanderson
and Page. Such records are given in our previous paper in the ' Proceedings ' ; they
were from the first so definite, and so constant, that they enabled iis at once to pass
on to a further development of the same method. It occurred to us that the method
aftbrded a means, not only of discovering the rhythm of the nerve disturbances as
they pass along the spinal channels, but also of investigating the line of communica-
tion existing between separated nerve centres, the mode of discharge of such centres,
and the determination of the direct paths, whether afferent or efferent in the
structure of the central nervous system. It is this last wider application of the
method to which we wish particularly to draw attention in the following pages,
inasmuch as although it has hitherto been possible by means of division or ablation
of certain portions of the central nervous system, e.g., the spinal cord, to trace by

-70 MESSRS. V. (lorcil AN'l) W lloUSLHV

the loss of function tlie probable patlis of transmission of afferent or efferent
impulses, yet such a mode of experimentation is always open to very grave
disadvantages and sources of fallacy. Sucli, for example, is the always recurring
possibility of functional changes proceeding beyond the desired lesion, &c.

The accomplishment of this further purpose, viz., the localisation of both paths and
centres by ascertaining the excitatory electrical effects in relation with them, was one
of the main objects we had in view.

In carrying it out we found it was unnecessary to employ the electrometer, and,
in fact, that it was advantageous to use the galvanometer, the record of which could
be more easily and more accurately noted, since its graduation admits of far higher
magnification. Moreover, witli tliis instrument it was possible, by employing a series
of stimuli, of known number and duration, to obtain quantitative results of definite
comparative value, as will be shown further on ; and thus to compare the electrical
effects evoked (1) in diffei^ent central paths by direct stimulation of these, and (2) in
any one path by excitation of different regions.

The plan upon which the present paper is framed is, first, to give a historical
retrospect of the work of authors who have opened up tlie study of electrical
changes in the central and peripheral nervous system ; second, to describe at length
our mode of experimentation, with special reference to the modifications which
we have introduced, then to compare roughly the results we have obtained by our
present method with those which had been previously ascertained by the graphic
method, and so introduce the description of the facts which we have discovered,
elucidating the physiology of the spinal cord both in its relation to the higher centres
and to the peripheral nerves. In describing the detailed results of our experiments
we have found that it is difficult, from the extent of ground covered by the subject,
to adopt a convenient arrangement of facts, and mode of describing the same, which
are free from the fault of repetition. We finally determined to gather the results
together into definite groups, each associated with the excitation and investigation of
particular regions, and refer in the briefest manner to any important principles which
were not directly elucidated by, but only involved in, the particular experiments
under consideration. Finally, we give at length a summary of the general conclusions
which we consider we are justified in drawing from our experimental results.*

* We here wish to express oui- great indebtedness to Professor Burdon Sanderson for placing at our
disposal the instruments and equipment of the Physiological Laboratory of the University of Oxford.
We are also under especial obligations to Mr. G. F. BuRCH, whose constructive skill provided us with
the requisite electrometers, and who very kindly gave his assistance in preparing our photographic
records.

Finally Dr. Howard Tooth has very obligingly carried out the laborious task of the microscopical
investigation of the spinal cords in those cases where we performed the section of columns antecedently
to the special experiments of the present research.

The expenses of the present investigation have been defrayed by grants from the Scientific Grants
Committee of the Royal Society, and from the Scientific Grants Committee of the British Medical
Association.

ox THE MAMMALIAN NERVOUS SYSTEM. 271

CHAPTER II.— HISTORICAL.

We consider it advisable to give a brief sketch or rather catalogue of the chief
facts which have been ascertained concerning the specific function of nerve centres,
and the conductivity of nerve fibres, so far as the central nervous system is concerned,
in order that the various points we raise in the rest of the paper may be rendered
more intelligible when the results obtained by our method are contrasted with those
discovered by other means. It is naturally impossible for us to give on the present
occasion a complete history of this vast subject ; we would only, therefore, allude
to those points towards the further elucidation of which we have directed the
present research. The procedures employed by various authors with the exception
of the galvanometric method may be enumerated as follows : —

1. Electrical and otlier stimulation with direct observation of the phenomena
evoked.

2. Stimulation with graphic record of muscular and other movements produced.

3. The method of anatomically observing the degeneration of nerve fibres consequent
upon their separation from nerve centres with which they are in functional relation.

4. The erabryological method of observing the development or differentiation of
tracts and fibres.

Of the foregoing methods, the first was the one by means of whicli the principles
of localisation were earliest determined, and in this connection it is scarcely necessary
to do more than allude by name to Hitzig and Fritsch, Fereier, Munk, Luciani,
Albertoni, and others. The results tliey obtained are to be classed with those
gained by means of the second method of recording muscular and other movements,
employed in the analysis of the functions of the central nervous system for the
first time by Francois Franck and Pitres, and in the subsequent investigations
of BuBNOFF and Heidenhain. They are open to the same objection, viz., that since
they involve motor function they necessarily exhibit the activity of two sets of central
mechanisms, and that therefore they only indicate the functions of the paths which
run in the central nervous system in so far as these functions are modified by those
paths being intimately connected with the lower or bulbo-spinal centres.

The degeneration method, in which localisation of a nerve jjath is accomplished by
means of studying th(j retrograde changes which nerve fibres undergo when they are
cut off from the lowest nerve centre, with which they are in relation, should they
happen to be paths for ascending transmission, or vice versd from the highest nerve
centre fur descending transmission was initiated by TiiRCK and followed up more
especially by Charcot and his pupils, as well as by a large number of investigators,
and neuro-pathologists of fill countries. This method is not free from possible
error, since in the case of nerve channels connected at eacli end with central
structures, we do not yet know what the nature of the connection must be wliich
enables the nerve channel to successfullv resist degeneration. Therefore, while the

272 MESSRS. F. GOTCll ANU V. HOliSLKY

existence of degeneration furnishes us with positive evidence as to the presence of a
direct path, its absence does not exclude the existence of such a path.

The embryological method, which afibrds a very fair control of the degeneration
method, was instituted by Fle(.'US1G and has since been extended by Bechterew and
others. It gives a fair indication of the proportionate number or quantitative relation
of the elements which go to make up any one channel, whether direct or indirect,
but beyond this, and the all important localisation of position, it does not carry us.
By a special histological method, introduced by Golgi, the anatomical relation of the
paths ill the developing cord have recently received a fuller demonstration, at his
hands as well as Kolliker's, Ramon y Cajal's and others.

It will be convenient to tabulate the facts according as they relate to

I. Centres.

II. Paths.

I. Centres.

Functional Activity of Centres.

As regards Centres, tlie position assumed in the introduction" is, we believe, as a
matter of fact, tacitly in the minds of observers, if not admitted, and the influence of
" basal centres" is not so overwhelmingly important as originally supposed by the
older observers. To conveniently combine, therefore, the results without doing
violence to the theoretical views expressed by anyone, it is only necessary to arrange
the facts in relation to the part wdiich is the seat of experimental investigation.

A. Cortex. ,

(1.) Latent Period. — The loss of time which intervenes between the application of
the stimulus to the cortex and the commencement of the resulting muscular contrac-
tion is -OG on the average in the Carnivora. (Fhanck and Pitres, Bubnoff and
Heidexhain, Schafer, ourselves.)

(2.) ExcitahiUti/. — This property of the cortex is greatly altered, i.e., either in-
creased, or diminished by —

(a) Severe haemorrhage ; (h) deep anaesthesia ; (c) cooling ; (d) drying ; (e) fatigue ;
( /) peripheral stimulation of the functionally corresponding and other parts of the
body. (Bubnoff and Heidenhain, Bozzard, Exner.)

The latent period may vary from the effect of any of these causes as well as from
the intensity of the stimulus. (Bubnoff and Heidenhain and other authors.)

* Viz., that the plan of the central nervous system, as regards its " motor" side, consists in a cortical
mass of grey matter, and a corresponding mass in the bulbo-spinal part of the central neural axi.«,
with segmental representation of function, while that finally the.se two great central apparatuses are
connected by the direct paths known as the pyramidal tract or excitable fibres of the corona radiata
internal capsule, and, we must add, of the lateral column of the spinal cord. According to this view
the basal centres and cerebellum act on the direct motor apparatus, if at all, as reinforcing mechanism.s.

ON THE MAMMA.LIAN NERVOUS SYSTEM. 273

(3.) Mode of Discharge. — The impulses generated in the cortex, and which pass
down to the muscles may, according to the duration and intensity of the stimulus, be
of the nature of {a) single discharges producing one muscular contraction ; or {h) com-
pound discharges producing tonic muscular contraction, or (c) a combination of tonic
and clonic contractions. (Fritsch, Hitzig, Ferrier, Munk, Franck, and all other
authors.) This last combination is to be looked upon as the complete discharge of the
cortex, vide infra and pp. 345, 349, &c.

(4.) Nature of the Discharge. — (a.) The single muscular contraction is more pro-
longed, and ceases more gradually than that ehcited by a single stimulus applied to
the motor nerve. (Franck, confirmed by ourselves.)

(6.) The tonic contraction is regarded by most authors ns a fusion of contractions
evoked by many discharges. (Franck, Schafer and Horsley ; vide infra. " Corona
Radiata")

(c.) The tonic and clonic contractions occur in the order mentioned, and are to be
regarded as tlie muscular response to a complete coi'tical discharge, i.e., comprising a
primaiy effect and after-effect.

{d. ) The rate or rhythm with which these muscular responses* appear to succeed
one another has been variously estimated by different observers in different animals
and with different instruments, and has been ascertained to be from 8 to 10 per
second.

(5.) Relation of the Discharge to the Parts of the Body. — {a.) Localisation of the
representation of the gross divisions of the body to definite areas of cortex. (Fritsch
and Hitzig, Ferrier, Munk, Luciani, Schafer, Beevor, Horsley.)

(6.) Localisation of the representation of segments of the gross divisions to definite
areas of cortex. (Beevor and Horsley.)

(c.) Localisation of the representation of the character of the various movements of
segments to definite areas of cortex. (Beevor and Horsley.)

B. Spinal Co)'d.

Under this heading we will group the jihenomena associated with the functional
activity of the system of bulbo-spinal centres, i.e., those in which the effei-ent paths
terminate. It may not be superfluous to add that the information furnished by the
graphic method does not differentiate the complex structure of a bulbo-spinal centre,
and that it can only yield a record of the combined action of mainly afferent and
mainly efferent corpuscles in the posterior and anterior divisions of the grey matter.
All methods of observation hitherto employed involve the activity of the whole
apparatus, and this must be borne in mind in considering the following facts.

* The tracing waves indicating the contractions are frequently summated. (Hoesley.)
Quite recently the value of these waves as indications of rhythmical nerve discharges has been con-
tested. (Wedenskii, Haycrait.)

MDCCCXCI. — B. 2 N

274 MP]SSRS. ¥. GOTCII AND V. IIORSLEY

( 1 .) Latent Period. — The interval of time occupied in the passage of a nerve iiii])ulse
from the afferent to the efferent side of a (reflex) bulbo-spinal centre is given on
p. 481, to whicli reference is directed. (Wunut, Exner, and others.)

(2.) Excitability. — This function of the grey bulbo-spinal matter is modified by the
same causes as those which influence the Cortex, see pp. 272 and 483.

When the excitability of the cord is raised, e.g., that of the distal segments after
section, these latter may discharge in a coordinated fashion, see p. 423, as was first
observed by Schiff.

(3.) Mode of Discharge. — The bulbo-spinal centres, like tlie cortex, present three
modes of discharge, but not in the same degree, e.g., the combined sequence of tonus
and clonus though sometimes present, see j^p. 483-499, nevertheless occurs but rarely.
Further, as evidenced by muscular responses, the bulbo-spinal centres appear to dis-
charge at regular intervals under certain conditions of isolation and excitation, e.g.,
ankle clonus (V.H.). For further detail see p. 483.

(4.) Nature of the Discharge, (a.) As in the case of the cortex cerebri tlie charac-
ter of the muscular response to excitation of the cord is different from that seen when
the excitation is directly applied to the nerve.

(h.) The tonic contraction is usually developed more rapidly than that obtained
from the corona radiata.

(f.) The rhythm of the intermittent muscular contractions, fused or not, is ft'om
8 to 1 0 per second or frequently the early multiples of this rate.

(5.) Relation of the Discharge to the Parts of the Body. — (a.) Localisation of the
representation of the gross divisions of the body to difierent regions of the bulbo-
spinal appai'atus is well marked. (All authors.)

if.) Localisation of the representation of the segments of the gross divisions and tlie
character of their respective movements to different root-origins in the grey matter.
(Fereiee and Yeo, Schiff, Forgue, Beevor.)

IL Paths.

Functional Activity of the Efferent Paths.

The graphic method only permits of a limited analysis of paths, i.e., fibres, inasmuch
as there is of necessity included the bulbo-spinal system of centres for the purpose of
giving the muscuhir contractions used for record. Hence no result bj the graphic
method can be regarded as pure ; moreover, it only furnishes information upon the
functions of efferent paths.

In fact, there is no direct evidence forthcoming from any method of observation,
except that of the galvanometer, to prove that the paths themselves in the cord
are excitable. Indeed, it has long been held by many physiologists that all eti'ects
obtained from stimulation of the spinal cord are due to primary and progressive

ON THE MAMMALIAN NERVOUS SYSTEM. 275

excitation of reflex centres (Schiff, Chauveau). Consequently, it must be under-
stood that the following enumeration of facts is written and arranged in the light of
our experimental results, set forth in the present paper.

A. Corona Radiata.

(1.) Latent Period. — The loss of time intervening between the moment of appli-
cation of the stimulus to the corona radiata and the commencement of the resulting
muscular contraction is usually "04 sec. (Franck and others.)

(2.) Excitahility. — -The excitability of the fibres rapidly fails upon their exposure.
VuLPlAN showed that relatively the excitability of the corona radiata was higher
than that of the cortex. In this he was confirmed by the observations of Boche-
FONTAINE, CouTY, AscH and Neisser, and others.

(3.) Mode of Discharge. — The muscular contraction may be (a) a single twitch, or
{h) a tonus. The combination we have termed the complete cortical discharge, and
consisting of tonus and clonus, is never seen where the cortex is absolutely removed.
(Franck, Bubnoff and Heidenhain in part, Horsley, Schafer and Horsley.)

(4.) Nature of the Discharge. — (a.) The tonus observed is regarded as a fusion
of muscular responses which commence and end sharply with the beginning and end
of the excitation.

(h.) The rate of recurrence of the individual responses is unknown.

(5.) Relation of the Discharge to the Parts of the Body. — Localisation of gross
divisions of the body to certain fields of fibres issuing from cortical centres (all authors
named above).

B. Internal Capsule.

(1.) Excitcdnlity. — The excitability of the fibres forming the internal capsule is
very high. (Franck, Gliky, Beevor and Horsley.)

(2.) Mode of Discharge. — Indistinguishable from that of the corona radiata, quod
vide.

(3.) Nature of Discharge. — Ditto.

(4.) Relation of the Discharge to the Parts of the Body. — {a.) Localisation of the
representation of the gross divisions of the body to limited fields of fibres. (Franck
and PiTRES, Gliky, Beevor and Horsley.)

(6.) Localisation of the representation of segments of the gross divisions of the
body to definite fields of fibres. (Beevor and Horsley.)

(c.) Localisation of the representation of the character of various movements of
segments to definite fields or bundles of fibres. (Beevor and Horsley.)

C. Cms Cerebri.

(l.) Excitahility. — The excitability of the fibres has been determined by many
observers. (Budge and others.)

2 N 2

27r. • MESSRS. F. GOTCH AND V. HORSLEY

(2.) Rdatlon of the Discharge to Parl^ of lite. Body. — Localisation of IIk! represen-
tation of the gross divisions of tlie liody to certain fields of fibres. (Bhissaud and
others.)

D. Spinal Coi'd.

It will be scarcely realised, except by those who have made a special study of the
literature of the physiology of the spinal cord, how little has been done whicli would
enable us to carry out, strictly speaking, the differentiation of the subject of conduc-
tion in that organ, by arranging the facts in the same way as we have just done for
the higlier parts of tlie nervous system. So much so is this the case that we do not
intend to do more at this stage than indicate what we believe to be a fair estimate of
the experimental work already accomplished. In introducing our own work later on,
which directly bears on the points at issue, we shall draw attention to many of
the more salient details which require attentive discussion. We will, therefore
now content ourselves with furnishing a general review of the function of con-
duction in the cord. We mav note in passing that almost all the literature is to be
found in the writings of von Bezold, Eckhakd, Grunhagen, Schiff, Immanuel
MuNK, and others. We would, in the first place, urge, what we have reason to do
with greater weight later, that not one of the foregoing methods, i.e., graphic degene-
ration, embryological, &c., is capable of answering the question of conduction or its
localisation in an absolute manner. The great difficulty in considering conduction by
the cord is the disseverance of the action of the bulbo-spiual centres from that of the
nerve fibres which happen to pass through or by them, and this difficulty does not
appear to us to have met with the amount of attention that it would naturally seem
to deserve. In fact, to our mind the galvanometric method is the only means within
our reach at present by which a solution can be arrived at, and even that method
requires to be considerably further elaborated.

Dividing conduction in the spinal cord into the two great classes of —

1. Conduction of impulses downwards ;

2. Conduction of impulses upwards ;

we are able to summarise the facts which appear to be thoroughly reliable as follows : —
(1.) Conduction of Impulses dowmvards. — In Man and the highest Apes direct
conduction downwards, i.e., from the cortical centres to the bulbo-spinal system,
appears to be provided for in the upper half of the spinal cord by both the anterior
column, close to the margin of the anterior fissure, and, speaking roughly, by a
triangular area in the posterior part of the lateral column, just external to the
posterior horn of grey matter.

(2.) Conduction of Impulses upwards. — There is known to be an entrance from each
posterior root into the postero-external column of fibres, which run directly from the
ganglion of the posterior root as high as the medulla oblongata. These direct fibres

ON THE MAMMALIAN NERVOUS SYSTEM. 277

are fiu'ther known to gradually tend towards the middle line of the cord as they are
displaced by others entering the cord in proceeding upwards. Further these direct
fibres are strictly unilateral, so far as is known at the present time.

There are, in addition, also other known systems of ascending channels, but these
are not directly continuous with the root, but start indirectly from central mechanisms
which, to judge from the degeneration method, although this is of course not absolute,
appear to intervene. These are (l) the direct cerebellar tract which runs up the
posterior and outer surface of the lateral column, and (2) the autero-lateral tract
which occupies a similarly lateral position further forwards on the margin of the coi'd.

In addition to the foregoing, there is some evidence to show that intemuncial fibres
run in the lateral column in about the inner third of its centre, or a little posteriorly
to this point. Of the existence of channels for transmission of impulses upwards in
the opposite posterior column to the side of the root by which they enter the cord
nothing is known for certain.

History of the Galvanometric Method of Determining the Action of the Nerve
Centres and the Course of Nerve Channels.

As we have before frequently acknowledged, the real basis of the galvanometric
method was the discovery by DU Bois-Reymond* of the negative variation produced by
excitation in the resting electrical difference of a nerve path. Since that discovery,
the idea has doubtless occurred to various physiologists that, by this means, we might
discover the mode of functional activity of nerve centres. CATONt was the first to our
knowledge who directly employed the galvanometric method of determining such
variations for the investigation of the activity of nerve centres, and the localisation
of the same. He connected points on the external surface of the cortex with the
galvanometer, and he found that the uninjured external surface of the brain was
usually positive to a section of the same. When any part of the cortex thus
investigated was thrown into activity, the resting diiference showed distinct negative
variations. Thus, in the Monkey, after the cortex had been prepared and the
electrodes applied to the centres (Ferrier) of rotation of the head and mastication,
the negative variation or action current showed itself when these movements were
performed. Fui'ther, in the Rabbit, when the area of the cortex which subserves the
movements of the eyelids was investigated, the negative variations showed themselves
when the opposite retina was illuminated. Set.schenow| was the first to our know-
ledge who connected the medulla oblongata with the galvanometer. He noticed
certain periodic variations in the resting electrical difference, which he attributed to
periodic changes of the functional activity in the bulbar centres.

* ' Untersuclmngen Uber Thierische Elektricitiit.'

t ' British Medical Journal,' 1875 ; also 'Transactions of the IX. International Medical Congress, 1887.'

J ' Ai-chiv fiir gesam. Physiologie,' Pflugeb, vol. 2.5, 1881, p. 281.

278 MKSSRS. V. (iOTCII AND V. IIOIiSI.KY

We cdinnienccd our reseai'ches in 1888, and ])ublished a preliminary account of the
same as stated in the introduction to the present paper. Moreover, we demonstrated
our method in 1889 before the Physiological Society, and further, at the Tnternational
Physiological Congress, held in Basle, in September, 1889, we made a general
communication on the method, and showed a complete experiment to tlie physiologists
there assembled. An abstract of our paper in the ' Proceedings of the Royal Society,'
was published in the ' Central blatt flir Physiologic,' 1889, and the account of the
demonstration at Basle was published in October, 1889, not only in the ' Centralblatt
fur Physiologic,' but in tlie ' Progres Medical,' 1889, and elsewhere. We have
published, on the occasion of a priority discussion presently to be alluded to, in the
' Centralblatt fiir Physiologic,' 1891, the various publications we have made of our
method a,rranged in chronological order. It was not until the close of 1890 that we
learned that the galvanometric method was being employed abroad. On the 8th
November, 1890, there appeared in the 'Centralblatt fiir Physiologic,' a paper by Dr.
A. Beck, of Cracow, who, ignorai;it apparently of our ])ublications and demonstrations,
described the galvanometric method, and pointed out the value of it in determining
the localisation of centripetal or afferent nerve function in the brain and spinal cord.
This paper, however, was really but an abstract of a full paper which was presented to
the Academy of Science in Cracow, in 1 890, a copy of which we owe to tlie kind courtesy
of Professor Cybulski, and in which is given a brief reference to the Basle demonstra-
tion, but no reference to our publications in 1888, or to Professor Biedermann's
abstract of the same in the ' Centralblatt flir Physiologic,' 1889. The appearance of this
paper produced a priority reclamation by Professor Ernst Fleischl von Marxow, of
Vienna, pubhshed in the 'Centralblatt fiir Physiologic,' on the 6th of December, 1890.
In this communication Fleischl showed that, as long ago as the 7th of November,
1883, he had deposited in the archives of the Imperial Academy of Science in Vienna
a sealed letter, in wliich he announced that he had employed the galvanometric
method for the same purpose, vide infra, as Beck and with the same details.
Fleischl, in this reclamation, does not mention any more than Beck, the prior
investigations of Caton, or our publications and demonstrations of the last three years.
Although Fleischl appears to have independently thought of employing the galvano-
meter as an index of functional activity in the nerve centre, when it is the seat
of centripetal or afferent disturbance, his method of recording his idea in a sealed note,
discounts the credit that otherwise might fall to him. All these authors have dealt
with the employment of the galvanometer as an index of the changes going on in the
nerve centre, i.e., nerve corpuscles, when that nerve centre is directly connected with
the instrument. We will allude directly to the results obtained by Beck and by
Fleischl, but we wish to point out that from our own observations made in the same
way, we are not at present satisfied that the basis of these researches is entirely
trustworthy (see Chapter IX.), and that they deal with but one point in this extensive
subject. So far as we are aware, we were the first to determine by use of the electrical

ON THE MAMMALIAN NERVOUS SYSTEM. 279

method the locaHsation and quantitative estimation of either centripetal or eflfei'ent
impulses issuing from nerve centimes in functional activity, and, in addition, we first
employed it for the determination of the localisation of paths or channels of nerve
function, particularly in the spinal cord. After this preliminary statement, w^hich has
been rendered necessary by the publications referred to, we will refer to the results
which were obtained by Fleischl and Beck respectively, although, as we said just
now, these results are, we consider, for the reasons we discuss on p. 29(5, not to be
immediately accepted. Fleischl {loc. cit.) connected, by means of non-polarisable
electrodes, two symmetrical points on the surfaces of the cerebral hemispheres with
the galvanometer, and found that in the resting condition there was little or no
electrical difference. If, however, a sensory end organ, whose corresponding area of
the cortex was thus connected, were excited, the effect produced was an electiical
difference as recorded by the deflection of the galvanometer. This effect he especially
obtained when the visual centre discovered by Munk was connected wiih the
galvanometer and the eye illuminated, whereas little or no difference followed ex-
citation of cutaneous nerve endings. Finally, he found that profound narcosis with
chloroform or ether abolished this effect, and that special precautions must be taken
against cooling of the preparations.

Beck {loc. cit.) removed the brain, spinal cord, and sciatic nerve of the frog en
bloc and placed the preparation on a glass plate ; he then applied non-polarisable
electrodes made of kaolin and 0'6 per cent, salt solution to the longitudinal surface of
the spinal cord and connected them with a Hermann's galvanometer. He found
that there was constantly an electrical difference of such a nature tliat the centripetal,
i.e., proximal parts of the nervous system were always electro-negative to centrifugal
or distal parts. If then the sciatic nerve were excited this primary difference was
increased, provided the galvanometer electrodes were placed above the lumbar en-
largement. If, on the other hand, the sciatic nerve were excited while one of the
non-polarisable or leading-off' electrodes were placed on the lumbar enlargement, there
was observed a negative variation of the primary difference. In another series of
experiments on Dogs and Rabbits he connected two points of the surface on one
hemisphere with the galvanometer, and found that there were more or less regular
swaying movements of the needle^ and which he regarded as "action currents."
Excitation of the retina caused the visual centre of Munk to become negative to the
rest of the hemisphere.

Danilewsky'^"' has quite recently published the results of five experiments wliich
he performed in 1876, and in which he found that when non-polarisable electrodes
were connected with the cerebral hemispheres (whether superficially or deeply) and
with a sensitive du Bois-Reymond galvanometer negative variations of the restino-
difference were observable as consequences of various modes of sensory stimulation.
The portion of the hemisphere from which these effeccs were obtained was the
* ' Centi-alblatt fiir Physiologie, ' Api-il, 1891, p. 1.

•280 MESSRS. F. COTCII AND V. HOHSI.KV

posteri(M- pai't hIcmic nnd in most cases opposite to the side to which the excitation
was applied.

CHAPTER III.— DETAILED DESCRIPTION OF THE METHODS EMPLOYED IN

THE PRESENT RESEARCH.

The description of the methods may be best effected by considering, in succession : —

1. Tlie method of ansesthesia employed.

2. The operative procedure.

3. The recording and exciting apparatus.

4. The general procedure and the precautions used.

Section I. — Method of An/esthesia.

The anajsthetic employed in these experiments was, in almost all cases, ether, the
exceptional cases being those iu which some bronchial or nasal catarrh rendered it
advisable to substitute chloroform, this latter giving a more even and steady narcosis,
though at greater risk to the animal.

The physiological action of ether has received considerable attention lately, and a
few points in connection with it are of sufficient importance as bearing upon our
experiments to warrant a more detailed notice.

Ill the first place, it has been shown (Hooper, Semon and Horsley, Bowdttch,
and others) that ether when present in the circulating blood through inhalation,
appears to have a distinctly differential effect upon the two kinds (red and white) of
muscle, or upon their innervation, i.e., centres, fibres, or nerve endings.

Further that the direct application of the liquid or vapour of ether to the trunk of a
nerve causes paralysis of its physiological conductivity.* (Hooper and Biedermann.)

The investigations of FlourensI in 1847 had elicited the fact that the inhalation
of ether produced physiological effects, in which the functional activity of the reflex
centres was abolished before that of the conducting nerve paths, and this has been
substantiated by the work of other investigators.

It is with special reference to the action of ether upon these two structures that
the present remarks are introduced, since, with the exception of the few experi-
ments in which we employed tlie graphic method, our work has been entirely confined
to the study of the changes in nerve centres and nerve fibres.

It was consequently of primary importance for us to know to what extent the
inhalation of ether has a differential action upon nerve centres as distinct from nerve
fibres.

In this connection experiments (Horsley and Spencer) have shown that the

* Biedermann, 'Wien, Akad. Sitzber.,' vol. 97, 3 Abth., 1888.
t Flouuens, ' Comptes Kendus,' vol. 24, 1847, p. 161.

ON TUK MAMMALIAN NERVOUS SYSTEM. 281

instant effect of ether inhalation is a remarkable fall in blood-pressure, such a fall
in blood-pressure as the early experiments of Vulpian showed would of itself, if
sufficiently pronounced, cause a diminution in the functional activity of all the
nerve elements, such diminution occurring first and most markedly in the centres,
and even a comparatively slight fall may be followed by a loss in the excitability of
these, although the nerve fibres might not be appreciably affected.

The effect of ether, therefore, in diminishing the excitability of the cortex and
lowering its functional activity may be in part due to changes in the circulation.
It is probable, however, that the ether in the blood exercises a direct toxic eifect
upon the nerve structures, of a similar kind to that which is brought about by the
du-ect action of the vapour already alluded to. We are not aware of any experiments
as to the direct toxic action of ether vapour upon nerve centres ; but it is extremely
probable that such action occiu's and that the centres should be affected by an
amount of ether in the circulating blood which is too small to affect in any sensible
degree the fibres.

Whatever the share which the two factors, blood-pressure changes and ether in the
blood, may respectively have in the production of the effect, the result is that
profound general narcosis serves to abolish the functional activity of nerve centres
before that of nerve fibres. We have therefore made use of etherisation, with due
caution, to assist us in analysing the compound excitatory effects observed when a
complex structural arrangement of centres and fibres has been stimulated, and we
have in all cases observed great care in noting as accurately as we could the degree,
whether profound or sHght, of narcosis at the time of each experimental observation.

There is one possible disadvantage in the use of large quantities of ether during a
considerable period of time which is not very obvious at first sight, and that is the
amount of vapour which is present in the air of the room. It did not occur to us that
this coyld act injuriously upon the preparations under investigation until we noticed
in three prolonged experiments in which, owing to the method of inhalation, a large
quantity of ether was used in a warm close room, that both the exposed cortex and
the sciatic nerve suddenly lost their excitability, which they did not regain. The
possibility of this being due to that injurious action of the ether vapour dissolving in
the liquids upon these structures, which was pointed out by Bcedermann, theji
occurred to us. The effect of the vapour in the room Was increased, perhaps, by its
dissolving in a warm bath containing 0'6 percent, saline, which mixed with the blood
of the animal was employed for irrigation, &c., of the nerve structures. Although
the failure might have been due to other causes (exposure, &c.), stiU. the fact of its
not occurring when special precautions were taken to avoid the excessive evaporation
of ether into the room and the contamination of the saline, leads us to conclude that
the cause mentioned is at any rate a depi-essing factor which ought, as far as possible,
to be excluded.

The use of chloral or morphia would undoubtedly entirely exclude any error of tliis

MDCCUXCI. — B. 2 O

282 MESSRS. F. (MVPCII AND V. IIOKShKY

kind, liiit it has the enormous disailvuntage that it is impossible to iiltcr tlu; degree of
narcosis, and, as will be seen, such alteration is an essential condition iu experimcntH
of the kind which we have undertaken.

As regards the practical aduiinistration of the anaesthetic, the best results were
obtained by pushing the etherisation at first to a profound degree, so as to abolish
entirely superficial reflexes, and always by causing profound narcosis before per-
forming any operation, such as division of the spinal cord, &c., which would, if the
narcosis were less deep, entail depressant effects upon the centres concerned with
systemic life through the damaging influence of shock. Subsequently when, the
operative procedure having terminated, the actual observations were being made, the
narcosis was rendered less and less intense as the slow collapse inseparable from the
experimentation asserted itsoll'. This slow collapse, since it involves the gradual
failure in excitability of the nerve structures and primarily of the nerve centres,
has the same effect as an anaesthetic in producing narcosis, which was preferably
intensified by the administration of ether in small repeated doses rather than in few
larffe ones.

o

Section II. — Operative Procedure.
1. Exposure of the Cortex and of the Corona Radiata.

{A.) Cortex. — The animal, having been deeply anaesthetised in the manner stated,
was further immobilised on a firm support, upon which a metal vessel was placed so
as to be underneath the thorax. This was kept filled with hot water and, combined
with suitable coverings for preventing loss by radiation, served to keep up the
temperature.

The dura mater was then exposed lege artis to the necessary extent and over
the required region, all haemorrhage from the bone being instantly arrested by the
use of soft modelling wax, &c. As soon as the dura mater was cleared the wound
was closed, and kept covered with sponges soaked in hot O'G per cent, saline solution,
whilst the further operations necessary for the experiment were undertaken.

The cortex was finally exposed by taking up the dura mater by means of iridectomy
forceps or fine curved needles, and dividing it, care being taken to keep it always
protected with hot saline sponges unless an actual experiment was in progress.

{B.) Corona Radiata. — The cortex having been thoroughly exposed, a sharp
scalpel was passed horizontally through the hemisphere in the plane of the centre of
the coronal gyrus, and the upper and anterior thirds raised as a lid of a box. Pieces
of amadou were then gently laid and pressed on the edges of the cut, thus arresting
the free bleeding from the vessels of the pia mater. Care was always taken to
determine precisely the topography of the section, and to accurately apply the
electrodes to the fibres coming from the requisite area of the cortex. The necessity

ON THE MAMMALIAN NERVOUS SYSTEM. 283

of this care is obvious to those who are familiar with the limitation of the fibres in
the corona radiata, but we think it well to mention the fact, as a malposition of 2 nun.
is sufficient, with the stimulus we employed, to prevent the production of the effect.

2. Exposure of Peripheral Nerves.

The great sciatic nerve was usually selected for this purpose, and was exposed in
the thigh for 6 centims. of its length. It was then divided near the knee, and its
central end ligatured, care being taken to prevent any pull upon the structure, or any
injury to the artcria comes nervi ischiadici which was always included in the ligature.
When not required for purposes of immediate experiment it was irrigated with 0"6 per
cent, saline solution, and covered over by the skin and muscular flaps. During use,
whether for purposes of excitation or in connection with the non-polai'isable electrodes,
it was always so fixed that no movement of the animal could pull upon the structure.

3. Preparation of the Spinal Cord for Observation.

The spinal cord afforded naturally more difficulty in its preparation, and this
requires therefore a more detailed description.

The objects in view were —

[a.) To expose and divide the cord.

(6.) To preserve as far as possible its circulation.

(c.) To fix the spinal column so that no accidental movements of the animal should
affect it.

{d.) To avoid the depression due to cooling, drying, &c.

(e.) To secure for the purposes of electrical investigation as complete an isolation as
possible of its different parts.

These requirements were met as follows : —

(cf.) Exposure and Division. — The muscles were rapidly exposed over the dorso-
lumbar region, and then partially extirpated and cleared from the vertebral laminae ;
haemorrhage was treated by frequent irrigation with 0'6 per cent, saline at a tempera-
ture of 50° C, and pressure of hot sponges. The exposed vertebral arches were then
carefully removed piecemeal for 6 to 8 centims., by the aid of powerful but fine-
pointed bone forceps.

In those cases in which the reflex affects of the lower fragment of cord were inves-
tigated, the cord was only exposed for a veiy short distance by the removal of one
lamina and then divided under profound anaesthesia. The small wound was then
closed and kept covered with hot sponges.

When a long portion of the cord was exposed the theca was divided in the middle
line with great care, the division being especially free at that portion of the exposed
area which was to remain in continuity with an unexposed region, so as to avoid

2 o 2

284 MESSRS. I'. COTCII AND V. IIOIISLEV

any danger of strangulation wlion the freed i)ortion of cord was subsequently raised.
On reflection of the dura, a ligature was cautiously passed around mm end of the
exposed cord, the end chosen varying in accordance as it was desired that the
observed portion of cord should be in conununication with the bi-aiu or with the
sciatic nerves. If the former, then it was knotted gently round the lower end so as
to securely close the vessels, and the cord divided immediately below the ligature.
The (central) end was now raised by the ligature, and the nerve roots exposed and
divided one by one imtil the whole portion of the exposed cord was freed from all
attachment except at the upper end, where it was continuous with the central
unexposed region. In the case of experiments upon the cord severed from the brain
but in connection with the sciatic nerves, the ligature having been applied round the
upper portion of the exposed tract, the cord was divided on the central side of the
ligature and the exposed portion freed downwards in a similar manner to that just
described.

(b.) Preservation of Circulation. — The circulation in the exposed portion was
maintained as far as possible by the ligature of its cut end, including the main vessels,
and by keeping the cord in connection with an undisturbed portion, and avoiding any
strangulation of that connection.

(c.) The Immobilisation of the Spinal Column. — The fundamental importance of
fixing the vertebrse necessitated the employment of a special clamp. This clamp was
applied so as to firmly grasp in its powerful vice-jaws the transverse processes of the
spinal column in the immediate neighbourhood of the exposed cord. The jaws were
fixed on a stem, and approximated by an ordinary double screw. In order to avoid the
extraneous electrical efiects which the presence of metal surfaces in contact with moist
cut tissues necessarily involves, the jaws were made of stout pieces of ivory. The
stem carrying them was fixed to a powerful upright attached to the experimental
table, and so arranged as to secure a fixation vertically above the preparation. (See
Plate 29.)

((7.) Cooling and Drying. — The exposed cord was kept bathed with the warm
saline solution until it was actually the subject of experiment, when, if it were raised
in air, care was taken to keep steaming sponges in its immediate neighbourhood.

(e.) Isolation. — The necessity of isolation for purposes of galvanometric observation
will be referred to later on. This Isolation was produced by raising the ligatured end
of the exposed portion of cord, so that the portion swung in air without pulling upon
its deep attachment. The electrodes, &c., were adjusted as described in the succeeding
sections.

4. Division of the Exposed Portion of Cord by a Longitudinal Incision.

For certain purposes it was desirable to observe the galvanometric effects in each
half of the cord independently. These were (1) the determination of the comparative

ON THE MAMMALIAN NERVOUS SYSTEM. 285

value of the cord excitatory effect direct and crossed when the cortex is excited,

(2) the determination of the bilateraUty of repi'esentation in one hemisphere, and

(3) that of the extent to which fibres cross from one side of the cord to the other.
We therefore devised the plan of dividing the exposed portion of cord l)y a longi-
tudinal incision, either in the lateral or antero-posterior direction.

The former was effected as follows : —

The cord, having been exposed, was severed, freed from its attachments, as previously
described, but not ligatured. It was then kept bathed with warm saline, and placed
upon a thin sheet of warmed cork with a shallow groove in it. The posterior fissure
could always be easily seen, and a longitudinal incision was then carried through it from
the jDosterior to the anterior fissure by means of a razor or sharp scalpel, the incision
commencing at the attached end of the cord. It is essential that in this operation
the instrument should neither pull nor press upon the cord unduly, but, since the
division involves both pressure and pull, the infliction of injury can only be miti-
gated by slowly making drawing incisions, first only through the posterior fissure,
and subsequently deeper and deeper until they have passed into the thickness
of the cord. The exposed portion is partly divided in this way down to its point of
transverse division for a distance of 2 centinis. or more. Irrigation with warm saline
being maintained, the edges of the wound are gently separated, when the central
canal of the cord will be seen in the bottom of the cut ; with this as a guide, it is easy
to continue by similar incisions the mesial division, luitil a complete separation of the
cord into two halves is effected. (See Plates 31 and 33.)

Each half thus prepared was now ligatured close to its cross section, and raised in
air for the necessary isolation when it was desired to subject them to electrical
observation. They were kept apart by the ligatures being carried to the respective
ends of a glass T-piece, or of two vulcanite rods arranged as a V.

Frequently the cord, when thus prepared, bled rather freely from the central end
of the cut.

The sunplest way to check this was found to be to press into the lips of the cut a
small fragment of soft dry amadou, and leave it in contact with the bleeding point.l

It might, perhaps, have been expected that such division would seriously impair
the physiological conduction of its fibres. It will be seen in the following pages that,
as far as the lateral tract is concerned, such impairment is not evidenced by our
observations, but that undoubtedly the posterior columns in the immediate neigh-
bourhood of the incision are injuriously affected by the process.

We have also divided the cord into a postei'ior and anterior half by gently raising-
it wlien freed and passing through it a thin-bladed knife from side to side opposite the
attachment of the ligamentum denticulatum, and carrying it forward to the free end
of the preparation.

2R(i MKRRRS, F. fiOTf'IT AND V. HORSLEY

5. Preparation of the SjiiuaJ Cord for Excitation.

Another mode of preparation remains to be referred to, that, namely, in which the
cord, when severed from its connection with the brain in order to investigate the
functions of the lower fragment, is to be excited.

It was necessary for these experiments to sever the cord as high as possible,
though, as a matter of fact, it was found advisable not to perform the operation at a
higher level than that of the 7th dorsal vertebra, otherwise, serious impairment
of respiratory and vaso-motor functions followed, and profound shock vitiated the
experimental results. When the upper end of the lower fragment of cord was to be
excited, this was furthei- exposed, and a piece 5 mm. in length excised on the peri-
])hL'ral side of the section, so that by looking Into the gap the structures on the cut
'surface of the cord could be seen and the excitation localised. The gap was made
absolutely dry with amadou and small pieces of sponge when the cord was to be
excited.

Finally, a series of experiments involved the division of the cord at two levels. A
portion was thus cut off, both from the brain and from the lumbar plexus, and prepared
so that one end could be connected with the non-polarisable electrodes and observed
whilst the other was excited. In all these cases the upper division was effected first,
so as to diminish the shock which the further operative procedure connected with the
lower division caused. The end with which it was desired to connect the electrodes
was then carefully exposed, ligatured and freed from its connections, as already
described under (3), whilst the other end was prepared for excitation.

6. Preparation of the Root-s of the Nerves.

The posterior and anterior roots were in some instances prepared for excitation, in
others for observation. In both cases the roots chosen were those formino- the cauda
equina, and they were exposed by a suitable opening of the spinal canal and theca in
the manner previously indicated. The roots were carefully sepai'ated and when
required each one was ligatured near its peripheral attachment and divided. Great
care is necessary to avoid pulling, drying, &c., in the case of the roots. (See
Plates 34 and 35.)

7. Section of Columns of the Cord.

In many experiments it was necessary to make a section of one or more of the
columns of the cord between the part excited and the part observed. The method
employed was to expose a portion of cord for the pui-pose and then to make the
section by means of Beer's cataract knife, completing it by means of a fine-cutting
needle or occasionally fine-pointed sharp scissors. In every case the limitation of the
section was ascertanied by post mortem examination. When a hemisection of the cord

ON THE MAMMALIAN NERVOUS SYSTEM. 287

was to be made, an additional precaution was adopted to ensure its accuracy, the
cord being fixed by a fine needle.

In some instances it was desirable that such intervening localised sections* should
be made some weeks previous to the actual experiments. This notably increased the
labour of the research, but is an essential control to such an experimental enquiry as
the present one. The animal was in all instances etherised and the seat of operation
having been carefully shaved, disinfected, &c., the necessary exposui'e and section were
then made under all aseptic precautions ; the wound uniting by first intention when
treated in the manner described by one of us in previous experiments of this kind.

In conclusion, it may be pointed out that since almost each experiment involved
exposure, &c., of several difl:erent portions of the centi'al nervous system, we adopted
the plan of performing these slowly, taking at least an hour or more in the preliminary
operative part of the experiment. We found that in this way less general shock
occurred than \\hen the various parts were rapidly prepared one after the other.

Section 3.— Kecording and Exciting Apparatus.

The recording and exciting apparatus used in the present research was chiefly that
adapted for determining the comparative amounts of the electrical changes in the
spinal cord and nerves at rest and when subjected to an excitation of definite intensity
and duration.

The apparatus and its arrangement will be best indicated l)y separating it into the
following groups : —

(1.) The apparatus used in connection with the observation of the electrical changes.

(2.) The apparatus used in connection with the excitation.

(3.) Extra apparatus used for determining the characters of the muscular contrac-
tions.

1. Tlie Electrical Apparatus for Observation.

Both tliese instruments were used to indicate the electrical changes evoked in the
spinal cord and nerves.

(a.) The capillary electrometer was made by Mr. G. J. BuRCH.t The mercurial
column was magnified about 300 times by the special microscopic arrangement
employed, and was sufiiciently sensitive to show perceptible movements when con-
nected with a difference of potential of j^q^qq Daniell ; its reaction was quick enough
to enable it to respond to a difference of x^o Daniell when connected with that

* The localisation of all points or regions of the cord exposed is expi-essed in terms of the body of the
vertebra, or intervertebral disc, opposite to which the section was made. We fui-nish in Appendix A the
topographical relations which the spinal segments bear to the superficial origins of the several nerves
from the spinal cord.

t Proceedings of the Royal Society.

288 MESSRS. 1-'. (.ajTCH AND V. IIOKSLKY

difference by means of a rheotorae for only xo"b"o second. Tlie ocular of the microscope
was fitted with one of Zeiss's micrometers, the scale being worked across the field
by a screw mechanism. Tn the particular arrangement employed, the scale bore such
a relation to the actual dimensions of the capillary that one division corresponded
to YijVff millimeti'e of the object. A difference of potential of yooo Daniell when
connected with the poles of the electrometer produced a movement of the mercury
amounting to r^y n^ilhmetre, and this, when viewed l^y means of tlie eyepiece,
amounted to five divisions of the scale.

The element of uncertainty in connection with the movement of the electrometer
under different conditions and the difficulty of obtaining reliable I'ecords of changes
differing from one another by constant but small amounts, induced us to abandon the
instrument for the quantitative observations which make up the bulk of the present
research. It was, however, often used as a means of ascertaining the existence and
character of electrical changes during the initial stages of a new series of experiments,
and proved in this respect a useful guide.

The movements of the meniscus were in most cases observed by the eye, but in some
cases they were photographed upon a travelling sensitive plate, as in the experiments
described in our previous publication.

(h.) The galvanometer was the instrument upon which we relied for results suscep-
tible of quantitative comparison. It was made by Messrs. Elliott upon the lines of
Thomson's reflecting instrument, and had a resistance at 16° C. of 20,364 ohms. The
light magnetic system was effectually damped by being enclosed between two plates of
glass 2 mm. apart, the aluminium vane of the ordinary Thomson being dispensed with
to secure a decrease in the inertia .of the system.

The degree of sensibility employed was such that the needle and mirror gave a
deflection amounting by the reflecting method of observation used to 650 scale, when
the instrument was connected through a resistance of 10,000 ohms (external to its
own resistance) with a difference of potential amounting to "01 Daniell, and when
only Yoo of ^^'^'^ current in the circuit was allowed by the shunt to traverse the instru-
ment. The small mass of the system enabled it to respond to currents of very short
dui-ation, thus with an additional resistance of 10,000 ohms in the circuit, a difference
of '01 DanieU produced a deflection of 5 scale when connected by means of a rheotome
with the galvanometer for a period of only -nfoo second.

The deflections of the magnetic system were measured by the usual reflecting
method ; a biconvex lens of 4 inches focal length was, however, introduced ahnost
3 centims. in front of the source of light between it and the concave mirror (of
40 inches radius) of the system. The screen behind which the light was placed was
pierced by a circular aperture bisected by a vertical cross wu'e, and the reflected
image on the scale was thus a large well-iUuminated disc with a sharply-defined
vertical shadow dividing it. The position of the edge of the vertical shadow on the
scale could be thus observed with great accuracy, since the moving illuminated dissc

ON THE MAMMALIAN NERVOUS SYSTEM.

289

was of sufficient size to embrace in its field a considerable range of the figures of
the darkened scale on each side of the shadow of the upright.

Fig. 1.
/ fiECOHDING emeu IT.

CalUxtcci I, CeM.

Tr»jiTrr>TKTnrBOfr>r 0 "^rs na B"o-«"rrwTm"aTnf

G coCvcvn<ym

^ I /tJOxciUriq

x.JLiccbrcbCe^ .

/i c ix)i> uf.n (J Kc i/.

» >-8r^9 (TO 0-,TlT)-a to p .' ^ jiJ./>P"3 0

JI EXCITING CIRCUIT.

2 JJdnicLl CtLLSk

The arrangement by which either of the above-described instruments could be
connected with the experimental circuit, and thus with the particular structure,
whether spinal cord or nerve, under investigation, is shown in fig. 1.

In the above arrangement the method of compensation and the method of
connection with the tissues call for a few descriptive details.

(c.) The compensator was of the pattern used in the Physiological Laboratory,
Oxford, and was similar in its general plan to that described by Burdon Sanderson,
as used* in his work on the electrical properties of DioncBa*

The total amount of wire in the instrument offered a resistance of 10 ohms, of this
a portion at one end, 125 centims. in length and of 1 ohm resistance, was exposed and
lay tightly stretched upon a boxwood scale. Upon this boxwood bed a heavy block
carrying a wire was allowed to slide. The block was furnished with a pointed index,
the under surface of which, covered with platinum, formed the sole contact with the
wire ; its position was easily read upon the scale beneath it.

The battery used in the primary circuit of the compensator was the Callaud
pattern, which is a Kaoult battery witliout any porous cell. This had been found by
one of us to be the simplest and most satisfactory battery for the purpose. As used
in the present research, it consisted of a glass jar containing about a litre, into this

* ' Phil. Trans.,' 1882.

MDCCCXCl. — B. 2 P

290 MESSRS. F. OOTCIT AND V. TIORSLEY

about 300 griiif?. of sulpluite of copper crystals were j)]aced, and tlie wliolo filled up
with distilled water; a coil of sheet copper, with insulated connections, is innuer.sed
in the crystals, and a ring of zinc in the superficial liquid.

The cell was always prepared forty-eight hours before use, and since the subsequent
alterations take place with great regularity and are slowly produced, the cell,
when evaporation is prevented, is very steady, and is thus particularly suitable for
the purpose indicated.

Tlie value of a given interval between different points of the compensator wire
when the Callaud cell was coupled up with it was always estimated before each
experiment and contrasted with that produced by a known fraction of a carefully
prepared Daniel! cell. This latter was determined by the balancing method employed
by DU Bois-Rfa'MOND and described by him in connection with his use of the " Rund
Compensator " in his standard work on the technique of physiological experiments
involving observations of electromotive changes. The measurements by the com-
pensator, although made with the Callaud, were thus all translated into terms of
a Daniell cell.

{d.) The non-polarisahle electrodes which formed the connection with the tissues
consisted in some instances of two pairs. Each pair was then joined to the two
screws on one end of a Pohl's reverser (without cross -wires) to the middle screws
of which the wires forming the main circuit were connected. Either pair of
electrodes could thus be rapidly connected with the recording instrument by
turning over the switch. In this way the changes in each of the two prepared
sciatic nerves, or in each separated half of the isolated and longitudinally divided
cord, could be ascertained one after another in rapid succession.

In the majority of experiments only one structure was examined, and one pair of
electrodes was therefore used and introduced into the main circuit.

Each electrode consisted of a U -shaped glass tube containing a saturated solution of
zinc sulphate with a well amalgamated and annealed zinc dipping into one limb.
This form was preferred in consideration of the experimental necessity of keeping the
electrodes exposed for an hour or more, in order to minimize the alterations due to
evaporation.

It was of the first importance that the connection of the electrode with the
investigated structure should be so firmly attached as to suffer no displacement from
any chance movement of the animal, and that they should be of such a nature as to
readily permit an adequate isolation of the structure from all its neighbouring tissues.
This was effected in the case of each electrode by means of thread cables made out of
the soft strands of such wicks as are used in ordinary paraffin lamps, since these
I'eadilj imbibe and become soaked with moisture. Two strands 10 centims. long, were
usually taken and soaked in 0"6 per cent, solution of NaCl ; they were then plastered
with powdered kaolin, made into a paste in similar liquid, and thus united into a
double cable. This cable was passed under the structure investigated and tied gently

ON THl'] MAMMALIAN NERVOUS SYSTEM. 91

round it ; the two ends of tlie tie were then united together, thus forming a cable of
four strands. The end of this cable was fixed into a plug of kaolin moistened with
similar sodium chloride solution, which was vniited in the usual way with sulphate of
zinc kaolin paste, and this with the liquid in the one end of the U-tube. Since from
4 to 5 centims. of this yielding cable thus hung between the rigid part of the electrode
and the attached structure, this latter could be moved so as to isolate it from the
neighbouring tissues without interfering with its electrode connection, whilst, owing
to the soft cable being tied around the structure no displacement of the point of
contact was possible in consequence of such movement.

2. The Exciting Arrangements.

The method of excitation involved the use of {a) electrical, (b) mechanical, (c)
chemical stimuli. Of these the electrical, as alone admitting of accurate graduation,
is the most important.

(a.) Electrical Excitation. — The exciting arrangement involved in almost all cases the
production of a rapid succession of induction shocks. These were obtained from the
secondary coil of an induction apparatus of the general pattern of that devised
by DU Bois-Reymond, and usually employed in physiological research. The par-
ticular coil used was made and standardised in Berne, under Professor Kronecker's
direction, and differed from those ordinarily used in having a scale which denoted for
every position of the secondary coil, with reference to the primary, the relative
intensity of the current induction effect. (Bowdituh.) Since the ordinary scale,
that of distance in centims. is most generally known to physiological workers, we
append the comparison of the numbers of the two systems showing their relative
intensity.

Dq Bois-Reymond inductorium scale.

cm. 0. 1. 2. 3. 4. 5. G. 7. 8. 9. 10.11.12.13.14 1.5.16.17.18.19.20.^1.22.23.24.25

Divisions of222'5PC5222'2'^'='C>'^2"=><3<='<='Ooi>ooooo.-i
Kroneeker in- ®''50(M(M(M»ocococi-.'COco^i^oeO'^<W(Hi— i,— i
ductorium 2 S" S d" S °^ =°" '^" ^' "^^ ^" ^" ^" ^' ^ S

scale.

.2 u

O !>
g O

Two Daniell cells coupled for intensity were used in the circuit of the primary
induction coil ; this contained an automatic electro-magnetic vibrator, wliich closed
and opened the circuit 50 times per second.

The interrupted circuit was a derivation bridging the primary coil (Hehnholtz side-
wire), so that each interruption should induce make and break currents of approxi-
mately equal intensity, thus avoiding the accumulation of polarisation after-effects in
the excited tissue. The electrodes were well insulated platinum wires with points
1 millim. apart. When placed on the tip of the tongue, a slight acidity only was

2 p 2

5!92 MESSRS. F. GOTCH AND V. HORSLEY

perceptible with the secondary coil at 2000, the tingling effect becoming first dis-
tinctly felt at 4000, and painful at 12,000.

The arrangement of the exciting circuit is given in the preceding figure 1, p. 289,
the only feature calling for remark l)eing the use of a revolving mei'cin-iul key in the
secondary circuit.

This key consisted of a hard paralliii bed, supporting two semi-circular narrow
vulcanite troughs containmg mercury. Into these dipped the two ends of a loop of
stout platinum wire, which was carried by a horizontal revolving vulcanite arm.
Each mercurial pool was connected with one terminal of the secondaiy induction coil,
so that when the platinum loop connected the two pools a short circuit was made for
the secondary coil. The arm was driven by a clock at such a rate that a whole
revolution occupied a period of 10 seconds.

By means of the key (D.B.) shown in the plan, the secondary coil was additionally
short-circuited, and the revolution of the arm thus made ineffectual ; when, however,
it was desired to excite the preparation, the key was opened for one revolution only,
and the secondary coil was thus disconnected from any short circuitous arrangement
durim>- 5 seconds. The interrupted induction currents were thus allowed to traverse
the tissue between the platinum points of the exciting electrodes for this period only.

(b.) Mechanical stimulation. — The importance of obtaining evidence of physiological
effects by different methods, as a control to the definite results obtained with
eleotrical excitation, led us to employ mechanical stimulation wherever practi-
cable. The simplest form of mechanical stimulus, and the most effectual, is un-
doubtedly that obtained by the complete and sudden severance of a tract of fibres.
This we accomplished, when the position of the parts allowed of it, by the use of sharp
scissors, thus severing, for instance, the sciatic nerve, and observing the electrical
chancre which the division evoked in the spinal cord, and vice versd. As, however, in
some cases the metal blades of the instrument coming into contact with the moist
tissue caused localised electrical changes, derivations from which affected the
galvanometer and electrometer, we substituted the stimulus occasioned by the sudden
tightening of a ligature, and also that caused by crimping the tissue, nerve, or cord,
by squeezing it suddenly (sometimes so as to divide it) between the jaws of a pair of
specially constructed ivory scissors.

The method of stimulus being used only as a control, the disadvantage of few
repetitions being possible was not a serious one.

(c.) Chemical Excitation.— In addition to the previous methods of excitation, we
have made use of strychnia, in intraperitoneal injections, in order to obtain pronounced
reflex discharges from the spinal cord.

Further, we have frequently employed, with notable advantage, the method of
exciting the brain first investigated by Magnan, viz., the production of nerve
impulses from central apparatuses by the toxic influence of certain substances,
especially the essence of absinthe.

ON THE MAMMALIAN NERVOUS SYSTEM. 293

Just as the mechanical stimulus affords, perhaps, the l^est control of the electrical
stimulation of fibres, so the chemical, since it excites centres powerfully, affords a
means of producing electrical changes in the efferent fibres from such centres which
cannot possibly be due to actual escape of the exciting current.

For the performance of this method we have prepared the cord or nerve for
connection with the non-polarisable electrodes as described previously, and then
exposed for a short distance the external jugular vein, applying a small clip to the
latter.

Two minims of the essence of absinthe were then injected with a hypodermic
syringe inserted through the vein wall into the freely flowing blood stream.

The animal was then carefully observed and the first twitching indicating the
commencement of a cortical discharge noted. It coincided in time with the beginning
of large electrical effects as evidenced in the galvanometer. For further facts relating
to this method we would direct attention to the Chapter dealing specially with the
I'esults thus obtained, vide p. 511.

3. Extra Apparatus.

The appai'atus used for investigating the muscular changes evoked by excitation of
the central nervous system was either adapted for a direct record of the character of
the muscular contraction or for ascertaining the moment of its commencement.

When necessaiy the muscular contractions, whether evoked by coi'tical or spinal
excitation, were recorded by means of a spring lever, made on the pattern of that
employed by FiCK for obtaining isometric muscular effects. The muscle was attached
so that any changes in its tension were communicated to the spring, the small move-
ments of which were magnified fifty times by a lever attached to its end. The lever
recorded its movements on the blackened paper of an ordinary drum driven regularly
by a clock.

When it was desired to ascertain the moment of the commencement of the
muscular contraction, however evoked, the method employed by Tigerstedt was used.
The Mammalian muscle was attached to the light " unterhrecher" which carried a
weight of 10 grm. placed upon its axle, to ensure a steady pull upon the large muscle
and a proper tension of its attachment. The unterhrecher, as in Tigerstedt's method ,
formed a key in a separate circuit, which included a battery of three Grove's cells and
one of Smith's* new electromagnetic signals. The break of the key and the consequent
movement of the signal was recorded upon the glass jjlate of a spring myograph
(du Bois-Reymond's Federmyoyraphion) which travelled at a rate of 2 centims. in

i~oo second.

* 'Philosophical Majjazine,' 1889.

294 MESSRS. V. GOTCH AND V. HOUSLKY

Section 4. — The General Procedure and the Precautions used by the

Authors.

A . The Method of Experiment.

In any individual experiment in which excitatory electrical changes were observed
the procedure we adopted was as follows : —

The animal having been anajsthetised and innnobilised, the structure to be
investigated was exposed and prepared in the manner indicated. It was always,
whether spinal cord or nerve, divided, and the soft cables of the no n-polari sable
electrodes were then tied, one round the structure close to the point of division, the
other about 1 centim. away round the surface. The ligature upon tlie cut end
enabled this portion of the tissue to be suspended with its cables in air, without
dragging on the remaining portion of the tissue which rerxiained in situ. The cables
were then connected witli the non-polarisable electrodes. (See Plate 29.)

The electrical difference between the two contacts was first observed and balanced
by the introduction by means of the compensator of a suitable difference opposite in
sign to the tissue difference. The general characters of the latter and the amount of
the balancing difference were then noted, both the galvanometer and electrometer
being used for this purpose.

One of us devoted himself to the observation and record of the electrical changes ;
the other to the maintenance of the animal in a uniform condition of anaesthesia, to
the prevention of the drying, &c., of the mvestigated tissues, to the observation of
the muscular contractions, and to the excitation.

The tissue to be excited, wliich had been exposed with that under investigation,
was then finally prepared, and the points of the electrodes were brought into contact
with it under all the precautions to be immediately described.

The electrodes were kept short circuited as already stated by means of a key under
the control of the observer at the galvanometer, &c. When everything appeared
favourable and the galvanometer needle had been brought to its zero position by
suitable compensation the revolving exciting key was set going and the control key
opened for one revolution of the revolver ; an excitation of definite duration and
intensity was thus applied to the excited tract, whether cortex, corona radiata, spinal
cord, or nerve. The extent of any deflection of the galvanometer needle or the move-
ment of the meniscus of the mercurial column was noted, and special attention paid
to its character. The extent recorded was always that between the jirevious resting
position and the point where the moving recorder, whether spot of light or image of
meniscus, stopped and commenced its return. If no return occurred liut only a
deflection produced which continued to creep on then the observation was disregarded
as worthless. Such an effect was found to be generally associated vv'ith some move-
ment of the animal which caused a slight displacement of the electrode contact.

ON THE MAMMALIA.ISr NERVOUS SYSTEM. 295

Whilst one observer noted the electrical eft'ect, the other noted the extent and
character of any muscular movements by which the animal responded to the stimulus.

Tt need scarcely be said that the whole method involves the use of special pre-
cautions to avoid the introduction of fallacious and misleading effects, especially
when it is desired to obtain a succession of results which ^vill admit of quantitative
comparison.

These precautions are, in the opinion of the authors, of the utmost importance and
must be considered in some detail.

B. Precautioiu in Connection with the Method.

The special precautions may be grouped as follows : —

(1.) Those connected with the isolation of the particular region under observation.

(2.) Those connected with the condition of the non-polarisable electrodes.

(3.) Those connected with the condition of the animal.

(4.) Those connected with the localisation, &c., of the excitation.

( I . ) The Isolation of the Particular Region Observed.

It is, in our opinion, an essential condition for the accurate employment of any
method of localisation which relies upon the evidence of electrical changes in a given
region, that the region in question should be as far as possible isolated. In the
galvanometric experiments alluded to In the History (p. 279), as carried out upon the
cerebral hemispheres, such isolation was not affected ; a door was thus left open for
introduction of errors which it is not easy to control.

We have repeatedly had occasion to observe that when a pair of non-polarisable
electi'odes is placed upon the cord lying in situ, or upon the surface of the exjDosed
brain (see fig. 26, Chapter XL) electrical diffierences present themselves and influence
the galvanometer, this being evidently due to the fact that the parts with which the
contacts are made, since they form one directly continuous mass with the structures
around them, lie in the path of the derivations of currents, whose primary electro-
motive source is far I'emoved from the electrodes.

In this connection we may refer to the derivations of the electrical difference
between the different regions of the beating* heart which are present in the body.

On connecting two points on the surface of the exposed brain by means of non-
polarisable electrodes with the galvanometer, any reflex movement of the scalp muscles
lying outside the exposed region was found to evoke electrical changes in the points
of contact, and if structures so far removed from the seat of observation can affect the
contacts, how much more easily will these be affected by sources of electromotive differ-
ences situated in the deeper fijbres, &c., of the tvain. The fact that eleqtrical

* Waller, ' Phil. Trans.'

-!"! MKRSRS. F. GOTCll AM) \'. HO|{S[iEY

changes inauifrst tli(-iii.scl\ (^s hctwccn two siiriace contacts is tlms by itself no proof
lliat tliese elcflrical changes have tlieir source in ilic tissue in tlu^ immediate neigh-
liouiliuod of the contacts. It is, however, easy to ascertain whether the opposite con-
chision is or is not true, by sUcing off the two portions of the surface to which the
electrodes have been attaclied, destroying their vitality, and replacing them in contact
with the subjacent tissues, so as to act as mere moist conductors. If under these cir-
cumstances electrical changes are observed in the electrode areas, then it is clear that
since all structnml and physiological continuity is destroyed, these changes must be
due either to (1) physical properties of the observed region, which are in no way asso-
ciated with physiological vital changes ; or (2) to the physical spread of derived
currents fi-om electrical changes of physiological origin in the dee])er uninjured tissues.

The further distinction between these two alternative causes is effected by observing
to what extent the electrical changes disappear in consequence of systemic death,
since this obviously will affect the second class but not the first.

It is, in oiu- opinion, essential in all exact experimental investigations carried out
upon a particular region of such a mass of conducting material, as the body of an
animal constitutes, to use a method which can by strict investigation carried out along
the above lines be shown to exclude such discrepancies.

It may be pointed out that whilst we have selected the brain mass as a typical
instance, the same objections apply with equal or greater force to the investigations of
the spinal cord in situ. When it is remembered that the exposed cord lying in its
cavity is brought into immediate connection with a large mass of muscles, some unin-
jured, some cut for operative purposes, and with the whole contents of the abdominal
and thoracic cavities, it is not to be wondered at that the slightest movement of the
animal should cause very large electrical changes between any two portions of the
surface of the exposed cord, these being simply due to an alteration in the position of
muscles or other structures, all of which are the seat of electromotive change. Such
alterations must seriously influence the particular effect which may manifest itself
when any two portions of the exposed cord are connected with the galvanometer.
The force of these considerations is strengthened when it is borne in mind that
the particular effects which the electrical method is to gauge, are excitatory in
character. Of what value would any excitatory effect be, if, when it is evoked,
there are, as is almost always the case, not only marked excitatory electrical changes
in other organs, muscles, &c., but genei'al movements and displacement of subjacent
parts as the result of the excitation ?

Enough has now been said to point out the necessity for the utmost possible isola-
tion of the observed region, if it is desired to ascertain by the electrical method the
presence of electrical changes due to electromotive differences occurring in that region
only, and which may be therefore interpreted as indicating the presence therein of
excitatory conditions.

The most complete method of isolation is obviouslv the removal of the structure

ON THE MAMMALIAN NERVOUS SYSTEM. 297

under observation from all contact with other tissues. This is not possible in the
case of the Mammalian central nervous system ; it is, however, quite feasible in the
case of the spinal cord to ensure an isolation which we find to be sufficient. This, as
has been detailed in the section dealing with the operative procedure, as i-egards its
essential feature consisted in always dividing the cord, freeing it for several ceutims.
from all its attachments, ligaturing the divided end, and suspending this portion in
air by means of the ligature (see Plate 29). The cord is thus only in connection with
the structures by its deep end, and any spread of electrical currents, &c., which have
their source in extraneous regions, can only occur in this portion of cord as " extrapolar
effects " ; further, the electrodes being placed one upon the extreme end (cross section),
the other only half to one centim. on the proximal side of this, such extrapolar deri-
vation effects, if present, must be still more duninished by the distance, 3 centims.,
between the electrode contacts and the deep connections of the coi'd.

The necessity of the above careful isolation has been impressed upon us by the ease
with which it is possible to introduce errors even where this mode of connection is
carried out. If, for instance, either the proximal contact be allowed to slip down and
to approach too closely (within 1 centim. or less) to the deep structure, or if the
latter, by the presence of large quantities of liquid, &c., be brought into connection
with a portion of the isolated tract close to the proximal electrode, then, even with the
remaining electrode contact well isolated, electrical effects manifest themselves between
the contacts, which we have no doubt ai'e really due to extrapolar spread from
changes in the deeper tissues. Thus, for instance, the anterior roots, even when
divided and one end isolated, present great difficulties, owing to the shortness of the
tract intervening between the contacts upon them and the cord from which they spring.
Electrical excitatory effects can be obtained on exciting the brain, which are apparently
situated in the anterior roots, but Avhich are in this arrangement largely due to
electrical changes, situated in the cord, and occiu- as extrapolar derivations in the
anterior root. This is readily proved by using the method of control previously
referred to, viz., cutting off the attachment of the root to the cord and then replacing
it in its old position, when, in spite of the want of structural and physiological con-
tinuity, the changes are still found to occur.

We cannot lay too much stress on the necessity of careful isolation as far as prac-
ticable, and of careful investigation by both the methods indicated, and other similar
ones, of the extent to which observed electrical effects are due to mere physical
spread through the moist tissues of changes in other regions than the observed ones.

It will be understood, therefore, that in the following experiments, when the cord
or sciatic nerve was observed, the length of the tissue afforded every facility for
placing the electrode connections at a safe distance from the point where each of these
structures came into relation with the general mass of the body, and in consequence
a position could be chosen in which all danger of extrapolar spread might with
ordinary care be guarded against. When, however, the roots were observed, since

MDCCCXCI. — B. 2 Q

298 MESSRS. V. OOTCll AM) V. lloUSLKY

the distaiict' In ([iicstion was necessarily slnut, tlic diinj^crs of extrapolar spread were
increased, and the experiuKMital results had to be controlled with great cai'e ; this
danger was most formidable in the case of the anterior root, as owing to its anatomical
relations it could not be so conveniently separated from its contiguity with the mass
of the cord, and the errors thus introduced have been at present sufficiently serious
for us to abandon these roots for direct connection with the galvanometer, and rely on
those expcriuuMits pcifbrnieil iijxm (he sciatic nerve witli nil the posterior roots cut.

(2.) The Condition of the Non-poJarisable Electrodes.

It is scarcely necessary to observe that one obvious precaution in connection with
the electrodes is to ascertain, before commencing the experiment, that between the
electrodes themselves there is only an extremely slight and constant electrical differ-
ence, and noting its amount in terms of the compensator.

There is one difhculty, however, which appears to be inseparable from a method
which involves a long bridge of kaolin moistened with 0"G per cent, of saline, exposed
to the air of the room, and that is the alteration of resistance due to slow evaporation.
This alteration does not affect the results wljen observed by the electrometer, owing to
the very high resistance offered by the latter. It may, however, seriously affect the
galvanometer results. It will be obviously mitigated by employing a circuit of very
high resistance, and this we always took care to do, the full resistance of galvanometer
and electrodes amounting to little short of 50,000 ohms.

It is, however, necessary to keep in mind the presence of such alterations, and
to ascertain by repeated measurement of the galvanometric effect produced by a given
difference in the compensation circuit, whether the resistance has so altered as to
vitiate the value of the results. Where experiments follow one another closely, as
was the case in all the instances in which the different columns of the spinal cord
were excited, any alteration would be obviously too slight to exercise an appreciable
influence upon the results, which admit, therefore, as far as this feature is concerned, of
strict quantitative comparison.

It is obvious that remoistening of the electrodes must only be effected at the
commencement of a fresh series of observations, and must on no account take place
during any given series.

(3.) The Condition of the Animal.

{a.) General Preliniinary Condition. — In all the experiments this was a very
important factor, which we early recognised as determining the success or failure
of the particular observations. If the animal were ill, or had been much exhausted
by the preliminary anajsthetisation, &c., it was noted that the electrical changes,
whether due to currents of rest or of action, in both nerve and spinal cord were
comparatively small in amount. This was especially the case in the Monkey.

(/;.) AncButhesia. — The degree of narcosis is an extremely important factor. It will

ON THE MAMMALIAN NERVOUS SYSTEM. 290

be seen in the chapters which deal with the different gi'oups of results, how the effects
obtained in slight anaesthesia difler from those obtained with similar stimulation in
profound narcosis. It was, therefore, extremely important to keep the condition
of anpesthesia as far as possible even, and to note as a guide to that condition
the presence or absence of muscular movements, when any part of the central nervous
system was stimulated.

In one respect, the degree of etherisation was most useful, since it served as
a valuable control against error due to simple electrical escape from the exciting
into the galvanometric circuit. It is evident, if on pushing the degree of narcosis
the electrical changes in the observed tissue are very much diminished or even
abolished, whilst the excitation by electrical current remains unaltered, that no
physical conduction of derivation currents from the exciting into the galvanometric
circuit can be present.

(c.) Collapse. — Since the experiments are often necessarily spread over a consider-
able period of time, the gradual collapse of the animal is not infrequent. Its onset is
readily distinguished, and can be allowed for by diminishing the amount of etherisation
(see p. 282), and by increasing the intensity of the stimulus used.

(c/.) HyperexcitahiUty. — In connection with this part of the subject must be noted
the variations in excitability of parts due to section.

As regards the cortex it has occasionally happened that it has been slightly punctured
by an electrode point, &c. Whenever this happened the effect was to heighten the
excitability of the immediate neighbourhood, and to produce thereby a very marked
epileptogenous change. A more extensive injury has the further effect of abrogating
not only this increase of excitability, but the normal degree as well.

As is well known, the excitability of nerve fibres is raised by their section, and
from a rapidly attained acme the excitability steadily diminishes, passing the normal
point and becoming subnormal. FRANgois Franck has shown the same to be true
for the reflex centres in the spinal cord, when division of the latter has been carefully
made high up and without notable haemorrhage. We have always borne this in
mind, and have discounted the rise and fall of excitability due to the section in each
observation.

(4.) Localisation of the Excitation, &c.

Many precautions had to be observed in carrying out the stimulation of the cortex,
spinal cord, and nerves respectively, especially when the excitation of these structures
was electrical.

A. Cortex. — As previously stated, it was, of course, found imperatively necessary
to keep the surface of the cortex both warm and moist, this being especially achieved
by replacing the skin and by irrigation with warm saline solution. To avoid short
circuiting and consequent irregular intensity of the exciting stimulus, the surface of
the brain was always carefully dried with soft sponge or amadou immediately before

2 c,, 2

300 MESSRS. K. COTCH AND V. IIOKSLEY

the application of the electrodes. Fatigue ^A' tlio cortex was avoided l)y suitable
intervals of rest.

The intensity and character of the stimulus is referred to under a])paratus, p. 291,
but it may ])e here stated that the electrodes were fine platinum points 1 to 2 nun,
apart.

B. Corona Radiata. — To expose the corona radiata it was necessary to raise the
cortex by a horizontal incision, and i\\('\\ to check the bleeding from the pial vessels
with pieces of amadou around the cut border of the area exposed. If after several
excitations the exposure had lowered the excitability of the fibres, the electrode points
were sometimes inserted 1 mm. deep into the substance of the corona, instead of, as
usual, resting gently on the surface. The electrodes employed were the same as those
for the cortex.

C. Spinal Cord. — The excitation of the spinal cord, especially for purposes of
differentiation of the columns naturally demanded special attention. We obtained
our initial generalisations by means of exciting needles fixed on either side of the
cord or pressed against the longitudinal columns, vide pp. 3G9, &c.

This method we soon discarded (Section G, Part V.,) for that now to be detailed,

viz., the excitation of the cut sections of the columns of the cord, inasmuch as the

differentiation of the bundles of fibres could thus be accomplished at the same time as

other objects in view. The cord having been previously divided In the manner stated

above, and the surface and surrounding tissues being carefully dried, the fine platinum

electrodes, 1 mm. apart were used. It was our aim to provide for the excitation of

as many fibres in any column as possible in order that the largest galvanometric

effect might l)e evoked, and at the same time to avoid extrapolar excitation of

neighbouring columns. These particulars were fulfilled by applying the two electrode

points, as indicated on the accompanying fig. 2, i.e., vertically on the anterior and

posterior columns, and horizontally on the lateral column in the region of the crossed

pyramidal tract.

Fig. 2.

Photograpli of a transverse section of the fresh spinal cord (Cat) at the 7th dorsal nerve, slightly
magnitied. The remarkable differentiation of the posterior columns in this animal (Scuiff) is
well seen.

That these measures were effectual in localising the stimulus was demonstrated by
the fact that the placing of the electrodes on the neighbouring sections of the grey
columns evoked no electrical changes, as evidenced in the galvanometer connected
either with the issuing nerves or with another portion of the cord ; so that the

ON THE MAMMALIAN NERVOUS SYSTEM. 301

positive electrical results following stimulation of a column, the posterior, for instance,
were to be attributed entirely to the localised stimulation of the structures upon which
the points of the electrodes rested.

A further point is the special diiSculty of maintaining the electrodes properly in
position, i.e., in contact with the cord, inasmuch as the spinal muscles on their
contraction shake the cord a little, although the spine be fixed. This, of course,
applies only to the end excited. Imperfect apposition of the electi'odes was best
avoided by holding them in the hand, the latter being suitably supported. As is
described on p. 292, the duration of excitation was provided for by the revolving
mercurial key ; consequently, after the proper contacts had been made as above, each
observatioQ was comparable with another as regards the degree of the excitation.
The important bearing this has on the quantitative value of the galvanometric readings
is sufficiently obvious.

D. Nerve. — In electrical excitation of the nerve, we were chiefly anxious at first to
avoid possibility of spread. Beginning with ordinary sheathed electrodes, into which
the sciatic nerve was laid, we very soon laid these aside for the more accurate and
simpler plan of applying platinum electrodes to the nerve raised in the air by a thread.

The mode of exciting a nerve in relation to its cross and long diameter respectively,
has for some time been the object of research (Hermann, Bernstein, and others).
Gad,^" in referring to the action of carbonic acid, and especially of alcohol, on the
excitability and conductivity of nerve fibres, raises this question again, and
strengthens the fact that transverse excitation is much more adequate (5-6 times)
than stimulation applied longitudinally. Apart from incidental polarisation it is clear
that the larger number of fibres lying in the principal axis between the poles chiefly
conditions the result.

To produce a maximal effect we have, therefore, always applied the plathium
electrodes, so as to bend the nerve slightly between the terminal points. (See fig. 3.)

Fig. .3.

e

n \ ■*— .

e

Inasmuch as we always sought to obtain, with as weak an excitation as possible, a
maximal effect for purposes of quantitative as well as qualitative comparison, it was
especially necessary for us to excite, without fail, all tlie fibres in the nerve.

The usual method of simply placing the nerve on the electrodes, as, for instance, in
using sheathed or hook patterns, only enables a few fibres to be excited — those
immediately in contact, or nearly so, with tlie metal. Thus, Beevor and HorsleyI

* ' Archiv fiir Physiologic,' (du Bois Reymond). 1889, p. 350.
t ' Roy. Soc. Proc.,' 1888.

302 IMKSSHR. F. nOTf'H AXD V irORRTiKY "

have seen tlin'troiifiation lA' a mixed nerve on applying the electrodes to one or otlier
side of the tnud< nf the liypoglossal above tlie descendens iioiii br:nifli.

We have, of course, t'ound that great differences are produced according to variations
in the condition of the Mammalian nerve under varying circumstances.

The nerve must not be dragged upon in any way and the electrodes must be in
good moist contact with the bared fibres, the intervention of fat or fascia or dried
epineurium being sufficient to diminish or prevent the excitatory effect.

CHAPTER. IV.— TllIO UKSTING ELECTlUCAIi 1)1 FKKIIKNCE IN TWK M.\M.\l.\lilAN

NKllVE AND SPINAL COliD.

The experiments which form the subject of the succeeding sections furnished us
with a large number of observations as to the amount and character of the persistent
electromotive difference which exists in the Mammalian nerve and spinal cord
between the cross section and surface.

The number alone of the experimental observations would be sufficient to warrant
their introduction at this stage in a chapter devoted to their consideration only ; but
as in addition they seem to furnish valuable side evidence as to the relations of the
spinal cord both to the cerebrum and the issuing nerves, such an exclusive study
becomes a matter of necessity.

It is well known that when by means of appropriate contacts the cross section is
compared with the longitudinal surface of a living nerve, an electromotive difference
is found to exist, of such a character that if examined galvanometrically the surface
is positive to the cross section. The current associated with this difference was
termed by its discoverer, DU Bois-Reymond, the nerve current {nervenstrom), and
by others, Hermann, Hering, the demarcation current, since it is presumed by them
to have its seat in the zone of tissue which at once bounds and divides the region of
living from that of injured and dead or dying nerve. The use of either of these terms
is, we consider, objectionable in the present instance, since our object is merely to
state the actual fact without involving any cause thereof; we will therefore designate
the persistent electrical difference just referred to the resting difference.

The difference was in all cases estimated in the following manner. By means of
the long compensator, described in the preceding chapter on apparatus, a known
difference of electrical potential was introduced hito one part of the electrode circuit
of a sign opposed to that present at the electrode contacts, and the former was then
adjusted until the galvanometer showed no current to be present in the circuit ; the
original difference between the two electrode contacts is thus given in terms of
decimal fractions of the constant extei-nal source of difference of potential. As
already indicated, this soux-ce was the Callaud cell, the E.M.F. of which was itself
estimated by a similar balancing method in terms of a carefully prepared Daniell cell.

ON THE MAJIMALIAN NERVOUS SYSTEM. 303

The special structures investigated fall into three classes — ■

A. Sciatic nerve.

B. Posterior root.

C. Anterior root.

D. Spinal cord.

A. Sciatic Nerve.

1. Amount of Difference.

The resting difference between the cross section and the longitudinal surface of the
Mammalian nerve was estimated by DU Bois-Reymond as "026 I), in the case of the
Rabbit;* a much less mai'ked difference, '005 D., was observed by IsRAELt in the
sciatic of the same animal. The amount of the difference in the sciatic nerve was
observed in various Mammals by Frederioq,;]: who obtained the following results : —

Cat . . .

. -017 to

•018 Daniell

Dog . . .

. -018 „

•024 „

Rabbit . .

•015 „

•028 „

Duck . .

• -024 „

•026 „

Horse . .

. -004 „

■007 „

Our own results were obtained with the Cat and Monkey. In all cases the sciatic
nerve of the anaesthetised animal was exposed in the popliteal space, ligatured, and
divided below the ligature.

The results obtained with 60 sciatic nerves in the Cat are given in Appendix B, I.,
these are separated into those in which, the cord, and thus the nerve, was connected with
the brain, and those in which the cord was divided. The average result of both sets is
the same, being -0094 Daniell in both. The influence of complete section of the nerve
at the sciatic notch was also observed in some cases, and was found not to affect the
amount of the difference for the first minute or two. The highest difference ever
observed was Cat (70)§ and Cat (75), when it amounted to -018 Daniell, the lowest
was "004. Whilst then the average amount is much lower than that which previous
observers above referred to obtained in the Rabbit, the maximum corresponds to that
noted by Fredericq in the Cat.

The results obtained in the case of the Monkey are shown in Appendix B, I. Twelve

* ' Gesammelte Abhandl,,' vol. 2, p. 250.

t Israel, ' Archiv f. Anat. u. Physiol.,' 1877 (' Physiol. Abth.,') p. 4.51.

t Feedeeicq, ' Archiv f. Anat. u. Physiol.,' 1880 (' Physiol. Abth.,') p. G5.

§ In all cases where a number in brackets follows the mention of an animal, e.g., Cat (70), it refers
to the page in our note ledger. The number of observations is so large that, for purposes of control and
correction, we have been compelled to thus check the record.

304 MFJSSRS. P. GOTCli AND V. HORSLEY

nerves were exaiuiiud and tlie avenige difference anioiiiited to only "OOS Daiiinll,
whilst the highest difference observed was "007, and the lowest '003. This diminu-
tion in the case of the Monkey cainiot be attributed to tlie smaller size of the sciatic
nerve in the common Macaque Monkey, which is generally experimented upon, since in
two cases large Rhoosus Monkeys with big nerves were used, and in tliese (333), (3G8)
the highest difference is "007, and the average '0055. (Compare the proportions of the
nerves as shown in Plates 30-35.)

Moreover, as the previous results obtained by Fkedericq show, the amount of the
difference does not viuy directly with the size of the nerve when different classes of
animals ai'e used, of which the high difference in the Ilal)l)it and the low difference in
the Horse is a notable illustration.

It will be seen that the spinal cord resting difference in the Cat and Monkey
exhibits the same contrast as that of the sciatic nerves.

It has been stated that the resting difference in the nerve is not perceptibly affected
by division of the spinal cord, when the division is carried out at the time of the
observation. It appears, however, that section of the cord one to two months prior to
the experiment does affect the amount of the difference. This influence, though small,
seems to be sufficient to enable the observer to judge, when the cord lesion has been
unilateral, upon which side it is situated.

Thus in an animal. Cat (227), in which the left lumbar posterior roots had been
divided, the left nerve diffei'ence was '007, that of the right '009. In another animal,
Cat (223), in which the right posterior column had been divided, the right nerve
difference amounted to "005, the left '008, and in an animal (259) upon which a
hemisection on the left side had been effected, the left nerve difference was '007, the
right -008.

In one case after section of both posterior columns (Cat) the difference in each nerve
was "012, an amount wliich is somewhat in excess of the average.

2. Alterations in the Resting Difference.

The alterations in the difference occurring whilst tlie nerve was under observation
may be divided into : —

(a.) Those due to physical changes, affecting the contacts and derivation currents,
and due to loss of moisture from evaporation, change of temperature, (fee.

(b.) Those only to be accounted for as connected with slow jAysiological changes in
the tissue, the chief agent in such alteration being loss of vitality, due to arrest of
the cii'culation, and to other causes.

(c.) Those directly connected with the production in the tissue of excitatory
physiological processes.

We will now examine tliese separately in detail.

(a.) Physical Clianges. — With regard to the effect of physical changes, such as

ox THE MAIMMALIAN NERVOUS SYSTEM. 305

temperature, drying, &c., these influence the electrical resistance of the tissue siir-
rounding by altering the moisture with which it is infiltrated. The result is to cause
an alteration in the balance j^reviously existing between the currents due to the tissue
and the compensator source. The character of this will l)e made evident by citing a
case.

The sciatic nerve, prepared as above, and hanging free in air, was connected by its
cross section and surface with the galvanometer electrodes, the resting difference
between the contacts was compensated at first by a difference of "01 D. The nerve
was then moistened between the contacts, this necessitating a reduction of the
compensation to "0085 D. This alteration is obviously due to the fact that the
increased moisture has decreased the resistance in the surface of the nerve, and
thereby increased the intensity of all currents which are present in that region.
These ciuTents are, first, a portion of the balancing current, and secondly, the derivation
currents between the surface and cross section of the nerve. Whilst these two are
equally affected between the points of contact of the electrodes, the result as regards
the galvanometer circuit is very different. This will be rendered clear if we remember
that the tissue difference is usually that obtained by connecting two surface points of
a series of closed circuits, through which currents flow, and that the resistance of the
circuits varies inversely with the amount of tissue moisture, so that the amount of
spread, and thus the particular derivation on the surface, is diminished by the
increased moisture. The final result is that, as regards the galvanometer circuit, there
is now a current in the direction opposed to that due to the tissue difference, this
being due to the over compensation, and rendered more distinct by the lessened
resistance in the whole galvanometric circuit which the moistening involves.

The effect of drying must be the reverse of that indicated. It became noticeable
in our prolonged experiments upon the Mammalian nerve and spinal cord, since it was
essential for our purpose to perform the experiments in a room at about 70° F. As
far as possible the errors due to this were obviated by the use of steaming sponges
placed under the structures, but a slow steady rise in the amount of the difference
during the first 5 or 10 minutes after the electrodes have been applied, and
maintained unaltered, is usually seen in the cord, vide infra, and although not a very
noticeable feature in nerve, may, when present, be partly attributed to this cause.

As just indicated the temperature was kept tolerably constant, but, as Hermann
and others have shown, the gradual rise in the temperatui'e of the experimental room,
may cause a similar slow increase in the amount of the difference. This alteration
is, however, strictly of physiological origin, being related to a change in the molecular
condition which lies at the foundation of the diff"erence.

This brings us at once to the alterations which may be more especially ascribed to
physiological changes.

(b.) Slow Physiological Changes not Obviously Excitatory. — The alterations of this
kind in the resting difference due to physiological changes are in the case of nerve

MDCCCXCI. — B. 2 R

306 MESSRS. F. GOTCII AXn V. HORSLKV

usually a fall in (lie amount. Sometimes it has happened that a ra])i(l I'ise for the
first few minutes has been observed which did not admit of explanation as being
caused by either of the two agents just indicated ; and the presence of such a rise
to a very marked degree in the case of the spinal cord as well as its ]>resence in the
isolated nerves and muscles of the Frog, renders it probable that in the Mammalian
nerve, when present, its origin is physiological.

The fixll is undoubtedly connected with loss of vitality, and may vary between the
slow dimiiuition which follows arrest of the circulation, and the rapid fall produced by
injury of the nerve in immediate proximity to the surface contact.

The alterations which especially call for remark here are those which are entailed by
the death of the animal on the one hand, and by the severance of the nerve from the
spinal cord on the other. The first is probably due to several factors whilst the
second is due both to arrest of the local circulation and the destruction of the con-
tinuity of the nerve fibres with their trophic centimes in the cord.

In the case of the sciatic nerve, the alteration seemed to us to be the same in
extent whichever event occurred ; that is to say, the resting difference for the first
30 minutes slowly diminished in amount at the average rate of momj Daniell per
minute, this rate of diminution becoming less afterwards.

It is, however, questionable to what extent in the suspended and isolated nerve, the
physiological conditions which ai'e dependent upon continuity with the cord are
maintained by any facilities which the preservation of vessels affords for keeping up a
circulation in the nerve, since this must be, undoubtedly, greatly impaired by the
exposure. The conditions of vitality would rather seem to be linked with the
maintenance of physiological connection with other nerve structures, around which
an active cii"culation is still being carried on, and which, therefore, retain their normal
vital characteristics. That this is probable is shown by the fact that when, as some-
times occurs, the nerve difference, owing to rapid drying, shows a steady rise in
amount, which rise continues even when no evidence of circulation in the exposed
nerve is obvious, the rise is immediately counterbalanced and converted into a fall on
systemic death of the animal, this fall occurring pari passu Avith the loss of excita-
bility in the central nerve structures.

(c.) Physiological Changes Connected tvith Excitation. — The alterations due to
excitation were after effects, that is, they followed the development of the transient
excitatory change and consisted in a permanent slight diminution in the previously
existing difference, such as has been observed to follow nerve excitation in the case of
the Frog.

The slow after effect can be readily distinguished from the transient rapid electrical
changes due to the actual presence of excitatory processes evoked by the application
of the stimulus, since these are synchronous with the excitation, whereas the after-
effect is very variable in its duration, but always follows cessation of the stimulus, and
persists for many seconds.

ON THE MAMMALIAN NERVOUS SYSTEM. 307

We have noticed that it is more pronounced when the etherisation has become
shght in degree.

In addition to the after-effect due to excitation there are seen under certain circum-
stances, particuhirly that of inadequate etherisation, changes in the nerve diflference,
whicli in the galvanometer indicate their presence by the needle now slowly rising
30 to 50 scale, and now falling. These disappear when the anaesthesia is rendered
more profound, and are much more marked in the case of the spinal cord ; they are
probably due to the occasional discharge of groups of minimal nerve impulses, and are
analogous to those fii'st observed in the medulla of the Frog by Setschenow.

B. Posterior Root.
1. Amount of Difference.

We have made six experiments upon the posterior root. In these a lumbar posterior
root was exposed, ligatured, and divided as close to the ganglion as possible, without
involving this structure, and the root thus left connected with the cord was suspended
in air by the attached thread. The particular root chosen in five cases was the 7th of
the lumbar series, since this is the largest in the Cat of the roots forming the lumbar
plexus (see Plate 35). The galvanometer electrodes were attached, as in the nerve, by
means of thread cables moistened with "6 per cent. NaCl, to the cross section and the
longitudinal surface, and the amount of the resting difference determined by the
balancing method.

The result in the five animals was found to be "026, "02, "02, 'OlS, and "016 Daniell
respectively.

Hence although the 7th lumbar root is less than half the size of the sciatic nerve,
the average amount of the resting difference, "02, is about twice as large. (The Gth
lumbal" root was observed once, the difierence being ■012.) This remarkable cii'cum-
stance is very significant when the proximity of the structure to the spinal cord is
taken into consideration, and deserves a more extended investigation. It may be
pointed out that Schiff observed that there was a resting difference between contacts
when placed upon a more central and a more distal portion of a continuous nerve, the
central contact being positive to the other ; also that excitation of any portion of the
tract caused an excitatory change, in which the tissue under the first contact became
galvanometricaliy negative to that under the second more remote one.

Further, both Grunhagen and Bernstein have noticed that the central portions
of the nerve are more excitable than the distal portions. Now whether the persistent
difference be fundamentally the same in kind as the transient excitatory electrical
change or not, there is an undoubted quantitative relationship subsisting between
the two. It is, therefore, not surprising that there should be found an increase
in the amount of the resting difierence obtainable in the posterior roots as compared
with the sciatic nerve.

2 R 2

308

MKSSH.«!. F. flOTCMI .\\l) V. I!OI!SI,KV

2. ^llU'ralions in RvsLiikj Dljj'trence.

((<.) Physical Changes. — -The alterations due to drying probably luive the same
influence as that already alluded to in the case of the sciatic nerve. The amount of
the difference noticed immediately after section generally increases slowly and steadily,
and, at least partially, from this cause. There is, however, a noticeable check in the
rise after the first five minutes, after which the difference then only in some cases
begins to fall. Hence either the rise is accentuated at first by some other agency, or
it is counterbalanced by the depressant effects due to loss of vitality.

(b.) Physiological Changes. — The amount of the difference in the root diminishes
with greater rapidity than in the case of the nerve trunk, in consequence of systemic
death. A similar fall occurs if the root is severed from its central connection. This
is illustrated by the following series of experimental observations made before and
after the death of the animal.

Left 7th Lumbar Posterior Root.* (Cat, 3G2.)
Root Divided near Ganijlion and Central End Connected with Electrodes.

Excitation experiments carried ont
Systemic death

Time of

Amount of

observation.

diii'erence.

11.42

•019

r 11.45

•02

<^ 11.51

•0195

[11.55

•019

12.0

12.3

•0185

12.12

•018

Root cut off from Cord.

Time of observation.

Amount of diilerence.

12.13
12.14
12.15
12.16

12.18

•0175

•017

■016

•0155

•015

In another animal after systemic death had caused a decline in the amount of
difference, severance of the root caused no further appreciable decline.

* See Plate 35.

ON THE MAMMALIAN NERVOUS SYSTEM. 309

The root, therefore, in this respect behaves, as far as these two experiments enable
us to judge, like the nerve.

(c.) Changes following Excitation. — In consequence of either direct excitation of
the root or its indh-ect excitation through the cord, the amount of the ditlerence is
decreased, an after-effect following the passing away of the true excitatory change
similar in character to that seen in the nerve.

The rising and falling in the amount of the difference when the anaesthesia of the
animal is not sufficiently jirofound is rather more marked in the case of the posterior
root than in that of the mixed nerve.

C. Anteeiok Root.

With regard to the anterior roots, we have only performed one experiment. There
are considerable difficulties in the way of connecting the anterior root satisfactorily
with the electrodes, so as to exclude all possibility of obtaining results which are
derived from electrical changes in the cord. These difficulties are mainly the shortness
of the root and the characters of its anatomical relations with the cord, it being
difficult to obtain a sufficient length. Moreover, our purpose being rather to obtain
the evidence of excitatory than of resting electrical changes, the shortness of the root,
when combined with the movements of the animal, caused serious errors in observa-
tion, due to the displacement of the electrodes and their being thus brought near to or
in contact with surrounding structures. In the single case in which the 6th left
lumbar anterior root was divided and its central end connected with the gfalvanometer
electrodes, the difference (Cat) was found to be only "0045 Daniell, whereas the 6th
lumbar posterior root of the same side, when examined, showed a difference of
•012 Daniell.

D. SriNAL Cord.

The fact that the cord exhibits, like the nerve, both a persistent difference between
its longitudinal surface and its cross section and a true excitatory effect, was dis-
covered by DU Bois-Reymond, and has been since confirmed by other observers, notably
Setschenow. The general features of the same were described by us in our preliminary
communication made to the Royal Society, but a systematic analysis of a number of
observations has never, that we are aware, been made, nor has the amount of the
persistent difference been determined. In the course of our investigations we have
noted the amount and characteristics of the resting difference in the spinal cords of
sixty-three Cats and fov;rteen Monkeys, and the results arranged in groups are given
in Appendix B.

The most striking characteristics of the resting difference in the cord and the
contrast between this difference and that of the nerve may be roughly seen after the
death of the animal (Cat, Rabbit, &c.) by rapidly ex])osing a length of cord and of

310 MK.SSIJS. F. ClOTCll AND V. liOiJSI.KY

the sciatic nerve, removing the exposed portion of each, placing it on a glass plate
and brincrinir one electrode in connection with the cross section, the other with the
longitudinal surface of the preparations. It will be found that in the Cat the nerve
dill'crence amounts to little over •01 Daniell, whereas the cord difierence amounts to '025
to "03 Daniell or more. That this contrast is not merely a ciuestion of cross sectional
area is shown by experimenting with the dorsal cord ol' a small (young) animal and
with the lai-ge nerve of a full grown adult animal of the same species, when the same
relations will be found to still hold good, viz., the cord difference twice to three times
the amoimt of that of the nerve.

For purposes of experimental investigation however, the spinal coi'd was exposed in
the lower dorsal or lumbar region under all the precautions mentioned in the
paragraph upon operative procedure ; it was then divided and freed as required on
either the central or peripheral side of the division for 4 or 5 centims. from all its
attachments, the severed end was ligatured and this portion of cord raised from the
canal and suspended in air by a ligature, still retaining its continuity with the
exposed cord at its deep attachment. The raised portion was in some cases that on
the central side of the section, in which case the continuity with the brain was pre-
served, in others that on the peripheral side, in which case the connection with the
sciatic nerves was preserved.

] . — Amount of Difference.

(a.) Whole Cord. — If an average of the various results obtained in the case of fifty
Cats is taken, the exposed cord being normal and always in connection by its deep
attachment with a portion of cord in situ, but without distinction of the nature of this
attachment, it will be found that the average resting difference in the Cat is '032
Daniell, the highest difference being "046 Daniell, the lowest "014 Daniell.

A similar average of the results in nine Monkeys gives a resting difference of "022
Daniell, the highest difference observed being "029, the lowest -013.

It is thus evident that the relationship between the amounts of the resting
difference obtainable in the Cat and Monkey holds good for both the sciatic nerve
and the cord, the amount in the Cat being very appreciably larger than that in the
Monkey. The notable increase in the size of the resting current in the cord as
compared with the nerve of the same species will be rendered evident if we place side
by side the average results of the two. Thus : — -

Cat cord _ "032 _ 3-3
Cat nerve '0095 1 ' *

Monkey cord _ "022 _ £4
Monkey nerve '005 1

That this increase in the proportion of about 4 to 1 is not merely due to increased

ON THE MAMMALIAN NERVOUS SYSTEM. 311

size has been roughly indicated previously ; this will, however, be more clearly
exhibited by the consideration of the ratio which a similar comparison between the
difference in the cord and in the posterior root of the Cat shows : —

Average Cat cord _ -032 _ 1-5

Average Cat posterior root -022 1

It is probable therefore that the larger differeiice is in some way correlated with
the structure of the cord and the trophic and other physiological influences which
that structure implies.

This explanation is supported by an analysis of the results obtained in the two
species of animals when the cord is placed under different conditions as regards its
central connections. Such analysis involves the grouping together on the one hand of
all those determinations in which the cord was still in connection with the cerebrum and
its central end investigated, and on the other hand of all those in which the portion
of cord examined was that of the lower fragment unconnected with the cerebrum.

The first group shown in Appendix B, III., gives the average result in the case of
the lower dorsal cord of fourteen Cats, this average being •034 Daniell ; the highest
amount observed being '046, the lowest •025.

It also gives the average result in the case of five Monkeys, this being ^025 Daniell;
the highest reading being "029, the lowest •018.

If now, we turn to the second group, the average in the Cat of all the readings in
the dorsal and lumbar regions respectively when the Cat's cord is severed from the
brain, amounts to : —

Dorsal region, 24 cases, average ^029
Lumbar ,, 10 „ „ '033

34 cases

Gross average = "03 Daniell.

Highest = ^043 ; lowest = •014.

In the Monkey a similar average of four sets of experiments gives '019 Daniell ;
highest -027, lowest -013.

A comparison of the two groups of results may now be made as follows : —

Cat-
Cord connected with brain "034

Cord severed from brain •OS

Monkey —

Cord connected with brain '025

Cord severed from brain '019

312 MESSRS. F. GOTCII AND V. IIORSLHY

The comparison shows that the amount of the resting difference in the anajsthetised
animal is hir>^er when the cerebral connections are intact.

This conclusion is supported by direct experimentation upon the inlluencc of
severance. Thus in an anaisthetised animal (Cat) the cord was exposed and divided
at the level of the 4th lumbar vertebra and the central end ligatured. It was then
suspended in the usual way and the electrodes attached to the cross section and the
longitudinal svu'face. The resting difference amounted to •044 Daniell, whereas, when
the cord was severed higher u]i at the 7th dorsal vertebra, a notable fall to "037
Daniell occurred.

In another experiment (Cat, 315) the cord was divided at the 13th dorsal vertebra
and similarly prepared. Its central end was then connected with the galvanometer
electrodes as above. The amount of" the difference between the surface on the central
side of the cross section and the cross section itself amounted to "04 Daniell. A
hemisection of the cord at the 8th dorsal vertebra was now made and the difference fell
rapidly to '38 Daniell, the rapid fall then ceased, but a slow fall continued for some
time.

These experiments suggest that severance in an anaesthetised animal of the cord
from the brain causes a direct diminution in the amount of the resting difference
below the point of severance, whether by the interference with the circulation which
such severance may imply, or by this aided by the interruption in the path which
joins at any rate some of the (pyramidal) fibres with the cells with which they are in
direct connection and which govern their nutrition. As will be seen later there is no
evidence of a similar diminution being produced when the cross sections are made
upon a portion of cord already severed from the brain by an interruption higher up.
Hence the inference that the fall is correlated with the interruption (in the pyramidal
fibres 1) is strengthened.

Finally the region of cord investigated has some relation with the amount of the
resting difference. This is shown m Appendix B, III., in which observations made in
the dorsal region are separated from those in the lumbar region.

The average of the dorsal region = '029 ; highest '043, lowest '014.

The average of the lumbar region = "033 ; highest '04, lowest '028.

Although the average seems to show that the difference in the dorsal cord is
less than that in the lumbar, yet an examination of the highest and lowest limits
suggests that the preponderance, in the latter case, is due to the fact that in the
dorsal cord several very low readings were obtained. (See Table (2) ; Appendix B,
III, Cats 194, 289, 349.)

We have endeavoured to get some notion of the relations of the different regions by
exposing in an ancesthetised animal (Cat 36) the whole cord. This was then divided
into three equal lengths, comprising the cervical, dorsal, and lumbar regions respec-
tively, and the three placed in a warm moist chamber for examination.

ON THE MAMMALIAN NERVOUS SYSTEM. 313

The examination was made as rapidly as possible in the order given below, and at
the various regions indicated.

Lumbar segment —

From 4th lumbar "OSS to about 11 th dorsal '04.

Dorsal segment — -

From about 11th dorsal '043 to about 3rd dorsal -037.

Cervical segment —

From about 3rd dorsal "035 to about 3rd cervical •037.

After a quarter of an hour's interval, the experiment was repeated in the reverse
order, commencing with the cervical and ending with the lumbar segment. The
results were as follows : — ■

Lumbar segment "035 '036,

Dorsal segment -030 "021,

Cervical segment . . ' -031 -029.

"&

Li this experiment, the dorsal cord resting difference is greater than that obtained
from either enlargement, but it is more affected by the injurious influences of
severance and loss of circulation. It must be remembered, that there are three
possible factors in the severed cord which determine the amount of the resting
difference in any given region ; the number of fibres, the amount of grey matter, and
the maintenance of vitality through circulation, &c., this last being influenced by size
of vessels, &c. All these vary in different regions, but the first is preponderant in the
dorsal region, the others in the enlargements.

(&.) Cord Divided Longitudinally. — In the course of experiments to be detailed
hereafter, we had occasion several times to divide the cord longitudinally into two
halves. This division was carried out under all the precautions indicated in the section
dealing with operative procedure, and was always in the form of a longitudinal cut
connecting the anterior and posterior fissures, and extended from the cross section of
the divided cord fin- three centimetres. Since it was our main object to establish a
quantitative comparison between the crossed and the direct excitatory cord eflect
evoked by cortical stimulation, the cord was generally divided in the lower dorsal
region, and the lower end of the fraginent on the central side of the division was split
in the fashion just described. (See Plates 31, 33.) The results in seven Cats and five
Monkeys are given in Appendix B, III.

Each half of the split coi'd was connected with a pair of non-polarisable electrodes
by the cable arrangement already indicated, and the amount of the difference between
the cross section and the surface in each then determined. It will be seen that the
average amount for each side in the Cat is '02 Daniell, the highest and lowest readings

MDCCCXCI. — B. 2 S

314 MESSRS. F. (iOTCII AND V. HOIISLHY

heiiig -020 mkI 01 in the case of tlie left s'ulo, and "027 and '014 in that of (he liL^hl
side.

In the Monkey, the average result for the left side is 'OK), that l\n- the il;^rht 'Ol 1.
It will he noticed that in each animal the snni of the two readings is slightly larger
than the average for the unsplit cord.

A further table (4) give the results of a similar splitting operation carried out on
the u[)per end of the lower peripheral fragment of the divided cord in two Cats.

In one Cat (137) the left half of the split cord showed a resting difference of '007,
and the right -016 ; in the other Cat (143) the left half showed -021, and the right
"013. The amounts are considerably below those just mentioned, one reason doubtless
being the fact that the portion of cord upon which the operation was carried out was
severed from the brain.

(c.) Cord previoudy operated upon. — It has already been pointed out that, while
severance from the brain seems to induce a condition in the separated cord in which
the amount of the resting: difference is less than it otherwise would have been, further
operations upon this severed fragment do not seem to notably affect the difference,
provided they are made in such situations as to leave a part of the cord both intact
and in situ between their seat and the portion observed.

That is to say, the difference in the lower fragment of a cord severed from the brain
is not further diminished by exposing it again in its continuity and performing there
either section of one or of several columns.

It is different, however, when such operative lesions have been performed several
weeks before the experimental investigation. In these cases the average amount
of the resting difference observed was "022 Daniell. that is less than the normal,
but the differences between the experiments are best shown by reference to the column
in Appendix B., III. (5). They will then be found to vary between '012 (cord one
month after section of posterior roots), and "OSS (cord four montlis after hemisection
on left side). This last is an exceptionally high result when compared with that
obtained after a very similar two months' previous hemisection when the amount of
the difference was only "017, and also that obtained after a one month's previous hemi-
section on the opposite side, when the amount was "018.

In a case of one month's previous lesion of both posterior columns, the amount of
the resting difference was found to be "022, and after a lesion of one posterior column
it was found to be "025.

If "030 is taken as the average amount of the difference in the normal Cat, after
severance of the cord from the brain, then a descending series occurs in these cases, as
shown in the table, in all of which with one exception the demarcation current is
below the normal, and the average of these, when this exceptionally high and contra-
dictory result is excluded, would be "019 Daniell.

ON THE MAMMALIAN NERVOUS SYSTEM. 315

Table of Results after Previous Operations.

'012 after section of posterior roots.

•017 ,, liL-niisection.

•018 „

'022 „ section of both posterior columns.

•025 „ M one

Total . . -094

Average. . "019 Daniell approx.

2. Alterations m Resting Difference.

The electrical difference between the surface and cross section of the spinal cord
alters during observation in the manner already described in treating of the sciatic
nerve and root, but the alterations are much more marked.

{a.) Physical Changes. — The effect of drying is the same as in the case of the
nerve, the amount of the resting difference continuing to rise steadily after exposure
and isolation of the cord unless great care is taken to keep it moist. It is difhcult to
ascertain with pi-ecision to what degree this physical change is capable of causing a
rise in the difference, this rise being about '0001 Daniell in five minutes ; but that it
is by no means the sole agent is shown by the fact that even when the exposed
cord Is kept moist by steaming sponges with as much care as possible, the rise con-
tinues, though more slowly than when no such precautions are used, as also by the
facts to be referred to in the succeeding paragraphs. It is, however, important to
keep -in view in the consideration of alterations supjoosed to be due to strictly physio-
logical agencies the influence of these purely physical ones.

{h.) Phijdolugical Changes not obviously Excitatory. — The alterations due to
physiological changes are associated with both a rise and a fall in the amount of
difference, and are much more marked in the case of the coi'd than in that of the
nerve or the root. The amount of the rise is demonstrated in the following obser-
vations, in all of which the cord having been exposed, ligatured, and divided, was
suspended by its unattached end and connected with the electrodes in the manner
described before.

It was then observed that in every case the resting diflerence between the surface
and cross section increased rapidly for the first five minutes or more after the isolation
liad been made, provided that the cord was still in connection by its deep attachment
with a part which, being in situ, was in its normal physiological state of nutrition.
The amount of this initial rise varies considerably in different cords, but the average
in 19 cases amounted to "0016 Daniell, the highest rise being "OOS and the lowest
■001. The duration of this comparatively rapid rise is, on the average, about five

2 s 2

310

WKSSRS. F. COTCll AND V. HOUSl-KY

minutes. 'I'his initial increase is probably of the same character as that observed
by DV Bois-Reymond, Hermann, Engelmann, and others in both muscle and nerve,
and is in some way related to the develo2)ment of changes prodtieed by the cross
section.

The difference continues very slow ly to increase in the unstimulated cord, at a rate
of about '0001 in five minutes, this change being probably connected with interstitial
drying. There is, however, a marked increase in the difference when the cord is
aroused by successive stimulation, and this last rise is the most striking and novel
feature displayed by the cord : it will be treated of under the excitatory changes.

If systemic death occurs, the difference immediately begins to decline, and this is
such an invariable result that we have often observed the first approach of death by
the behaviour of the galvanometer needle. On separation of the cord liom the body
a similar decline occurs, the rapidity of which is shown l)y the following results : —

Decline in Difference following Systemic Death.

Cat (246) . . .
Cat (2.58) . . .
Cat (323) . . .
Cat (345) . . .

•029
•031
•039
•04

5
minutes.

10
minutes.

15

minutes.

20
minutes.

25
minutes.

.30
minutes.

•028
•03
•038
•037

•026

•028

•024
•026

•023
•025

•022
•024

• •

•023
•03

From the above oases it will be seen that on systemic death the resting difference
declines to such an extent that in half an hour the total fall amounted to '007, "008,
and "009 Daniell. How far this fall is associated with the cessation of the circulation
in the actual portion of exposed cord under investigation is a moot point, but the
reasons brought forward on p. 283 with reference to the similar question in the case of
the sciatic nerve are applicable to that of the spinal cord. The local circulation is
undoubtedly seriously impaired by the exposure ; it is, therefore, rather to the
structural connection with the portion of cord in situ, in which the circulation is
adequately maintained, than to the integrity of its own blood supply that the main-
tenance of the difference in the case of the livintr animal must be ascribed. It is
wonderful what prolonged exposure a portion of the spinal cord will sustain without
losing its excitability, provided only that the local conditions of adequate moisture
and warmth are fulfilled and tha,t a structural connection is kept up with normal
unexposed cord substance.

The above decline is contemporaneous with a marked lessening of excitability. As
is shown in the chapter upon cord excitation, all evidence of excitatory change in
response to stimulation disappears a few minutes after systemic death.

If the fragment of exposed cord is wholly removed from the animal, the difi'erence

ON THE MAMMALIAN NERVOUS SYSTEM. 317

i-apidly subsides, as is shown in the experiment described on p. 313, as made upon
such pieces of spinal cord.

The decline in the difference thus produced, seems to occur with greater rapidity
than in the case of systemic death. It is illustrated by tlie following experiment
upon the cord of the Cat (3.')4), in which, at the close of an exposure and series of
investigations, which had lusted an hour and a lialf, the exposed piece of cord was
cut free from its deep attachments, and the consequent decline in the resting
difference noted.

This difference had }>reviously been rising aU through the experiment, ever since
the cord which had been exposed and divided at the 1st lumbar vertebra was first
examined. The portion investigated galvanometrically was on the central side of the
above section, but the upper part of the cord was not connected with the brain, since
a second complete section at the level of the 7 th dorsal vertebra had been made for
])urposes of stimulation.

The amount of the initial restuig difference between the cross-section and surface
of the investigated portion observed (at 11.4) was '033 Daniell ; at tlie close of the
experiments (12.39) it was "038 ; the exposed portion of cord was then severed from
its deep attachment so as to be wholly detached ; in two minutes (ll2.39) it was "037 ;

(12.41) .... -030

(12.43) .... -034

(12.45) .... -033

(12.47) .... -032

That is to saj', a fall in the difference amounting to OOG occurred in ten minutes ;
in another hour the difference had fallen to half its original amoiuit = -019 Daniell.

Both the slower fall in consequence of systemic death and the more rapid f;ill in
consequence of a complete detachment from the rest of the nervous sjsteTu, contrast
with that produced by the change following injury sustained l^y the cord even when
in situ. The mechanical injury which seems most fatal to excitability, is that pro-
duced by stretching of the cord, a sudden pull heing followed almost immediately by
the disappearance of any excitatory effect and by a rapid fall in the resting
difference Thus, in one animal (Cat 126), the difference was '035, but owing to the
arrangement for fixing the spine, &c., becoming loose, the suspended cord was pulled
by a movement of the preparation, and the difference fell suddenly to '017. The
whole cord, however, was not injured, but only the portion observed, for a new
dissection exposed a fresh part and a higher section was then made, which gave a
difference of "028.

(c.) Physiological Changes connected with Excitation. — It has been already indi-
cated that a characteristic feature of the spinal cord difference is the alteration pro-
duced in it by stimulation. It will be seen by the exjjerimental details, which are
set forth in the following chapters, that marked excitatory electrical effects ai-e evoked

318 MKSSRS. F. GOTt'll AX1» V. IIOHSI.KY

ia the cord by .stimulating either the cortex of tho biiiin, the difVcrent regions of the
cord itself, or the issuing posterior roots. These effects resemble in character those
evoked in the nerve trunks themselves ; the electrical change commences with the
stunulation, is opposed in direction to the resting difierence, and subsides on the
cessation of the .stimulus. Fullowing this true excitatory effect is an after-effect,
which is characterised by an increase in the previous difference.

In any giveii experiment the after-effect shows itself thus. The cord having been
exposed, divided, prepared, and coiuiected in the usual way with the electrodes by its
cross section and its longitudinal surface, the amount of the resting difference was
noted. This difference rises as just stated for the first few minutes, and then remains
very nearly steady. The cord is now excited in any of the three ways mentioned
above ; an electrical change occurs, i.e., a current opposed to the difference is present
in the galvanometer circuit which, on the cessation of the stimulation, subsides. The
galvanometer needle, which evidences the existence of this current, having moved
from its resting zero position through a certain deflection, returns to its ja'evious
position. If the nerve were the tissue under investigation, this return would be not
quite complete, but with the .spinal cord it is characterised by continuing beyond the
zero position to a point considerably the other side, after which the needle slowly
moves back again towards zero, which, however, it does not reach. As the result of
excitation, therefore, the previously balanced difference is no longer compensated ; a
current persists in the circuit in the direction of that produced by the difference, and
to compensate this, and thus bring the needle back to zero, the balancing circuit has
to be readjusted.

In short, the excitation has caused, after the usual negative variation, a
permanent rise in the resting difference between the two electrode contacts. This,
positive after effect is not unknown in other structures, it occurs occasionally in both
muscle and nerve preparations of the Frog (Hering, Head), but the conditions
necessary for its production are but imperfectly ascertained. It is, however, as far as
we know, invariably present in the case of the spinal cord of the Cat and Monkey,
though its amount is very variable in different preparations.

In almost all cases the amount of the rise is dependent rather upon- the condition
of the prepax'ation than either the intensity or duration of the cord stimulation. It is
greatest as a consequence of the earlier stimulations, and becomes less and less with a
repetition of the stimulus.

If, therefore, a series of experiments are made involving the stimulation of
the coz'd at successive intervals, the resting difference rises at first rapidly, and then
more slowly, until after 20 to 30 stimulations in from 30 to 45 minutes, the rise
finally practically ceases.

The stimulus used, whether applied to the cord or to structures connected with the
cord, was in all cases that of the alternating induced current previously alluded to in
the chapter on the experimental method. The amount of the total rise observed in

ON THE MAMMALIAN NERVOUS SYSTEM.

319

different cases varied according to the seat of the stimulus which evoked the cord
change. Tiiis seat may be either (A) Bruin, (B) Nerve, or (C) Cord, and the cases
in which we specially noted this change are as follows : —

A. Cortical Excitation, (^ord connected with Cortex.

Initial difference.

Final difference.

Total rise.

Cat (114) ....

•037

•039

•002

„ (124) . .

•046

•047

•003

„ (126) . .

•036

•040

•004

„ (308) . .

•033

•036

•003

„ (319) . .

•030

•035

■005

„ (323) . .

•032

•033

•001

Monkey (234)

•026

•029

■003

Average

rise -003

B. Nerve Excitation (sciatic). Cord severed from Cortex.

Initial difference.

Final difference.

Total rise.

Cat (121) ....

•031

■034

•003

„ (143) .

■032

■035

•003

., (145) .

•028

■035

■007

., (148) .

•023

•026

•003

„ (153) .

•025

•029

•004

„ (192) .

•023

•030

•007

„ (194) .

•017

•020

•003

„ (327) .

•023

•034

•Oil

„ (339) .

•022

■om

•008

„ (344) .

•022

•025

•003

„ (170) .

•026

•029

•003

Average rise "005

3:iO

MESSRS. F. CO'l'CII AND \'. lloUSLKV
C. CoUD Excitation. Coril severed from Cortex.

Initial differencp.

Final iliiTerencc. 'i'o

till rise.

Cnt (196) ....

•017

•030

(ii;!

„ (2-14) .

•027

■029

002

,. (245) .

•021

•02(5

005

„ (357) .

•032

■038

006

., {366) .

•030

•037

007

„ (355) .

•037

■041

004

,. (371) .

•028

•035

007

„ (375) .

•032

■045

013

.. (378) .

•027

•029

002

„ (243) .

•027

•029

001

„ (354) .

■033

■038

005

Average rise

006

These results indicate (1 ) that the rise in tlie difference is occasioned not merely by
the direct application of tlie stimulating agent to the cord, but as a consequence of
the presence of a series of excitatory processes, whether these are produced by nerve
impulses entering below by afferent channels, or from above by cortical efferent ones.
This conclusion is supported by the fact that a similar rise follows a large reflex
discharge of energy from the cord, when this is produced in the strychnised animal
by sensory stimulation.

(2.) They also show that the rise is least in the case of the excitatory cord changes
evoked by cortical stimulation, in which case the limit of rise is not only small, but
soon attained, the average rise in the instances given under A being '003, the
maximum = "005, the minimum = "001. When the sciatic nerve or posterior root of
the cord (severed from the brain) is excited, the rise is seen to be much more
pronounced, being on the average "005, maximum •Oil, minimum "003. Finally,
when the columns of the cord itself are excited, the rise is greater, the average being
"006, maximum '01 3, minimum "002. (The average rise in this case would be still
greater if the three very low and exceptional readings of '002 wei"e omitted from the
table; it would then be "0075.) In the .same experiment performed on the Monkey
(232) the rise in half an hour following twelve excitations was "006.

It would thus appear that one of the mahi features in the rise is the extent to
which the nerve structures in the cord are tlirown into activity ; the result of
cortical excitation is to awaken impulses in a more limited area of the cord than is the
case with excitation of the sciatic afferent nerve fibres, and this latter does not cause
such a widespread awakening as is produced by direct cord stimulation.

From these results alone no conclusions of a definite kind can be drawn as to the
relations subsisting between the cord and the brain on the one hand, and the cord
and the nerves on the other, but if the amount of resting difference may be truly

ON THE MAMMALIAN NERVOUS SYSTEM. 321

taken to be an indication of the amount of potential energy, which the nerve material
is capable of makhig kinetic in the form of a nerve impulse, then this remarkable
rise, and the conditions which determine it, would appear to show that the physio-
logical characters of the structure upon which the storage of energy depends, are such
as to be rendered more efficacious with use, and that the central structures (cells) of
the cord play the most prominent part in this influence of functional use upon
efficiency.

CHAPTER v.— ON THE ELECTRICAL EFFECTS EVOKED IN THE SPINAL CORD AND
MIXED NERVES BY EXCITATION OF THE CORTEX CEREBRI.

The primary object of this work being the determination of the nature of the
impulses which issue from the excitable or so-called motor cortex, considerable
importance attaches to positive results on this point. As, however, usually happens
in the employment of a new method, we have found that more actual advance was
made by applying it to the study of the lower centres and fibres through which these
impulses pass than by attempting to elucidate the functional relations existing
between one part of the cortex and the rest of the encephalon.

One fundamental fact stands out, however, prominently, viz., that, as we shall see
directly, it is possible to ascertain and judge the nature and comparative amount of
the electrical changes which accompany the descent of the cortical impulses in the
spinal cord, and so to learn the character of the cortical discharge. Practically, there-
fore, the results of our special investigation of the excited cortex will best be
arranged according to the part of the nervous system in which the electrical changes
were observed. A further subdivision of such a classification will be necessary in
order to bring out the wider points of interest which have received elucidation by the
use of this method, and this is furnished by the summary and arrangement we have
given on pp. 272-276, of the facts discovered by other methods, of which the most
notable are those of simple inspection and of graphic record respectively.

In accordance with this plan, therefore, we will commence with considering the
case in wliich we observed the electrical changes in the dorsal cord, with the object
of ascei'taining the character of the descending nervous iinpulses in consequence of
stimulation of the cortex.

We have employed in these experiments on the relations of the cortex cerebri to
the bulbo-spinal and peripheral systems fifteen Monkeys and thirty-two Cats.

MDCCCXCI.-B. 2 T

'6-2-2, MESSIiS. F. (lOTCIl AM) V. IIOKSMIY

ExClTATlii.N OV C-DUTIiX.

1. Electrical Changes Ohsei'Ded in the Spinal Cord.

Tlie iirraiigeinent expressed in the above lieadings is the fundamental exj)eriment
designed to elucidate, if possible, the nature of the impulses which pass from the
cortex to gain and traverse the bulbo-spinal centres. The desirability of avoiding
shock, which so markedly lowers the excitability of the cortex, led us to choose to
investigate the electrical effects which accompany the impulses as they pass thi'ough
the mid dorsal region, so that the section of the cord was made in the Cat imme-
diately below this point, and in the Monkey, just above or through the uppermost
part of the lumbar enlargement. (See figs. 4 and 5.)

This selection of the seat of section had the additional advantage that it enabled
us to proceed at once to the determination of further points, viz., the localisation of
the upper and lower limb areas in the cortex, and thus to generally control the results
obtained, as will be shown presently. We will first describe a typical experiment.

The anaesthetised animal having been arranged as described (Chapter III.) the dura
mater was exposed over the so-called motor area of the lower limb ; the spinal cord
was then exposed at the level of about the 7th dorsal vertebra, raised in air and
connected to the non-polarisable electrodes. The character of the resting electrical
difference between the cut end of the cord and the uninjured surfaces of the colunnis
was noted as set forth in Chapter IV.

(a.) Effects seen in the Electrometer.

The electrodes were, in the first instance, connected with the capillary electro-
meter.

The dura mater was then opened, and the surface of the cortex excited {vide p. 299).
The effects observed in the electrometer were —

(1.) A persistent negative variation of the difference lasting as long as the excitation
was applied to the cortex.

(2.) A series of intermittent negative variations commencing (sometimes after a
short interval had elapsed) from the cessation of the excitation, and continuing for a
variable period according to the state of the cortex.

We may with advantage at once briefly review the general questions raised by
this result, as such a discussion will make clearer the object of further researches.

The effect seen in the electrometer was so identical in character with that obtained
from the muscles in an epileptiform convulsion started by similar excitation of the
cortex as to suggest that the essential features of the muscular convulsion are
wholly due to the character of the cortical discharge. This is illustrated by the

ON THE MAMMALIAN NERVOUS SYSTEM.

323

Fig. 4*

Fig. 5.

^^njTfCWivJ^

t- -.

ESifi5L^^f

f"

tJ

^

V^i-

lit

.

Cat

Adult 0

- ni.D

^Y^'^'^'^T)-bl50<)<)-J

-

'-<^'^'^<r(nr(rsx^^'^

'

KD

— —

xin.D

i

iv:l

1

VE>

''4

\

1'

^}

K£^iUi^^H

^iS^',-^ i,c

L^ID

Macacus

Rhesus ' \

AdiiU %

-

.—■ «r<7Cn

.fS'SWswa-J'-

5>^^»^

1 -•. AMD

r

1

* The central nervous system displayed in this and succeeding figures is a photographic reduction of
a full size photograph of an actual dissection, as shown also in Plates 30-.33. By this means, exact topo-
graphy of all experimental details (i.e position of electrodes, &e.) has been preserved. It is, perhaps,
hardly necessary to add that the amount of cord exposed is shewn in Plato 29.

2 T 2

324

MESSRS. F. (lOTCII AND V. HORSl.KY

annexed woodcut (fig. G), in wlue'Ii two records, one of tlie muscle, the other of the
movement of the electrometor, are contrasted. The lower tracing, G, is that of the
muscular movements of the rectus femoris of the Cat ; the upper, E, is that of the
movements of the projected image of the mercurial meniscus of the electrometer
photographed on a sensitive plate. Tlie juxtaposition of the records was efl'ected by
a photographic reduction of the two tracings. It is necessary to point out that exact
comparison is not possible, since the two records were obtained from different animals ;
they are, however, comparable in this sense, that since we have selected in each case
fits of about the same degree of intensity, the general fused character of the first
stage, and the interrupted character of the second stage, is similar in each.

Further, the change in character of the interrupted stage itself as the fit draws to
its close is very noticeable, the great increase in the extent of each clonic spasm
accompanying the slower rhythm is, as will be seen, a feature common to both records.

It will now be advisable to consider in more detail the character of these electrical
variations.

A reference to the jjhotograph of the movements of the electrometer, in fig. 6,
will show that the first stage of the effect produced in the cord by exciting the
cortex, i.e., the tonic variation lasting during the period of excitation, is a complete
fusion of impulses, so that it is with our present instruments impossible to detect
any waves or intermissions in it. In this respect the tonic stage exhibits no

Fi.'. 6.

c ctcitatton '

E. Photograph of movements of electrometei'.
G. Tracing of contracting muscle.
t. Tracing of time signal with intervals of i^th sec.

difference from the tonus obtained by exciting the corona radiata after the cortex
is removed. It is otherwise with the graphic method, see p. 273, in which a
rhythm or period can be seen, showing the tonus to be an imperfect fusion of
contractions. Until the question is re-examined with still more delicate instru-
ments, it must be left entirely open, as the absence of intermittence in the
electrical record may be due to instrumental inadequacy. We cannot, however,
refrain from suggesting that it may be possibly a genuine phenomenon, and that
the waves seen often on the tetanus curve, in the graphic method, may be in
relation with blocks in the motor path ; one such block we have found to be offered

ON THE MAMMALIAN NERVOUS SYSTEM. 325

by the bulbo-spinal centres to the passage of the impulses, fi'om the spinal cord out
into the nerve (see p. 458, Chapter X.). The further recent experiments of Wedenskii*
seem to suggest that these waves in the muscular record may be to some extent
in relation with the block offered by the motor endings of the nerves in the muscles,
since he finds that such muscular intermittence may be produced by stimulation
of the nerve itself with rapid rates, an intermittence which, not being synchronous
with the rate of stimulation, but having its own rhythm, is presumably referable
to the neuro-muscular mechanism.

As regards the after effect, the clonic stage, it is frequently separated from the
tonic stage by a distinct pause or interval ; this is of course visible in all methods of
exploration, to which the electrical investigation of tlie spinal cord is no exception.
When it occurs, therefore, it is obvious that after the cortex has been roused to
discharge, thus evoking the tonic stage, it possesses the power of again emitting
impulses when thrown into a high state of excitation. The period at which this
clonic discharge commences is variable, there being in some instances a pause or
period of delay, whilst in others the intermittent or clonic stage may even ovei4ap
the tonic condition, which, consequently, ceases prematurely, while yet the excitation
is being continued.

The clonic stage, or after effect, has a distinct rhythm (see graphic method, p. 273),
in which the muscles respond to intei'mittent impulses. Exactly the same is seen in
the electrical changes of the spinal cord. This stage, or after effect, as it truly is, ■
presents three chief features for study : — (l) Its commencement ; (2) its development ;
(3) its mode of termination ; these being common to both the muscular and the spinal
records.

(1.) The clonic stage invariably begins with small, often single discharges, vide fig. 6,
these soon increase in strength. This is of course closely imitated by the graphic
recoi'ds of the muscle, the contractions of which, minimal at first, become, later,
maximal.

(2.) The further development of the stage is marked by summation of the impulses,
the rhythm of which, directly measured, appears to be about 10 per sec, but we
have not yet had time to thoroughly investigate this point. ( Vide graphic method,
p. 273.)

The summation of the electrical variations very obviously harmonises with all the
other facts relatmg to the persistence of the muscular contraction, when produced
by cortical stimulation. (See also Francois Franck and Bubnoff and Heidenhain,
loc. cit.)

(3.) The termination of the series of intermittent variations is marked by their
becoming fewer, lai'ger, and finally ceasing abruptly {vide fig. 6), vei-y rarely
diminishing to final disappearance.

* Wedenskh, ' Ai-chives de Physiologic Normale et Pathol.' (BEOWN-SliQUARD), Jan., 1891.

326 MRSSRS. F. fiOTCH AND V, IIOUSLEY

The absolute vakics obtained in the electrometer varied in the two stages, as
estimated by the eye, and the rough estimate thus formed is supported by the better
evidence of the photographs. They average 2 divisions, rarely 3 divisions for the
tonic stage, and 1 "5 divisions for the clonic stage.

It would appear that this is true for the Carnivorous animal, as represented by the
Cat (eleven animals observed). Whether these electrometer figures are the same for
the Monkey we are unable to say, as we devoted our attention in the experiments on
that higher Mammal (seven animals observed) to photographing the results.

The absolute electromotive values obtained by the electrometer and recorded photo-
graphically, are being examined by Mr. G. F. Burch at the present time, and we,
therefore, will postpone the further consideration of this most important point.

(/;.) Effects seen in the Galvanometer.

The total effect produced in the galvanometer when it is connected with the spinal
cord, and the cortex is excited, i-esembles the records of the muscular contractions, in
that there are two distinct stages. It need, however, hardly be said that the slow
swing of the galvanometer needle is incapable of recording intermittently, and con-
sequently both stages are composed of a series of summated effects. The rate of
movement of the needle is notably different in the two stages, and the close observa-
tion of this feature proved of much value in other experiments, notably those on the
corona radiata, see p. 337.

In the first stage the needle generally begins to move soon after the commence-
ment of the cortical excitation (see Method, p. 299), but often owing to one of the
depressing circumstances mentioned on p. 272, the cortex is not normally excitable and
the effect it produces when stimulated is consequently delayed. The needle swings
steadily during the excitation, but when this ceases there is a distinct check,* and
then as the after-effect develops the needle slowly swings on and gradually comes to
a standstill at the end of the second stage or after effect. The mode of termination
of this last is slow and deliberate in its gradual diminution, thus contrasting markedly
with the abrupt cessation of the galvanometer excursion, when the corona radiata or
spinal cord is excited. This gradual dying out of the effect is so absolutely
characteristic of the cortex that it can be used for differentiation.

Though well aware of the small value to be attached to the quantitative use of the
galvanometric readings thus obtained, we venture to add a few remarks on these in
view of the novelty of the point.

When the central end of the whole cord is connected with the galvanometer, and

* What is said on p. 299, &c., relating to possilsle errors and fallacies may be here remembered as
showing that this check is purely a physiological phenomenon, there being no possibility of the excitation
current affecting the delicate galvanometer.

ON THE MAMMALIAN NERVOUS SYSTEM.

327

one cortex {i.e., lower limb focus) excited for a definite period, an electrical excitatory
variation is produced in the coi'd, the amount of which is indicated by the following
average of several observations.

Cat

Monkey

Duration of
excitation.

Average strength
of excitation.

Average
variation.

5 seconds
3-5 „

8500
4500

193°

175°

The brain of the Monkey was stimulated for a shorter time, and with far weaker
excitation than that used for the Cat, in consequence of our desire not to make the
cortex hyperexcitable. Even then the excitatory effect in tlie Monkey is nearly as
large as that in the Cat. The truth of this position is also evidenced by numerous
observations we have made under other circumstances, the Cat always requiring a
stimulus, which in the Monkey would have evoked a far higher excitatory effect. As
with all results obtained by the method of averages, no doubt this point would repay
accumulation of observations, since, although no absolute value can be given to the
figures, they seem to show specific differences. With the cord divided longitudinally,
and the excitatory effect observed in each half, this absolute difference apparently
disappears (see p. 354).

Before proceeding further, it is essential that we should indicate the experimental
evidence on which we base our interpretation of the electrical changes in the cord.

That the changes are due to the discharge of excitatory impulses from the cortex is
clear, but it may be asked what evidence apart from the parallelism discussed on
p. 324 there is to show that they are the concomitants of the passage of actual nerve
impulses through the portion of exposed cord, and not the x'esults of conduction from
the electrical changes evoked in all the surrounding muscles which are thrown into
spasm.

The evidence which shows that the change is one localised in the nerve fibres of the
exposed cord may be grouped as follows : —

(1.) The position of the electrodes being arranged to ensure isolation, as described
in Chapter III., p. 295, this isolation is sufficiently complete to allow of the surrounding
muscles being thrown into activity without any electrical change in the exj^osed
tract being evident in the galvanometer.

(2.) In observations made in the Monkey with the freshly exposed cortex, the
initial and immediately succeeding stimulations of the lower limb area evoked effects
which, with the strength of the stimulus employed, were so exactly circumscribed
that, since the cord was divided and the lower limbs thus cut oft" no demonstrable

328 MESSRS. F. GOTCll AND V. IIUUSLKY

muscular iiiuvcmr-uI oI' ;iny kind existed, yet a UKukcd and distinct electrical change
of the usual kind was present in the cord and evidenced both in the electrometer and
the galvanometer.

When, further, owing to prolonged excitation, exposure, or slighter degree of
etherisation, hyperexcitability of the cortex became established, subsequent excitation
evoked a more general discharge, and although the muscles of the trunk and upper
limb were now thrown into active contraction, the change evidenced by the electro-
meter and galvanometer remained similar in character to that previously observed.

(;i.) If in any given instance with the central end of the spinal cord in llio lower
dorsal region exposed and connected with the galvanometer electrodes, the cortical
area for the upper limb was excited, no effect was observed in the galvanometer and
electrometer, although the muscles of the upper limb were thrown into convulsion.

(4.) The injury (cutting off, &c.) of the cord between the exposed region and its
deep connections abolishes the effect in the cord, even although the muscles are
thrown by cortical excitation into violent spasm.

(5 ) In addition to the above, the whole mass of results to be detailed in this and
the succeeding chapters is convincing evidence of the truth of our interpretation,
since, as will be seen by separating the fibres in the two sides of the cord (longitudinal
section), by interrupting the fibres on one side only (hemisection), &c., we obtained a
differentiation in our results which admits of no other explanation than the one
upon which we base our deductions, namely, that the electrical effects observed are
due to changes localised in the exposed cord, and, moreover, in the experiments
under discussion localised particularly in the nerve fibres which form the pyramidal
tracts.

The fact above alluded to of the absence of electrical efiects in the dorsal cord when
the area for the upper limb was excited in the Monkey, involves matters of such
importance that we now pass on to consider it in some detail.

(('.) The Localisation of Efferent Rejiresentation in the Cortex as ascertained hy the
Electrical Method.

It was clear from the commencement of this research, that our method afforded
means of differentiating the centi'es in the cortex and their correlated fibres in the
spinal cord. We accordingly made observations of the following character : —

Having exposed the cortex freely, and connected, as before, the cut dorsal cord with
the electrometer, we proceeded to obtain the usual result by exciting the lower
limb area with a minimal but adequate stimiUus, evoking thus an eftect without at
the same time developing an epileptic hyperexcitability of the cortex. We then
explored the rest of the surface of the so-called motor region with the electrodes and
the same strength of excitation, to see how far we could determine whether there was

ON THP] MAMMALIAN NERVOUS SYSTEM.

329

any electrical change in the cord, i.e., the pyramidal tract, really belonging to the
lower limb, when other parts of the cortex were excited. As will be now shown, our
instrument gave no indications of such difiused effects, but in this relation we must
draw attention to the animals used (Cat and Monkey).

Considering that the minute differentiation of efferent motor function in the cortex
of the Cat is relatively insignificant compared to that in the Monkey, we employed
the former animal for the preliminary investigation of the accuracy of the general
position which is involved in our first statements. But we also found in the Cat that,
while with a criven streng-th of excitation, stimulation of the lower limb cortical area
gave the definite result of a movement of two divisions in the electrometer (the con-
nection of the cord, &c., being as stated before), the excitation of the occipital lobe
and the temporo-sphenoidal lobe respectively gave no result at all. Upon this point
it may be remarked that in some instances we were able in the Cat even to differen-
tiate, as we easily could do in the Monkey, between the upper limb and the lower
limb areas. That is to say, that if the dorsal cord were observed, we got a well-
marked effect by exciting the lower limb focus (as indicated by Ferrier), whereas
excitation of the fore limb focus produced no visible change in the cord's state, as
evidenced by the electrometer.

Thus the following facts were noted in the Cat (126).

Cord led off' at Lower Dorsal Reffiou.

Regions of cortex excited.

Strength of
excitation.

Effect in Electrometer.

Tonic stage.

Clonic stage.

(a) Lower limb focus ....
(t) Occipital lobe

(c) Temporo-sphenoidal lobe .

(d) Fore limb focus ....

10,000
10,000
10.000
10,000

Rise 2 divisions

Nil
Nil
Nil

1"5 divisions
Nil

Nil
Nil

For the points a, b, c, d excited, see Plate 30.

Turning now to the Monkey, we carried this mode of investigation still further by
the employment of minimal stimuli. The following general result was obtained. The
dorsal cord being observed, the greatest effect was produced when the excitation was
applied to the centre of the lower limb area, i.e., just at the hallux focus,'"' and this
effect diminished as the electrodes were removed from that point towards the mesial
surface of the hemisphere or downwards over the convex surface towards the superior
frontal sulcus. (See Plate 32.) It was interesting to observe that opposite this sulcus.

MDCCCXCI. — E.

* Bekvor and Horslev, ' Phil. Trans.,' 1890, &c.
2 V

330 MKSSKS. F. (lOTCIl AND V. HORSLKY

wliere, according to Bkkvoii and one of us, tlie upper and lower limb areas tend to
overlap, with the same stimulus as in the latter case this region was the last from which
any effect could be produced in the fibres of the dorsal cord. Even in the case where
the whole cortex was very excitable, and where a generalised effect was thus easily
evoked, iind where, consequently, electrical effects in the cord followed stimulation of
the upper third of the upper limb area, the general effect was, nevertheless, most
markedly graduated from the centre of the lower limb area in diminishing order as
the electrodes were removed downwards and passed tlie level of the superior frontal
sulcus.

We may now allude to the obvious inference that might l)e drawn from these
experiments, namely, that although there is no sharply marked line of demarcation
between the cortical foci, there is nevertheless in the facts we have just stated a
certain amount of evidence against the assumption that the lower limb, for example, is
represented to any marked degree in the upper limb area of the cortex. We do not
speak with great positiveness on this point, because it may well be that the instru-
ment we chiefly used in this branch of our work, the electrometer, was not sufficiently
delicate to show the extremely minute variation which might be supposed to result
from the excitation of only a very few fibres. That this indeed seems probable is
sliown by the results of two experiments which we performed with the galvanometer
instead of the electrometer. Unfortunately, however, they do not enable us to speak
with greater certainty in the directions indicated, because in each instance the cortex
was hyperexcitable from the commencement of the experiment. This, however, is
clearly a branch of enquiry which might with advantage be followed out by subsequent
observers.

In this connection the experiments of Shereington"'^ are very suggestive, and we
shall refer again to this part of the subject in desci'ibing our experiments on bilateral
representation.

((/.) Tlie Amount of the Gahxmometric Readings.

Although, as has been stated, no exact comparison for quantitative purposes can
be made between different deflections of the galvanometer, when, being connected
with the central end of the exposed spinal cord, the cortex is excited, yet the
amounts of the different readings are in themselves of great interest, owing to the fact
that it is always possible with an excitable cortex and a sufficiently strong excitation
to obtain very large galvanometric deflections. It is obvious that the deflection is
dependent not merely upon the intensity but the duration of the stimulation, hence
the excitation was limited by the I'evolving key, previously described, to 5 seconds.
Since, however, the cortical discharge lasts a variable time after the stimulus has ceased,

* ' Proceedings of the Physiological Society,' 1890.

ON THE MAMMALIAN NERVOUS SYSTEM.

331

the galvanometric effects display, os might be expected, great inequahties ; these,
however, will be seen to be in strict relation with the amount of visible contraction
evidenced in the muscles of the trunk, &c., and hence are really determined by the
force and duration of the cortical discharge. Bearing this in mind, the following
table is very instructive, as indicating the size of the deflections, which vary, as is
seen, from 63 to 510 scale. The intensity of the stimulus, the particular cortex
excited, and the character of the fit evoked, are in each case noted.

It will be seen that the average of these readings, which Avere all taken in four
animals (Cat), as far as possible under similar conditions of etherisation, &c., amounts
to 193 scale. It will be also seen that it is always easy to produce by a sufficient
intensity of stimulation readings which are in excess of this.

Galvanometbic Effects produced in Spinal Cord of Cat by Cortical Excitation.

Intensity of

Galvanometiic

Muscular

Cat (315) . . .

Excitation.

5'' left cortex

Eifect.

Movement.

8,000

63

Slight fit

8,000

„ right „

115

)» J)

9,000

„ left „

108

)» fi

9,000

„ right „

230

Good fit

„ (317) . . .

8,000

,. left

50

Very slight fit

10,000

)1 i) J)

162

Fair fit

12,000

J) H )1

235

Good fit

8,000

„ i-iglit „

1.30

Slig-ht fit

10,000

?i n J)

185

Fair fit

12,000

•f i» ))

230

Good fit

„ (319) . . .

10,000

„ left „

90

Slight fit

12.000

ij »i i)

170

Fair fit

10,000

,. 1-iglit „

150

)» ))

12,000

?) IJ V

260

Good fit

., (324) . . .

6,000

„ left „

260

Slight fit

8,000

>j )) 5'

340

Fair fit

10,000

Average. . . .

610

Powerful fit

] 7 = 3288

193

The interest of the results will increase when we compare the average amount with
the highest obtainable value of the galvanometric reading afforded by the effect of
cortical excitation in the sciatic nerve. It must be now pointed out that in all cases, it is
essential to reject observations in which there are sudden irregular changes in the resting
electrical difference of the cord, since these are in great measure due to slight failure
of the anaesthesia and to the consequent semi-voluntary discharge of cortical impulses.

2 u £

332 MKSSKS. i'-. (lO'irii and \'. iiitiisi,i;v

(<■.) Electrical Change i» ConJ praijitced hi/ " Scmi-rnfiniinri/" f'nrtirnJ Discharges.

During prolonged etherisation it is wrll kmiuii dial animals nfall kinds, including
Human beings, occasionally make unconscious, but serai-voluntary, purposive move-
ments. These movements are abolished by very profound anfesthesia.

Electrical ett'ects manifest themselves in the observed portion of exposed spinal
cord when these semi-voluntary movements occur in the upper limb and trunk muscles.
The effects are evidenced in the galvanometer connected with the tissue by deflec-
tions, which, though resembling the excitatory variations in being always opposed in
sign to that of the resting difference, vary very much in amount and rapidity of
development. It is obvious that the excitability of the cortex and the intensity of
the exciting agency, must be the chief agents in determining the amount of this
effect. The extent of such a deflection when, as evidenced by well-marked move-
ments of the muscles on both sides in the upper half of the body, a considerable
cortical discharge of energy was taking place, has amounted in the Monkey to 360
scale ; it was however, at once abolished by profound ansesthetisation. As a rule the
deflections amounted to from 20 to 60 scale ; no doubt they would have been larger,
had it not been for the fact that, since their appearance interfered with accurate
observation of cortical effects evoked by stimulation, we always took care on perceiving
them to so alter our conditions of anaesthesia as to abolish this source of error.

2. Electrical Changes in the Sciatic Nerve.

After having thus considered the electrical changes produced in the spinal cord by
the functional activity of the cortex, we next turned our attention to the more
complex question of what occurs in the peripheral nerves when the cortex is excited.
This is obviously more complex, because, although we are still exciting the cortex, w^e
have now introduced into the mechanism the bulbo-spinal system of centres, so that
we ai-e brought at once face to face with the question as to how far the changes which
we have just been studying in the spinal cord are modified by the as yet hypothetical
termination of the fibres of the pyramidal tract in the said bulbo-spinal centres.

We are justified in expressing this query in a still simpler fashion : what becomes
of the cortical impulses when they reach the bulbo-spinal centres? Other things
being equal, it might have been expected that we should have got very clear evidence
of the effect of active excitatory change in the peripheral nerve, seeing that if this
had been still in connection Avith the muscle, the muscle would have been thrown into
a powerful series of contractions. It was, therefore, with very considerable surprise
that we observed what a relatively small electrical variation was obtainable in the
sciatic nerve when the corresponding portion of the cortex was excited. The experi-
mental procedm-e was extremely simple, and as follows : — The cortex was exposed, as
described, and the sciatic nerve on the opposite side of the body was also exposed,

ON THE MAMMALIAN- NERVOUS SYSTEM.

333

ligatured, divided below the ligature, and its central end connected in the manner
indicated in Chapter III. with the galvanometer or electrometer electrodes. (See
fig. 7.)

We found that excitation of the cortex in most cases produced no perceptible
effects when the capillary electrometer was in connection with the nerve, and when
occasionally effects were seen, their complete absence in other instances, and their
rapid disappearance on repetition, threw doubt on the observations.

With the galvanometer, however, effects could always be seen, though these
were very small in amount as compared with those observed in the spinal cord. In a
few cases, however, in which the excitation applied to the cortex was very intense, so

334

MESSRS. F. GOTCH AND V. IKIHSLEY

US to arouse very violent and prolon^'cd bilateral lit.s, tiie (loUection.s observed
approached in size the smallest of those obtained in the cord. The general results
■will be seen by the following table, in which the effects observed in (he nerve on the
opposite (i.e., corresponding) side to the excited cortex of eleven Cats and two Monkeys
are separated from those observed in the nerve on the same side. The table is so
arranged that the weak fits are given first and the stronger prolonged ones last.

It will be seen that the amount of defiection varies from a mere trace to 80, the
larger readings being always obtained in the case of powerful prolonged and bilateral
fits. The average of all the readings in the nerve opposite to the excited cortex is 31,
whilst under no circumstances, even with the most powerful fits, is an effect obtained
equal in amount to the change evoked in the spinal cord.

Intensity
of

Duration
of

Effect in
Nerve

Effect in

Nerve on

same Side as

Corte.x.

Cbaractur of Fit.

Excitation.

Excitation.

Opposite
to Cortex.

Sees.

Cat (291) ....

12,500

5

2

Weak

„ (293) ....

10,000

)5

7
6
5

11

„ (296) ....

11,000

n

'.I

}9

1>

9

, ,

T)

12

Slight

20

. ,

Fair

„ (299) ....

12,000

))

12

20
10

10

Weak
Fair

„ (290) ....

*>

)»

18
10
22

2

11

„ (71) ....

(3,000

JJ

20

'1

Monkey (54) . . .

5,000

■^

16

18

5

1

Good

„ (-217). . .

u

•>

28
18
16

18
5

Moderate

Cat (298) ....

12.500

1'

28

Good

„ (75) ....

))

»)

35
28

)i

10,000

"

35
55

A^iolent
Prolonged

„ (292) ....

12,500

)»

62
46

80

(Prolonged povrerful

r fits

„ (79) ....

»»

.,

H

1 Prolonged powerful
/ fit.s

11

„ (301) ....

12,000

■»

70

80

Prolonged powerful
fits (morphia)

Average . .

28=868

31

ON THE MAMMALIAN NERVOUS SYSTKM. 335

If we select out of the whole table tliose readings which were obtauied with tits of
a unilateral character, thus disregai'ding the whole of the larger numbers, we find
that the average amounts to 20, this being very much below the average unilateral
effect in the cord. (See Chapter VII.)

It might be imagined that this difference in amount between the nerve and cord
effect was entirely connected with the difference in sectional area and thus in resist-
ance ; that this is not the case is shown by the following considerations : —

(a.) The total resistance in connection with the preparation is made up of the
electrodes with their cables and the tract of tissues, of these the resistance of the
cables is far in excess of that of the tissue, and hence any difference in resistance
between the spinal cord and nerve preparations due to the sectional area of the tissue
is but a small fraction of the wdiole.

{h.) The absence of effect in the capillary electrometer, when connected with the
nerve, shows that any unusual resistance which may be offered by the nerve as com-
pared with the cord is not the cause of the diminution, since the amount of the
electrometer movement is unaffected by changes of resistance.

(c.) The sectional area of the nerve of an adult large Cat is nearly as great as that
of the dorsal coi'd of a young animal or a small Monkey, yet the effect is small in the
former and large in the latter structure.

{d.) As will be seen in Chapter VII., on bilaterality of representation, it is possible
to obtain large effects in only one-half of the cord, split longitudinally. This portion
has apparently less sectional area than the sciatic nerve, but the electrical change in
it, as will be seen by reference to the tables, is far in excess of the nerve change.

(e.) The change in the cord is undoubtedly one connected with the fibres of the
pyramidal tracts, it is therefore with reference to their sectional area as compared
with the area of the motor nerves in the sciatic nerve that any criticism on this head
should be directed. There is, however, every reason for supposing that the pyramidal
tracts m the dorsal region are smaller in cross section than the sum of the anterior
roots of the lumbar plexus.

These considerations serve to emphasize the conclusion to which, as it appears to us,
the foregoing results tend, viz., that the extraordinary difference in quantity between
the electrical effect in the cord and in the sciatic nerve when the cortex is excited,
must be attributed to an alteration m the quality and quantity of the nerve impulses
in their passage from the cord into the nerve, and that the structure of the spina,
centres, through which the impulses must necessarily pass to reach the issuing nerves,
so influences the transmission as to cause this striking change. This view is
strengthened by the remarkable confirmation which is given to it by experiments in
which both the spinal cord is directly excited and the centres in it discharged reflexly
by stimulating its posterior roots. It will be seen (Chapter X., p. 478, and Chapter
XL, p. 494) that under these circumstances the electrical effects produced in the sciatic
nerve are extremely small and resemble in amount those above referred to.

33G MESSUS. [>'. (lO'l'C'll AND V. HOHSLEY

Tlie lar"-est ellccts we liiuc ovt-r sefii in (.he nti'vi- in consequence of cortical
excitation liave been obtained \>y the use of absinllic!. 'i'he powerful general fits
which followed tlic inli.Hliution into the Ijlood of this drug being evidenced in two
animals (Cat) by eftects which have amounted to 272 scale. It need hardly be
pointed out that such fits are of extremely prolonged character, and that in the spinal
cord electrical changes occur which give deflections in very many cases too large to be
read (off screen).

Cortical Localisation as Evidenced by Elcctncal Changes in the Nerve.

The electrical changes in the nerve, as evidenced by the galvanometer, although
small, are quite definite, and aftbrd additional proof of the localisation of cerebral
representation. Thus, when an adequate stimulus was employed, which was,
however, only intense enough to evoke unilateral muscular contractions, the electrical
change in the nerve was entirely confined to that on the opposite side to the excited
cortex. This subject will be discussed in detail in Chapter VII. , which deals with
bllaterality of representation. It may, however, be pointed out that no other operative
interference than that already described as necessary for cortex and nerve exposure
having been performed, it is perfectly evident that if the stimulus was maximal, as it
usually was in this experiment, a bilateral fit could and did happen, so that from one
nerve an effect might be and was obtained due to discharge of both hemispheres.

It is, moreover, interesting from the point of view both of control and of localisa-
tion, that in cases where a strength of stimulus which did not evoke a general
discharge was employed, whilst the application of the stimulus to the so-called lower
limb area of the Cat produced definite electrical changes in the nerve, its application
to the focus of representation of the upper limb evoked little or no change, even
though the stimulus was follow^ed by marked epileptic convulsions in the upper
limbs. The localisation of the foci of representation of the movements of upper and
lower limbs in the Carnivora, which has been determined by other observers, is thus
corroborated by the use of the present method.

CHAPTER VI. -ON THE ELECTRICAL EFFECTS EVOKED IN THE SPINAL CORD AND
MIXED NERVES BY EXCITATION OF THE CORONA RADIATA.

Excitation of Corona Radiata.

1. Electrical Changes in (he Spinal Cord.

We will consider first tlie general character of the changes as evidenced by the
movement of the capillary electrometer.

ox THE MAMMALIAN NERVOUS SYSTEM. 337

(a.) Effects seen in Electrometer.

The coi'tex and spinal cord having been exposed, the central end of the cord
divided in the lower dorsal region was connected with the instrument in the manner
already desci-ibed, and the corona radiata displayed by the method detailed in
Chapter III., Section 2.

We have already stated that it was early shown by Francois Franck and others
that removal of the cortex and excitation of the corona radiata no longer gave the
characteristic sequence of tonic followed by clonic muscular contraction, but simply a
tonic contraction synchronous with the duration of the excitation, the first stage, as it
were, of the cortical effect. In our earliest experiments (see ' Roy. Soc. Proc.'), we
found that a similar difference occurred in the character of the electrical changes
observed in the spinal cord, as evidenced by the cajiillary electrometer. The chax'acter
of the change in the two conditions is well shown in the facsimile I'epresentations of
the photographic records which are given in our former paper, the corona radiata
effect being simply a persistent tonic negative variation, the amount of movement of
the meniscus being about one division of the eye-piece scale.

The small size of the electrometer change, and the consequent difficulty of appre-
ciating its comparative quantitative value under different circumstances, led us to
employ the galvanometer as an index of its amount.

(6.) Effects seen in Galvanometer. ■

When the effect in the cord is observed with this instrument it is seen to be unlike
that evoked by cortical excitation in this respect, that the deflection of the needle
commences with the application of the stimulus, and ends sharply with its termina-
tion. The amount of the deflection is thus very definite, but of less extent than
that obtained when the cortex is excited.

In the following table a series of observations made on three Cats are given, in
which precautions were taken to insure that the state of anajsthesia, &c., should be in
all cases as far as possible the same. The deflections vary from 37 to 175, the
average being 102.

MDCCCXCI. — B. 2 X

838

MESSRS. F. GOTCH AND V. IIORSLEY

Effect in Spinal Cord.

Intensity

Dnration

Effect

of

of

in spinal

Character of muscular effect.

excitation.

excitation.

cord.

Sees.

Cat (99) . .

12,500

5

42
37
50

Tonic contraction

„ (288). .

10,000

5

118
58

120
62

Well marked tonic contraction

)» n ))

11,000

5

130

110

71

170

Well marked tonic contraction

„ (309). .

8,000

5

125
175

W ell marked tonic contraction

8,000

5

135

)j n n

Average . .

129

)» »» n

15=1532

102

We would pass from this part of onr subject, since our knowledge of the process
of secondary degeneration has shown that in the arrangement of the experiment all
we do is to observe at one (the spinal) end of a column of fibres the changes evoked
by stimulating these fibres at the other (corona radiata) end, wex'e it not that the
above efiiect is capable of modification in a manner which demands consideration.
Moreover, we wish to emphasise the fact, that in its simplest form the corona radiata
effect, as observed in the cord, is an exanjple of the application of the galvanometric
method to the determination of directly continuous nerve tracts in the central
nervous system. This, as we shall see later in dealing with the spinal cord itself, is
one of the most valuable uses to which the method can be put. Thus when a definite
tonic effect, and that only, is evoked by the excitation, the application of the stimulus
to any fibres of the corona radiata, except those immediately underlying the focus of
representation of the movement of the lower limb, produces no electrical changes in
the lower dorsal cord, even though the muscles of other parts are sent into tonic
contraction. The modifications we have referred to above, are those which have been
observed by most workers with the graphic method, and consist in the presence of an
after-effect in no wise differing from the clonic stage of the full cortical discharge save
in duration and completeness. The assumption is natui-ally that in the arrangement
of the experiment, it is impossible to avoid exciting association fibres (fibr« arcuatse,

ON THE MAMMALIAN NERVOUS SYSTEM. 339

&c.), which arouse the neighbouring cortex cerebri, and cause a discharge from cor-
puscles in which the peripheral part under investigation is yet represented.

We have little doubt but that this is the true explanation, and believe that
excitation really limited to the fibres of the corona radiata is never followed by such
an after-eflfect. This will, however, be again referred to in Chapter VII.

2. Electrical Changes in the Sciatic Nerve.

As in the result of cortical stimulation, so in that of excitation of the corona radiata,
there 'is a striking ditference between the effect in the cord and that in the sciatic
nerve. We have never yet observed any effect from the nerve when the electrometer
was employed ; in the galvanometer, however, such effects are generally seen, although
the necessary stimulus has to be eo strong as to run the risk of evoking discharge
from the uninjured portions of cortex. Whenever such epileptic discharges obviously
occurred, as evidenced by* the muscular movements, the observation was regarded as
worthless.

The following table gives the results observed in eight Cats, all those being
excluded in which, owing to the degree of anaesthesia being less, the excitation
brought about general muscular movements.

As in the case of cortical excitation, so here, the deflections observed in the (corre-
sponding) nerve on the opposite side to that of the excitation are placed in a separate
table to those observed in the nerve on the same side, the latter being given so as
to show when bilateral effects were produced.

It will be seen that tlie amounts vary between 4 and 68, and that the average of
all readings is 26.

2x2

340

MESSRS. l'\ CiOTCH AND V. IIOUSLKY
Electkical Efiects in tlie Nerve.

Intensity of

I
Diii'iitiou of

I'JUect in cor-
responding

Effect in

M iiscular clloct

excitation.

excitation.

nerve on
same side.

observed.

nerve.

Scca.

Cat (290) . . .

12,000

6

7

0

Tonus only

9

0

H J)

„ (291) . . .

12,500

5

5

, ,

Weak tonus

„ (292) . . .

9,000

5

20

18

12

Good ,,

10,000

5

20

. ,

ri V

6

2

u )T

„ (297) , . .

12,500

5

12

0

Good

„ (298) . . .

12,000

5

4

0 •

Weak „

12,700

5

11

0

Good „

„ (299) . . .

12,000

5

46

21

Powerful bilateral tonus

61

20

» )» )»

10 '

0

Weak tonus

„ (303) . . .

10,000

5

30

12

Bilateral tonus

28

14

») n

22

8

a ^i

„ (305) . . .

8,000

5

52

14

Powerful „

63

33

Bilateral ,,

23

5

Feeble „

10,000

5

59

25

Good

10

1

Very feeble tonus

31

. 13

Fair tonus

68

42

Bilateral good tonus

23 = 615

26

Special attention must be drawn to the fact tliat higher numbers than those given
were sometimes obtained ; however, these large deflections accompanied general
movements of the animal, due, in part, to the fact that it is extremely difficult to
maintain a j^roper degree of anfesthesla. If this is too slight, the stimulus causes a
general awakening of the whole cerebral system.

The question of bilatei'ality will be referred to in the next chapter. It remains to
institute a general comparison between the galvanometric effects observed in both
cord and nerve when the cortex and corona radiata are respectively excited. Such a
comparison will lead to a better understanding of the part played by the bulbo-spinal
centres, though any complete inquiry into this difficult question must be postponed to
Chapter XI., which treats directly of the matter.

If we arrange the parts of the nei'vous system here dealt with in their anatomical
order, they form the following series : —

1. Cortex.

2. Corona radiata.

ON THE MAMMALIAN NERVOUS SYSTEM.

341

3. Spinal cord (dorsal region).

4. Bulbo-spinal centres.

5. Nerve.

6. Muscular nerve-endings.

The combinations of these parts involved in our experiments may be grouped as
follows, and to each group is affixed the highest and lowest galvanonietric reading
obtained in the case of the Cat : — ■

Part stimulated.

Part observed.

Galvanometric effects.

Highest.

Lowest.

Average.

1-3. Cortex . . .
1-5. „ ...
2-3. Corona radiata .
Z—o. „ „

Dorsal cord . .
Sciatic nerve
Dorsal cord . .
Sciatic nerve

510
80

175
68

50

2

37

4

193

29
102

26

From these figures, which apply to the Cat only, it is very evident that the specific
element distinguishing the cortex from the white fibres of the corona radiata leading
away from it is the energetic discharge of its own structure, the corpuscles. Reverting
now to the question raised at the beginning of this section, it is plain that the
average amount of the effect j^roduced by the corona radiata forms practically half, viz.,
102, that, viz., 193, evoked by the cortex, the remainder naturally corresponding with
the clonic after-effect. The levelling process effected by the block in the connection
between the pyramidal fibres and the nerves through the corpuscles of the bulbo-
spinal centres is so severe as apparently to reduce the increase due to the after-effect
to too low an intensity to make itself felt in the galvanometer in any marked degi'ee,
for it will be seen in the above table that the figures for the sciatic nerve are
practically equal.

Beyond multiplication of experiments to check and control these observations, we
did not see that further investigation in this direction would be so profitable as
examining the relations and action of the bulbo-spinal centres when separated from
the cervical region of the spinal cord (see Chapters VIII., IX., X., XL), since the
phenomena of conduction in fibres are far more easily studied than those in which
corpuscular mechanisms are involved.

342 MESSRS. F. GOTCH AND W llORSLEY

CHAPTER VII.— ON BILA.TERAL1TY OF REPRESENTATION TN THE CEREBRUM, AS
EVIDENCED BY THE ELECTRICAL CHANGES IN Till'; SPINAL CORD AND
MIXED NERVE.

The galvanometric method affords an excellent means of determining to what
extent any bilateral movements obtained in the lower limbs on exciting the cortex are
associated with impulses proceeding down both sides of the cord.

It is obvious that a most important question is involved in this inquiry, that,
namely, of the share taken in the production of bilateral effects by the different parts
of the nervous system. When bilateral movements of the lower limbs are evoked by
excitation of one cortex, it is conceivable (a) that in the portion of cortex excited,
muscular movements on both sides are represented ; (h) that the excitation has
aroused the corresponding cortical areas in the opposite hemisphere, and thus
produced the bilateral effect ; (c) that the excitation has aroused the basal ganglia,
cerebelluna, &c. ; (d) that the descending impulses, although unilateral when they
come mto relation with the spinal centres, are then brought into relation with both
sides of the body.

In order to investigate this, the first problem to solve is that referred to last under
(d). This, the present method, and that only is capable of doing, since by it we can
determine the characters of the actual descending impulses in the cord before they
come into relation with the bulbo-spinal centres. In this way we can ascertain to
what extent the impulses which descend the cord from the brain akeady show a
dual (bilateral) grouping.

We devoted a very large number of our cortical experiments to the further elucida-
tion of this question, feeling that the great importance of any approach towards its
solution, in consideration of its bearing upon the physiological characteristics of the
cortical cells, was a sufficient justification.

It is necessary, before stating our results, to give a short categorical account of the
work already done in this subject, and this is the more desirable since we cannot find
any such resume published.

Previous Work on the Subject.

The question whether both or only the opposite of the two sides of the body are
represented in one cortex cerebri on the efferent or motor side has been approached
experimentally by relatively but few authors. ■'■ As the results hitherto obtamed do
not by any means decide the important questions which ofler themselves for solution,

* The subject has also been considered fully from the theoretical point of view by Beoadbent
(' British Medical Journal,' 1876, pp. 333, 401), this author believing that bilaterality of representation
is always effected by commissui'es between the bulbo-spinal centi'e.s.

ON THE MAMMALIAN NERVOUS SYSTEM. 343

it is impossible to tabulate the facts in the definite manner used for the historical
retrospect of the graphic and other methods.

Moreover, since the interpretation of the results is completely dependent upon the
methods of experiment in each case, it will be better to arrange the facts as follows : —

(1.) Excitation experiments on the cortex.

(2.) Excitation experiments combined with division of commissures, i.e., corpus
callosum, &c.

(3.) Excitation experiments combined with the excision of the opposite " motor "
area.

(4.) Excitation experiments combined with hemisection of the spinal cord or bulb.

(5.) Ablation of one hemisphere.

(G.) Ablation of the "motor" area of one hemisphere followed by ablation of that
of the opposite side.

(7.) Excitation of the corpus callosum.

(8.) Degeneration of fibres after excision of jDortions of the cortex cerebri.

We will now briefly state the more important facts determined by the above
methods of experiment, postponing for the present any criticism.

1. Excitation Experiments on the Cortex.

It was first observed by Hitzig,''''' later by FERRiER,t Albertoni,J and Munk,§ and
subsequently by Franck and Pitres,|| that powerful excitation of one hemisphere in
the Carnivora produced movement not only of the corresponding or o23posite side of
the body, but also of the same side as that stimulated.

Franck and Pitres|| showed fui-ther that when the muscles of the same side were
thrown into action they contracted '01 second later than those of the opposite side, a
most important observation.

BuBXOFF and Heidenhain,!' Carville and DuRET,ir confirmed these facts.

Janicke'""- observed that while in dogs bilateral representation of the facial muscles
was very constant, in the limbs on the contrary, unilaterality was the rule, but that
to this there were exceptions. His views were confirmed by Unverricht.

LEWASCHEwtt noted that the movement on the opposite or coi-responding side was
a coordinated one (like a voluntary action), whereas that of the same side, besides
being late, was only a simple tonus.

* ' Untersuchungen iiber das Gehirn,' Berlin, 1874, pp. 48, 134.

t ' Functions of the Brain,' 1st ed., 1374.

X ' Lo Sperimentale,' 1876.

§ ' Gesammelte Abliandlungen,' 1890.

II ' Compt. Rend. Laboratoiro,' Marey, 1878, 1879.

^ Loc. cit.

** ' Centralblatt fiir klinische Medicin,' March, 1883, p. 177.

ft Loc. cit. See p. 346, infra.

344 MKSSRR. F. GOTCir AND V. HORSLEY

ScHAFEn and Horslky* noted that bilateral representation of the facial muscles
existed in \iuiou.s species of Monkey, but that the trunk and limb muscles were
unilaterally represented.

ScHAFEU and MoTTt confirmed these observations.

Beevor and Horsley;}; further showed in more detail that in the Bonnet Monkey
(Macac^^s sinicus) the limbs were unilaterally represented, but that other groups of
muscles, e.g., the tongue, buccinator oris, &c., &c., were bilaterally represented in each
hemisphere. They discuss this point also in another Paper on excitation of the
internal capsule,§ and confirm the above statement relating to the trunk muscles.

Semon and Horsley|| have shown that in all animals the vocal cord movements are
absolutely bilaterally represented in the excitable area of the cortex ; this bilaterality
was previously observed by Krause in the Dog ; it has been contested by Masini.

Brown-SequardU in one experiment on a Monkey found that excitation of the
gyrus fornicatus produced movements of the same side of the body, whereas excitation
of the paracentral lobule immediately above evoked movement of the opposite side.

AscH and Neisser** found from (but a few) experiments in Rabbits that excitation
of the cortex produced movement of the muscles on the same side, and, subsequently,
those of the opposite side, whereas excitation of the corona radiata gave movement on
the ojjposite side. The narcosis in their experiments was incomplete.

CouTYtt occasionally also observed the same phenomenon exceptionally in Rodents,
but only in the absence of narcosis. Steffahny^J observed occasionally bilaterality of
movement in Rabbits to follow excitation of one hemisphere. Braun§§ similarly
noted bilaterality in the absence of narcosis.

2. Excitation Experiraents on the Cortex combined with Division of Commissures,

i.e.. Corpus Callosum, dtc.

Exner|||| observed that in the Rabbit, after section of the commissures, and even
after ablation of one hemisphere (see No. 5), bilateral movements (nature not detailed,
possibly tonus, p. 189 of his paper), occurred.

* ' Phil. Trans.,' B., 1888.

t ' Brit. Med. Jour.,' 1890.

+ ' Phil. Trans.,' B., 1887, 1888, 1890.

§ ' Phil. Trans.,' B., 1890.

II ' Phil. Trans.,' B., 1890.

^ ' Coinptes Rendus de la Societe dc Biologic,' 1887, p. 2(31.

** ' Arcliiv fiir d. ges. Physiologie,' von Pflugee, 1887, p. 191.

tt ' Compt. Rend.,' vol. 96, 1883, p. 506.

XX Eckhard'.s ' Beitriige,' 1888, p. 97.

§§ Eckhard's 'Beitrage,' 1876, p. 127.

nil ' Wien, Akad. Sitzber.', 1881, 3 Ahth., p. 185.

ON THE MAMMALIAN NERVOUS SYSTEM. 345

Francois Franck and Pitres* found that in the Dog, after division of the corpus
callosum, the anterior and middle commissures, or even the pons, " les reactions
bilatdrales " persisted (? tonus only).

GLiKvt also found (see No. 4) bilateralitj of movement after dividing all parts in
the middle line to the mesencephalon, in the Rabbit.

Horsley| showed that in the Dog after division of the corpus callosura and com-
missures, as described by Franck and Pitres, although bilateral movements could
be obtained upon excitation of one hemisphere, yet the movements were not the same
in character on the two sides ; the true cortical effect of tonus followed by clonus
being only obtainable in the side opposite the excitation, whilst the effect on the same
side was only a feeble tonus {vide also Lewaschew) and often absent.

3. Excitation Experiments on the Cortex combined with Excision of the Oi^posite

"Motor" Area.

Franck and Pitres§ observed in the Dog the bilateral movements to persist
(? tonus only on the same side) even after the opposite motor area had been removed
(" centres corticaux opposes au centre excite "), and ExnerU also in the Rabbit noted
the occurrence of bilateral movements if one hemisphere were excited, even when the
■whole or major part of the opposite one had been excised.

HorsleyH found the same in the Dog, but noted that while the limbs opposite {i.e.,
corresponding) to the side excited developed the usual combination of tonus followed by
clonus, the limbs on the same side exhibited only tonus. Further, that as in Nos. 1
and 2 the complete (tonic plus clonic) discharge from the cortex of one hemisphere
could be obtained by adequate excitation without any bilateral movement whatever.

He also found the same to be true when the excitable region of one hemisphere
was ablated and absinthe injected into a vein. In the resulting epileptic convulsion
only tonus was noticeable in the limbs corresponding to the seat of ablation, whereas
the typical tonus plus clonus was exceedingly marked in the limbs opposite the sound
hemisphere. This differentiation was often very precise, i.e., the tonus on the side of
the sound cortex was very weak, even in the Cat.

In the Monkey the tonus mentioned was extremely slight, and possibly in Man is
absent (Obre) under these circumstances.

* Loc. eit.

t Eckhard's ' Beitrage,' vol. 7, p. 179, 1876.
J ' Brown Lectures and Reports.'

§ See for full discussion Francois Franck : ' Fonctions Motrices du Cervoau,' 1887, p. 59.
II Loc. cit.

^ ' Bi'own Lectures ' since 1885 ; also, ' Reports of the Brown Institution.'
MDCCCXCI. — B. 2 Y

346 MESSRS. F. GOTCH AND V. HORSLEY

Semox and Horsley^'' observed (li;it in all animals examined, i.e., Rabbit, Cat,
Dog, and Monkey, the l)ilateral movements of the vocal cords wei-e still perfectly
obtained on excitation of the one cortical representation after that of the opposite
side liad l>een reiuovod.

4. Excitation Experiments on the Cortex combined ivith Tlemisection of the Spinal

Cord, &c. (Eckhaed's Method).

Franck and PtTREst observed that in the Dog hemisection of the spinal cord on
the same side as that of the hemisphere excited failed to abolish the l^lateral move-
ments in the limbs.

LewasciiewJ followed the French authors by similar experiments in the Dog, from
which he deduced the same idea, viz., that the bilaterality was accomplished by means
of commissural fibres in tlie spinal cord.

Balighian^ observed that in the Eabbit the opposite, i.e., normally corresponding,
movements were not abolished by hemisection of the bulb opposite the lower border
of the pons and on the same side as that of the hemisphere excited, whence he con-
cluded that the crossing of the excitable fibres begins to occur as high as the lower
border of the pons.

Balighian also showed that in the Rabbit, excitation of one hemisphere failed
to elicit movement in the opposite limbs if the corresponding part of the spinal cord
had been divided.

Schiff|| found that in the Dog after section of the right crossed pyramidal tract
excitation of the cortex with four Leclanche elements jjroduced no result (whereas
before the division this formed a maximal stimulus), and only with a current of
fourteen elements were movements of both the hind limbs obtained.

Gltky^ showed by hemisection of the bulb that in the Rabbit the pyramidal tract
cr'ossed below the centre of the fourth ventricle, and that a tendency towards
unilaterality of representation prevailed.

Steffahny,'"'"" applying this method in Rabbits, came to the following conclusions :
that the path of the crossed impulses for the innervation of the extensoi's of the fore
limb is in the uppermost part of the cervical cord, the anterior column, and lower
down in the lateral column, and further that there are paths for bilateral movements,
and that these lie in close relation to those for the ordinary crossed effect.

* Loc. cit.
t Tioc. cit.

X Pi-lugkr's ' Archiv fiir die gesammte Phjsiologie,' vol. 2G, 1885, p. 279.
§ Eckhard's ' Beitrage,' vol. 7, 1876.
II Pfluger's 'Archiv,' vol. .30, 1883, p. 248.
1 Eckhard's ' Beitrage,' vol. 7, 1876, p. 186.
** EcKH-iRD's 'Beitrage,' vol. 12, 1888, p. 43.

ON THE MAMMALIAN NERVOUS SYSTEM. 347

Schiff/'' by limited section of tlie parts of the cord, was the first to find that after
division of tlie crossed layramidal tract in Dogs, no movement of the leg on the same
side as the lesion followed upon excitation of the opposite excitable area.

DupuYt stated that in the Dog opposed hemisections of the cervical region of the
cord when sepai'ated from each other by about 1-1 '5 cm. did not interfere with the
production of movements in the lower limbs on exciting the cortex cerebri.

5. Ablation of one Hemisphere.

To test localisation in principle Goi.Tz\ has in the Dog removed large proportions of
the brain, and recently^ succeeded in completing the ablation of one hemisphere.

The animal under tliese circumstances could walk, run, and use the limbs in all
automatic movements of feeding, &c., there being no obvious persistent hemiplegia as
in Monkeys and Men. For many weeks, however, there is in the Carnivora marked
hemiplegia of the opposite side to the lesion (all authors), and according to Hitzig
the loss of the " muscular sense " is permanent in these animals. This fact, as well as
the concomitant persistence of antesthesia to moderate tactile impressions, is also
confirmed by the observations of ScHiFF, and one of us (V. H.).

Semox and Horsley|| found that in all animals examined after removal of one
hemisphere the "automatic" i.e., respiratory, movements of the vocal cords were
perfectly bilateral.

6. Ablation of the " Motor " Area of one Hemisphere, followed by Ablation of that of

the Opposite Side.

Carville and DuretIT investigated the matter by removing one so-called motor
area, and observing the paresis caused thereby, noting that the paresis gradually dis-
appeared as if by substitution on the part of the opposite sound hemisphere.
Removal of this latter, however, did not reproduce the paresis of the limbs on the
same side.

7. Excitation and Degeneration of the Corpus Callosmn.

The results of excitation of the corpus callosum by Schafer and Mott,** as well
as the degeneration observed by SnERRiNGTONtt to follow localised ablation of portions

* ' Arcliiv f. d. gcs. Physiologie,' vol. 30, p. 248.

t ' Compt. Rend. Soc. do Biol.'

+ " Verrichtungnn ties Grossliiriis." Pfluhek's ' Ai-cliiv,' vol. 34, 1883, p. 50.

§ Demonstration at the 1st Internat. Physiol. Congress, Basle, 1889. See also the full account given
by Langley and Grunbaum, 'Journal of Physiology,' vol. 11, 1890.

II Loc. cit.

^ ' Archives de Physiologic ' (Paris), 1875, p. 446, &c.
** ' British Medical Journal,' 1890 ; also ' Brain,' 1890.
tt ' Proceedings of the Physiological Society,' 1889.

2 Y -2

;'!48 MKSS1«S. !•'. COTCII AXU V. IKHtSI.EY

of one cortex, suj;-<,'est that tlie corpus e;illu.suiu is a true coiiiiuissure hetweeu the
excitable or " motor " areas.

Bkown-Si'^QUARd'' had ])reviously shown tliat the corpus callosutn was excitable
about its middle third, and he conseqnently regarded its fibres as commissural between
the hemispheres.

8. Degeneration of Fibres after Excision of Portions of the Cortex Cerebri.

The phenomenon uf bilaterality of function has also Ijeen refei-red to the normal
exercise of those fibres which degenerate bilaterally (PiTREst) in the spinal cord after
lesion of one hemisphere. Although the degeneration method has established com-
pletely the existence of such atrophy of channels in both lateral columns of the cord
consequent vipon a unilateral cerebral lesion, we do not, unfortunately, know whether
these channels are " recx'ossed," as suggested by Suerrington,;|: Charcot, and others,
or what is their destination, or whether they are to be regarded as of constant
occurrence. As regards the latter point, it is certainly at present -considered that
they are not constantly affected by a cerebral lesion, and yet more rarely degenerate
after a hemisection of the spinal cord. As yet, therefore, the method does not afibrd
anatomical means of deciding the questions at issue.

Relation of the Foregoing Facts to our own Experiments.

In summarising the facts thus collated on this subject, it is difiicult to avoid discus-
sing the theoretical interpr-etations advanced by the authors quoted, but we do not
think that anything is to be gained by such a procedure, and intend now to merely
point out what conditions yet remain to be satisfied before anything like a full con-
clusion can be arrived at.

All agree that in the intact nervous system a nerve impulse from one hemisphere
may readily pass to the other, excite it, and thus bring about bilaterality of movement
as a result.

Nothing, however, can be judged as to such crossing to the other hemisphere being
necessary for bilateral function until it is clearly defined in which part of the body
such functional effects occur.

We must for this reason exclude altogether from the present discussion the
facial movements, since these § (see p. 344) are in great measure bilaterally repre-

* ' Compt. Reud. Soc. de Biol.,' 1879, p. 165 ; 1881, p. 204.
t ' Ai'chives de Pliysiologie,' 1884, p. 142.

X ' Journal of Physiology,' 1885, p. 177. See also summary by Tooth : " Gulstonian Lectures on
Secondary Degenerations of the Spinal Cord," 1889.
§ Cf. Paneth ; also UxvERRifnT.

ON THE MAMMALIAN NERVOUS SYSTEM. "49

sented in the excitable cortical areas of each hemisphere. The question, therefore,
narrows itself down to that as to whether the highest, i.e., purposive, movements of
the limhs of both sides of the body are represented in the cortex of one and the same
hemisphere. Theoretically, we are compelled to admit (following the teaching of
HuGHLiNGS Jacksox) from the evolutionary standpoint such bilateral representation
of the limbs. The important point for experiment to decide, however, is whether such
bilateral cortical representation exists in more highly differentiated animals to a suffi-
cient extent to cause movements. The opinion of most authors is evidently that it
does so exist. We feel, however, very strongly that the methods hitherto adopted by
these authors are not definitive, and do not establish the positions claimed. We are led
to this conclusion from the consideration of certain facts now to be discussed which have
come under our notice in the present research, as well as of others previously dis-
covered, the importance of which has of late become moi'e recognised. By welding
the fresh information, which the use of our method has given, to the old, we hope to
help forward the solution of this apparently simple but very complex subject.

(1.) Narcosis. — As is s^en from the foregoing retrospect, some authors have employed
narcosis to a greater or less degree. The statement that a narcotised cortex could in
any way " completely discharge " is always, of course, open to objection, and hence
observations in animals narcotised to unconsciousness have been held by some to be
incomplete. The answer to this objection, however, is simple and, we believe,
sufficient ; it is included in that of the next paragraph, in which " complete discharge "
of the efferent appai'atus of the cortex is seen to be effected even in unconsciousness.
Clinical experience of epilepsy also affords evidence of the truth of this contention.

To look at the question from the opposite standpoint, although electrical excitation
of the cortex is not in any way painful, still it is clear from observation of the
influence of such stimuli in imperfect narcosis that the effect spreads rapidly from centre
to centre, i.e., to the opposite hemisphere, &c., and hence renders any topographical
conclusions impossible.

The observations, therefore, which have been made in this way, though very
valuable as throwing much light on synchronous excitation, e.g., in epilepsy, &c.,
cannot, as yet, form a basis for the determination of the presence or absence of
Vjilaterality of rejaresentation so far as the limbs ai-e concerned.

(2.) An adequate stimulus which completely discharges the cortex at one given focus
of representation of one limb produces movement in that limb only and none in the
limb of the same side.

By the term "adequate stimulus tvhich completely discharges the cortex" at one
given spot, we mean an interrupted induction current of sufficient strength to evoke
a strong movement in the limb represented (in the present research the leg), and to
produce a slight excitation also (by overflow of nerve impulses) of the nearest lying
centre or focus, which in our experiments was naturally that for the fore or upper
limb. If the strength of the stimulus and the condition of the cortex be accurately

350 , MESSRS. F. GOTCH AND V. ]I()RSLKY

gauged btfurehaiid, so tliiit the latter is nut tlirown into a iiypcrexcitable state, the]i
the above mentioned phenomenon can invaria,bly be obtained, and this not only in
tlic Monkey but also in (lie Cat. It might be objected (hat the cortex was not
completely discharged, but we regard the overflow tA' tlie excitatory changes into the
neighbouring foci as sufllcient evidence of the required completeness. Moreover, the
converse position, viz., the appearance of bilaterality tells the same (story see pp.
354 and 35i)), and thus forms the crucial argument.

If we are warranted, therefore, in our view that the cortex under these circumstances
is completely discharged, the above described phenomenon negatives the practical
existence of bilateral re})resentatioii of the limb muscles in these animals.

(;3.) When (" Jnlateral") movements of both lunhs follow excitation of one hemi-
sphere after the excitahle cortex of the opposite hemisiihere has been thrown out of <jear
by ablation, division of the commissures, dc, dec, the movement of the limb on the side
opposite to the cortex excited is the com])lete cortical effect of tonic folloived by clonic
contractions, lohereas the movement of the limb on the same side as that of the cortex
excited is only a tonic contraction.

If bilateral movements of both limbs receive their originating impulses directly
from one cortex, it is not com})rehensible why the above-mentioned striking difference
in the kind of movement of the limbs should exist. This difference, specially insisted
on by one of us,* has been also observed by several authors (Lewaschew, &c.), and
it suggests that the movement noted on the same side as the excitation is of some
origin other than the cortex (its corresponding cortical apparatus being destroj'^ed, be
it remembered), for a simple tonus lasting during the excitation is characteristic not
only of the cortex but of the cerebellar or other lower centres.

(4.) When bilateral movements of limbs are observed, those of the limb of the same
side as the cortex excited are always later in commencement than those of the opposite
side. (France and Pitres.)

This delay in movement of the limb of the same side is attributed by most to loss
of tune in traversing basal commissures. There is no means of testing this view
except by some arrangement in which the exclusion of the said commissures is
provided. As far as the commissures in the cord are concerned, this might be achieved
by ascertaining the time relations of such excitatory electrical changes as will pre-
sently be shown to appear in each half of the longitudinally divided coi'd upon
stimulation of the cortex.

(5.) All experiments on tlds subject have included in their anatomical plan the
cerebellum, ivithout excluding its functional influence.

In our own experiments about to be described, as well as those of the authors
ali'eady quoted, the cerebellum has been left in normal connection with the central
structures excited.

In performing the experiment of exciting one hemisphere after ablation of the

* V. H., lor. cit.

ON THE MAMMALIAN NERVOUS SYSTEM.

351

other, it is clear that the association through the superior peduncle of the cerebellum
affords a means whereby the opposite lobe of that organ miglit be aroused, and so
produce the tonic contractions of the limb on the side of excitation. This view is
suggested to us more especially by the teachings of Dr. Hughlings Jackson, and it
obviously must be excluded before a positive opinion can be expressed as to the
bilateral representation of the limbs in the excitable cortex of one hemisphere.

From none of the foregoing researches can it be determined to what extent the
bulbo-spinal centres are associated with bilateral movements, since in all experiments
in which muscular contractions are taken as an index their functional activity is
included.

Having laid before the reader these general considerations by way of preface, we
will proceed to describe our own experiments, as far as they suggest fi-esh evidence for
or against the different views just enunciated.

Experimental Eesults in connection with Bilatekality.

The obvious elimination which the use of our method enables the experimenter to
obtain is the removal of the influence of the bulbo-spinal centres.

This elimination is effected by the division of the spinal cord in the dorsal region,
and the observation of the electrical changes occurring in its central end when the
lower limb ai'ea of the cortex cerebri is excited. It is, however, essential to divide the
descending tracts in the cord into two halves. This is done by splitting the cord
longitudinally in its antero-posterior plane. The mode of operation employed has been
described in Chapter III., Section 2,* and the preparation shown in Plates 31 and 33.

* It might reasonably be conceived that this operative procedure would seriously impair the con-
ducting power of the cord. Although, as we point out in Chapter III., the posterior columns suffer some-
what, this is not the case with the lateral columns, the seat of the pyramidal tract to be investigated.
An example of this is to be seen in the following case, where the effects in the divided cord of exciting
but one hemisphere do not, when summed, fall far short of the effects obtained before the longitudinal
section :

352 MESSRS. F. GOTCII AND \ . IIORSLKY

As regards the nietliod of observation, the electrical changes in each half of the longitu-
dinally divided cord were recorded by means of the galvanometer, since it was essential
to obtain results which admitted of relatively strict cora])arison as to their amounts.
Each half of the cord was therefore attached to an independent pair of non-polaris-
able electrodes, and an arrangement made by which either pair could at any desired
moment be switched into connection with the wires leading to the galvanometer, &c.

The result of this experimental Investigation may be divided into groups, each one
of which we must consider in detail.

l.—Excitation of Cortex (whole Encephalon intact). — Electrical Changes in each

half of Longitudinally Divided Cord.

The cord having been divided, and .split as just indicated, one cortical surface
was exposed and the animal having been brought as far as possible into a perfectly
steady, i.e., constant state of nai'cotisation, the same strength of stimulus was then
applied to the cortex, first one half of the cord being in connection with the galvano-
meter, and then the opposite half (see fig. 8). The results obtained, i.e., from the
opposite side of the spinal cord, and from the same side respectively, were then
gathered together and averages taken.

It will be best to begin with the results of special experiments. Of these, the
first we will refer to were made upon an animal (Cat 237), in which, from the move-
ments of the upper limb, it was very easy to ascertain when the muscular contractions
were unilateral or bilateral. It was seen that when the said contractions were .strictly
unilateral, the excitation of the hemisphere produced no result in the half of the cord
on the same side, but a marked result in the half of the cord of the opposite side, viz.,
275 degrees of the scale.

In this case complete unilaterality for the lower limb existed as far as the pyramidal
tracts in the coi-d were concerned, for the cortex was, as we have before indicated in
the first of our considerations, completely discharged, the discharge being evidenced
by the large effect in the corresponding half of the cord, yet no electrical change
could be seen in the other half indicative of descending impulses.

In another experiment made upon a Monkey (215), both cortices were exposed for
excitation. In this animal, it was easy to confine the excitation apparently to one
hemisphere, although, judging from the contractions of the muscles of the upper limb,
the cortex excited was very completely discharged. In this instance, the cord was
divided as described betw^een the 10th and 11th dorsal nerves; the resting electrical
dlfiereuce of the two halves was in the proportion of two to one, the smaller being
the right.

For the convenience of brevity, we will speak of the halves of the cord as the " right
cord " and " left cord " respectively.

Excitation of the left hemisphere gave, with the coil at 5000, an effect in the right
half of the cord of 425 degrees, and then, the right cord still remaining in connection

ON THE MAMMALIAN NERVOUS SYSTEM.

353

with the galvanometer, the right hemisphere was stimulated, the result being an effect
of only 22 degrees. This last result, in addition to being only a twentieth of that
evoked by the corresponding cortex was probably due to discharge from the left hemi-
sphere, for, after the cortex had been allowed to rest, this pair of observations was

Fig. 8.

repeated with the following striking results, viz., the right cord being observed, the
left hemisphere was excited, with a resulting variation of 3G5 degrees. The same
cord being still led off, the right hemisphere was excited, i.e., the hemisphere of the
same side, and no effect was produced in the galvanometer. The same degree of
narcotisation being maintained, and the same strength of coil being employed as that

MDCCCXCI. — B. 2 Z

354

MESSRS. F. GOTCII AND V. HORSLKY

just inentioneil, viz., 5000, the left cord was then coiuiocii'd witli the galvanometei- ;
excitation of (lie lii^hL hemi.si)here {i.e., thai of the corresponding side) gave 375 degrees,
and fxcitation of ihv h'i't hemisphere {i.e., tlial of the same side) only gave 8 degrees.
This mode of pairing the observations is clearly one likely to give the most useful
results, but we also paired the observations in the following way, viz., by connecting
first one side of tlie cord with the galvanometer and then the other, in each case
stimulating the same hemisphere.

The foregoing two selected experiments illustrate the conditions which are observed
with the hemispheres in a normal state. Before considering the further evidence
afforded by massing our observations, we must draw attention to the greatest
diihculty in connection with this branch of enquiry, and which prevents us from
speaking so positively upon this point, when the results noted at the beginning of
any experiment are added to those recorded at the end. We refer to the great
tendency of the cortex to become hyperexcitable after one or two excitations, and,
consequently, for the excitatory state to pass from one hemisphere across to the other.

This is especially liable to occur when through repeated excitation the cortex has
been thrown into a hyperexcitable state. This is, in fact, the objection which may
be urged against any conclusions derived from massing together results, many of
which beino- repetitions involve previous excitation. The same objection can also be
urged with equal truth against the results of most of the observations of earlier
experimenters ; but upon the question whether this objection accounts for the whole
of the results to be immediately given, we have no means of expressing any decisive
judgment. It is, however, easy to understand that a considerable proportion of the
bilateral effect noted in the two halves of the cord when all the results are taken
may be due to this circumstance.

The average amount of the bilateral effect obtained in the two halves of the split
cord when one hemisphere is directly excited is as follows : —

Cat — Average of 12 observations . . .
Monkey— Average of 31 observations .

Cord.

Opposite half.

Same lialf.

1.57
158

33
32

It will be seen that when all results are averaged together there is an effect on
both sides of the cord, and that its amount in the half opposite to the excited cortex
is tive times as great as that in the half which is on the same side.

The close identity of the figures in the two animals is remarkable, and not what
our early observations had led us to anticipate."

* One of us (V. H.), however, had already shown by experiments with absinthe combined with

ON THE MAMMALIAN NERVOUS SYSTEM. 355

The method under consideration, with both cortices exposed but intact, is not,
however, favourable for testing bilaterality, since we have often had occasion to
observe the great inequahty in the excitability of the two hemispheres, which appears
to be to a large extent connected with the necessary exposure of one cortex before
the other.

From the result of other experiments we are inclined to believe that the effects
(average 32) observed in the half of the cord on the side of the cortical excitation is
due to the passage of impulses which have descended from the cortex opposite to that
directly stimulated, but several interpretations are, of course, possible. In this con-
nection it must be remembered that we do not yet know whether a direct pyramidal
tract exists in the Carnivora (most authors denying its presence), and, further, that
possibly even the galvanometer may not show an excitatory disturbance if the fibres
by which the hemisphere of one side might be in relation with the same side of the
spinal cord were very few in number.

II. Excitation of Cortex. — Effect in Cord after previous Hemisection.

The next step was to ascertain the effect of a previously performed hemisection of
the cord between the encephalon and the observed region, the experiment being made
with the view of eliminating any presumably crossed discharge from the cortex
opposite to that excited.

The following experiment was made : In a Cat (225) the right half of the cord was
divided under antiseptic precautions at the level of the lower border of the 9th dorsal
vertebra. In this case there was well-marked motor paralysis in the right hind limb
with rigidity, and diminution of perception of sensory stimulation of the same limb.
The right knee jerk was exaggerated. When partly etherised the right hind limb
became flaccid and the left somewhat rigid, the knee jerk then on the right side was
greatly exaggerated and marked clonus present. These facts were noted just before
the experiment now to be described. Eighty-four days later the cord was exposed
and divided below the lesion, and the central end of the whole cord connected with
the galvanometer electrodes. (See fig. 9.) The two cortices were then exposed and
excited with an intensity of stimulus indicated by coil 8000 for 5 seconds.

The following definite results were obtained : —

In the first place excitation of each hemisphere evidently produced unilateral
epileptic fits. The effect in the descending (lower limb) fibres of the cord was as
follows, it being remembered that the right half of the spinal cord had been divided.
Excitation of the right cortex which produced unilateral fits on the left side of the
body, gave a variation of 82 degrees in the galvanometer, whereas excitation of the
left cortex, although it produced a good unilateral fit in the parts above the section,
i.e., right upper limb, gave nothing in the galvanometer, although the ivhole cord

iihlation of one hemisphere or excitable area in Cats, that apparently in that animal roniplotc unilnteiality
of the true cortical discliai'go {i.e., tonus //Z«s clonus) existed.

2 z 2

356

MESSRS. P. GOTCIl AND V. HORSLEY

was the seat of galvanometric observation. In Uiis way it was obvious that, when the
excitation was so limited to one hemisphere as to produce only unilateral nuiscular
movements, no impulses descended the fibres on the same side of the spinal cord

Note the hemisection of the cord on the right side at the level of the 9th doi-sal vertebra.

and all impulses passed down the side opposite to that of the hemisphere
excited.

That this conclusion was correct was confirmed by observing that even when on
repeating the experiment a bilateral fit was obtained from stimulating either cortex,

ON THE MAMMALIAN NERVOUS SYSTEM.

357

whilst the cord effect evoked by the right cortex was increased to 134 scale, an effect
was also evoked by excitation of the left cortex, though it amounted to only 15 scale.
The cord was next divided longitudinally and the electrical change observed on
the two sides with the following result, it being remembered that the right half of the
cord had been interrupted by the previous hemisection.

Left coi-tt'x

Right ,

Left ,

Right „

Left „

Right „

Excitation.

Left lialf.

Right half.

6000, 0 sees.
66u0,

8000. ;;

trace

60

trace

215

5, 6, 20

150

trace
trace

It is clear from this that in the half of the cord (left) which offered an
uninterrupted channel, the excitation of the cortex of the opposite side evoked large
results, whilst the excitation of the cortex of the same side side evoked no results
until a considerable intensity of stimulus was used and a bilateral fit produced. On
microscopical investigation the lesion in the cord showed that the whole of the right
half of the cord was destroyed, and the anterior third of the left posterior column, and
the left posterior median. The descending degeneration, therefore, affected the right
pyramidal tract in the lateral column only. The ascending degeneration affected both
postero-median columns and the right cerebellar and antero-lateral tracts.

It should be mentioned that Schiff's observations on hemisections made just
before excitation of the cortex show that the same relationships prevail as regards
muscular movements.

III. Excitation of Corona Radiata. — Electrical effect observed in each Half of Split

Cord.

The general electrical phenomena observed after excitation of the corona radiata,
and especially as contrasted with those elicited by exciting the cortex are described
in Chapter VI., but in attempting to still further elucidate the subject of bilaterality
we arranged a third variety of experiment, originally designed for the graphic method
by Fraj^ck and Pitres and others, viz., the investigation of the bilateral phenomenon
noted after excision of the cortex, and consequent upon excitation of the subjacent
corona radiata.

We have considered, on p. 338, the errors it involves, and have always endeavoured
to minimise them as far as possible just as in the case of the cortex. Although
the results are thus of necessity only to be accepted with reservations on the
points already mentioned, which we must again briefly notice, nevertheless they do
afford some very suggestive deductions.

358 MRSSRS. F. GOTCH AND V. HORSLEY

We will ilrst give the results obtained in the longitudinally divided cord by excita-
tion of one corona radiata.

In this series of experiments we connected either half of the longitudinally divided
cord with the galvanometer as in the cortex experiments previously described.

We have not yet seen any absolutely unilateral effect, as in the case of stimulation
of the cortex. Very often the effect in the same side of the cord -was extremely
small (once only a trace), e.g., 4°, 0°, 8°, &c., but it was always present. On massing
the results together, i.e., adding the observations in the Monkey to those in the Cat,
we obtained the following proportionate averages : —

Half of coi'd on same side as excitation of corona — 14° mean of 12 observations.
Half of cord opposite or corresponding to the corona excited — 50° mean of 13
observations.

The proportion, therefore, is 7 to 2, whereas the cortex proportion is almost 5 to 1,
and often purely unilateral. Consequently it appears that bilateral phenomena are
more easUy elicited by exciting the corona radiata than the cortex cerebri, provided
the same care is employed to avoid as far as possible in both cases errors due to
spread, &c.

The explanation of this is not far to seek, and, if correct, throws more light on
bilaterality. It simply consists in the obvious fact that removal of the cortex lays
bare a crowd of association fibres, the excitability of which is heightened by the
section removing the cortex, and that it is the spread of the excitatory effect along
these fibres to other central mechanisms in the encephalon, i.e., opposite cortex, cere-
bellum, &c., which produces the bilateral phenomenon. It is thus extremely likely
that bilateral tonus should occur after stimulation of but one corona radiata.

In accordance with what we have said before respecting the influence of ether, this
reagent afforded an opportunity of testing the truth of the foregoing views, since the
influence of bilateral central mechanisms could be by this means in part excluded, and
it could therefore be seen whether or no there was a corresponding modification in the
bilateral character of the effect.

We found that deeper etherisation invariably tended to restore in a great measure
the inequality between the two cord effects, the difference between the two being
much greater in proportion as the degree of anaesthesia increased. This again seems
to indicate that any markedly bilateral effect is due to an additional functional
activity in the cortex on the opposite side to the excited corona.

IV. Excitation of Corona Radiata. — Effect in Cord ivith previous Hemisection.

It was necessary now to repeat the design of the graphic experiments as employed
by Franck, Schiff, Lewaschew, and others, using, however, our present electrical
method, that is to say, observing how far a hemisection of the cord between the

ON THE MAMMALIAN NERVOUS SYSTEM.

•359

encephalon and the part observed interrupted the propagation of impulses fx-om
the corresponding corona fibres.

The hemisection was made in the case (Cat 225) before referred to, p. 356, two
months beforehand, at the 9th dorsal vertebra. For the observation the cord was
exposed, divided, and split longitudinally at the level of the 1st lumbar vertebra.
Both coronfe radiatse were then exposed and excited alternately, the eftects evoked
being given below.

Excitation.

Cord.

Galvanometric effects.

Left half.

Right half.

Right. Corona, 12,000, 3 seconds . . .
Left ,, „ „ . . .
Right „ 13,000, „ ...
Left „ „ „ . . .

30

Trace

51

4

0
0

In this case it is evident that the effect was almost completely unilateral.

In another case in which the hemisection was made at the time, there was no com-
plete unilateral effect, but excitation of both coronse elicited changes, the corresponding
corona evoking an effect which was twice as large as that produced by stimulation of
the one on the opposite side to the lesion. By means of etherisation the proportion
was increased to 3 to 1, but since the alterations in excitability, due to the lesion
being performed at the time, were disturbing factors, the influence of which could not
be gauged, the experiment was not pursued further. It may be fairly concluded,
however, that the more perfect the elimination of such complications, the more
complete is the uuilaterality of representation.

V. Excitation of Corona Radiata after Complete Removal of one Hemisjyhere.

Effect in each Half of Divided Cord.

In order to endeavour to ascertain whether the bilateral effect was in relation, not
merely with the discharge of portions of the cortex of the hemisphere opposite to that
of the excited corona, but also in relation with the basal ganglia and the cerebellum,
we have in one animal (Cat 309) observed the influence of complete removal of one
cerebral hemisphere.

The cord was, in this case, split longitudinally, and both corona) exposed for
alternate excitation, the results of several observations being given below.

360

MESSRS. F. GOTCII AND V. IIORSLKY

Excitation of one corona.

Eifcct in opposite
half of cord.

Effect in same
half.

Right, 8000, 5 seconds

I> )) n

II )) n

Left „ „

11 11 11

11 11 11

Average

1(50
61
62
54
54
50

20
8
20
34
11
25

G = 431

6= 118

72

20

The left cerebral hemisphere was now removed, and the effect of excitation of the
I'ight corona observed in each half of the cord.

Right corona

Average

Opposite side.

Same side.

100
75

27
30

2 = 175

2 = 57

87

28

The result is to show that the bilaterality continues, and apparently to the same
extent as before.

The experiment is one which needs careful repetition, and special precautions to
prevent errors.

VI. Excitation of Cortex and Corona Radiata. — Effect in Sciatic Nerves.

It has been already stated (Chapter V., Section 2, p. 332) that the excitation of the
cortex evokes electrical effects in the sciatic nerves ; and, further (ChajDter VI.,
Section 2, p. 339), that similar effects are evoked by excitation of the corona radiata.
A comparison between the amounts evoked in the nerves by excitation of one cortex,
or by observation of the changes in one nerve consequent upon alternate excitation of
each hemisphere, throws additional light upon the question of bilaterahty of repre-
sentation of the limbs now under discussion. It will be in the remembrance of the
reader that the nerve effects derived from cortical excitation are remarkably small in

ON THE MAMMALIAN NERVOUS SYSTEM.

361

amount ; in consequence of this, the comparison necessarily becomes less exact when
the metliod of averages is used.

The following table gives this comparison in the case of six animals, four Cats and
two Monkeys. In three cases, (a), (/;), (r), the eftect was observed in one nerve and
each cortex alternately excited; in the remaining three cases, ('/), {c), {/), the
observations were made upon each nerve, one cortex only being exposed and
stimulated.

(a) Cat (296)

(h) Cat (290)

(c) Cat (298)

(d) Cat (299)

(e) Monkey (54)

(/) Monkey (217)

Average

Intensity of
stimulation.

Effect in nerve
of opposite side.

Effect in nerve
of same side.

11,000

. 12,000

12,500
12,000

5,000
5,000

5
9
12
20
18
10
22
28
12
20
10
16
18
18
28
16

0
0
0
0
2
0
0
0

10
0
0
5
1
0

18
5

262

41

16 2-6

From this table it is seen that, in the large majority of cases, unilaterality strictly
prevailed. The small average amount of the bilateral effects, dependent, moreover,
as it is upon a very few high readings, suggests that these results are due to that
constant source of error in experiments upon bilaterality of representation, viz.,
hyperexcitability in other parts of the encephalon, produced by the operative
procedure.

Examination of our records shows that when one coroiaa is excited, and first one
nerve and then tlie other connected with the galvanometer, the following average
results were obtained : — •

Corona radiata excited.
Average of all effects . Nerve opposite side 26 (23 observations).

Nerve same side 1 1 (20 observations).

This average, however, is the result of all the observations made, some of them
being markedly bilateral in character. There are, however, no less than six cases
in which no effect at all was observed on the same side as the excited corona,

MDCCCXOI. — B. 3 A

362 MESSRS. F. GOTCH AND V HORSLEY

and four cases in which the effect was under 10, and averaged only 4. It is, there-
fore, only by including cases of bilateral effects produced by powerful tonic discharges
that the above total average is so conipai'atively high.

If, therefore, these higher figures were separated fi'om the Table, we should find that
the average effect on the same side would be not more than one-sixth of that on the
opposite side ; whilst, with evident and powerful bilateral discharges, the former effect
amounts to one-half of the latter.

Summary of Facts.

The facts brought forward in the preceding pages seem to show that it is possible
to obtain completely unilateral effects in both spinal cord and nerve when one
cerebral hemisphere is excited, but that an increase in the intensity of the stimulus,
and a diminution in the degree of narcosis, favour the production of bilateral effects —
the inequality between the crossed and uncrossed effect becoming less and less marked
in proportion as these favouring circumstances are augmented.

The conclusion to which the previous observations seemed to tend, that in one
cortex bilateral representation of the limbs exists, does not seem to be supported by
the present experiments, since such bilateral effects as may be witnessed are specially
brought out by agencies which may be supposed to bring into play other portions of
the central nervous system, particularly the opposite excitable cortex, the cerebellum,
and basal structures. We feel, however, that without definitely proving this, our
experiments have the result of making the question of bilateral representation in
one cortex an open one, at least for the Carnivora.

To this position all we wish to add is our view that the weight of evidence goes to
show that, where the excitation is properly limited to the cortico-pyramidal system,
unilaterality of representation of the lower limb muscles appears to exist.

How far this is correct only further researches by the electrical method will, we
believe, be successful in showing ; and we hope that others will forward the investiga-
tion of this point.

ON THE MAMMALIAN NERVOUS SYSTEM. 3G3

CHAPTER Viri.— ON THE ELECTRICAL EFFECTS EVOKED IN THE SPINAL CORD BY
THK EXCITATION OF THE VARIOUS PARTS OF THE SA]\IE.

Section 1. — Iiitrdiluct.ory.

Section 2. — Propagation of Impulses by the fibres of the Cord.

Section 3. — General characters of electrical effects in the Cord following its excitation.

Section 4. — Excitatory electrical effects in the Cord, as evidenced by the Electrometer.

Section 5. — Electrical effects evoked by localised stimulation of the Spinal Cord ; plan of

experiments.
Section 6. — Electrical effects in the Lumbar Cord, following excitation of the Dorsal Cord.
Section 7. — Electrical effects in the Dorsal Cord, following excitation of the Lumbar Cord.
Section 8. — Electrical effects in each half of the divided Cord.
Section 9. — Influence upon electiical effects in the Cord of intervening sections of the various

columns.
Section 10. — Summary of results of experiments.

Section 1. — Introductohy.

The electrical changes produced in the cord by excitation of the cortex, and of the
fibres of the corona radiata, whilst due to the passage of nerve imj^ulses along
tracts in the spinal cord, derive their interest from the further knowledge which they
give us with reference to the functions of the excited parts in the encephalon from
which those impulses spring.

It is otherwise with the material of this and the succeeding chapters ; since
the electrical changes, now to be described, are evoked by the excitation either of the
difterent parts of the cord itself, or its nerves. The method of determining the
characters of the functional activity of nerve tissue, by the study of those electrical
effects which undoubtedly accompany and indicate the extent of that activity, is thus
to be now applied solely with relation to the cord.

Since the functions of the cord ai'e naturally divisible into those connected
especially with its fibres — conductivity — and those connected especially with the
activity of its cells, of which reflex action is the example, the present method was
applied in any given experiment with special reference to the elucidation of one of
these two branches of enquiry, it being always borne in mind that, as a matter of
fact, the two groups of function overlap.

The question of the localisation of paths, in the fibres of the cord alone, has at
present been only approached in reality by the method of histological experiment,
including the embryological and degeneration methods, since previous physiological
observation, relying on movement as an index, has always included the whole of the
neuro-muscular mechanism.

The sole method for obtaining actual indications of the conduction of physiological
processes in the fibres of the cord, is that of determining the presence in them of
excitatory electrical changes. We have, therefore, carried out a very large number of

3 A 2

;^G4 MESSRS. F. GOTCII AND V. IIORSLEY

experiments upon the .spiunl curd under very different conditions, with the express
]iui-jiose lioforc n.s of ol)taining data wliich sliould give clear evidence of tlu! cnnduc-
tiou (iriuM\-c ini|mlst\s in its fibres and centres respectively. 'J he results furni-shed by
our ex})erinients may be spill up into the following groups : —

A. The experimental evidence of definite localisation of the channels of conduction
of nerve impulses afibrded ))y electrical changes in the cord, when evoked by stimula-
tion of its different parts, whether distal or proximal.

B. The evidence of the localisation in the cord of fibres which enter it by the roots
afforded by electrical changes in the cord when evoked by stimulation of its nerves.

C. The evidence of the relations of the spinal cord to the nerves afforded by
electrical changes in the nerves, when evoked by stimulation of separate parts of
the cord.

D. The experiments elucidating the complicated group of phenomena in relation
with the reflex activity and function of the spinal nerve cells. This may be expressed
as follows : —

The evidence afforded by the electrical changes, in both cord and nerve, of the
nature of the role of the nerve corpuscles in the cord.

To obtain a sufficient mass of evidence, to establish even a few conclusions in these
four subjects, a large niunber of experiments have been carried out, constituting by
far the larger share of the work we have done since our preliminary communication in
the ' Proceedings of the Royal Society,' in 1888.

The groups of data just indicated, in which these results are expressed, will be
treated of in the succeeding four consecutive chapters ; group A. in the present
Chapter VIII. , B. in Chapter IX., C. in Chapter X., and D. in Chapter XL, this
being the order which seems to us at once the most natural and the most likely to
present the conclusions in a logical and thus intelligible manner.

We therefore now pass to the consideration of the experiments, confining ourselves
in this chapter to the electrical changes evoked in the spinal cord by stimulation of
its different po,rts, and these only.

It is, however, first necessary to say a few words on the general question of
conduction in the nerve fibres of the cord.

Section 2. — Propagation of Impulses by the Fibres of the Cord.

It has been maintained that the nerve fibres of the spinal cord do not respond to
direct stimulation, in the same way that the fibres in a mixed nerve do,* and that it
is the cellular elements alone which respond to excitation, the resultant nerve impulses
being therefore, according to this view, indirect. The experiments to be detailed in
this and the succeeding chapters will give convincing proof, if that is necessary, of the

* Van Deen, ' Nederl. Tijdsclir. v. Geneesk.,' vol. 3, p. 393. Moleschott, ' Unters. z. Naturl./ 1860,
vol. 7, p. 380.

ON THE MAMMALIAN NERVOUS SYSTEM. 365

orroneous character of this view. The nerve iibres in the cord respond to direct exci-
tation, like those in the nerve roots and nerve trunk ; that is to say, nerve impulses,
with tluir accompanving electrical effects, are propagated in both directions along
any continuous nerve fibres which may exist in tlie excited area ; but, in addition,
complications are introduced by the connection of a large number of the fibres with
nerve cells, this connection causing now a possible decrease in the total electrical
effects, presumably by the blocking of the path and the falling out of certain
impulses, now an increase, presumably by the awakening of cells which lie in the path,
and the accession of fresh impulses generated in these structures.

We will first state in general terras upon what facts rests our present knowledge
of the mode of propagation of nerve impulses, from an excited area of the cord along-
continuous piaths to a distant unexcited area.

(«.) It has been already pointed out in the historical introduction that the
methods of histological investigation, particularly those associated with the presence
of developmental and degenerative changes, have unravelled from the skein of nerve
fibres in the cord certain tracts, and grouped them into columns of a continuous
character in the lateral and postei'ior regions respectively. The limits of our
knowledge have already been alluded to, but the inadequate character of the method
is shown by the large number of fibres which are displayed in each transverse section
of the cord, and the comparatively small number as to which a continuous connection
with other parts of the cord has been demonstrated.

{!>.) When we tiu'u to the results of physiological experiments, the only method
which has furnished satisfactory indications of direct physiological continuity in a
tract of nerve fibres is that employed by Woroschiloff, Schiff, and others, of
exciting the peripheral end of the cut cord below the medulla and observing the
muscular movements of the lower limbs, in one case with the lower part of the cord
intact, in another with a section of some structurally known column. In this way
it has been shown that a group of fibres in the lateral column forms a path of such
direct continuity between the seat of excitation in the cervical region and the lumbar
cord that its section interrupts the passage of the descending nerve impulses gene-
rated in the former region, and it is therefore inferred that this path is, physiologically
speaking, directly continuous.

(c.) Another method of determining the character of propagation in the spinal paths
is the classical method employed by Helmholtz in the case of the nerve trunks, that
is, the measurement of the transmission time. This, in the case of the fibres in the
nerve trunk {Frog, Kabbit), is generally held to be about 33 raeti'es per second,
though in Man the conduction along sensory fibres is stated to be twice as fast.
The experimental determination of this latter is, however, complicated by the
methods used, which include the measurement of the individual reaction time, and
inti'oduce, therefore, additional uncertain factors which blur the clearness of the
results.

3^fi MMSSRS. V. OOTCH AND V. IIORSLEY

Sliiiihir attoiiipts have been made, to (loteriuinc the rate of eoiuhiction of nerve
iui]iul.srs ill (he s|iinal cord, both centriAiffal and centripetal in character. Most of
the experiments liave been made upon Man l)y incnsnrint^r the reaction time, and are
therefore more or less mil rustworthy <n\ing to the conditlnns just mentioned. ExNER*
obtained thu.s a mean result for centrifuo-al impulses of II to 12 metres in 1 second,
for centrijietal im])ulses of 8 fiietres. On the other hand, von WiTTirn,t by the use
of a similar method, had previously obtained a rate of 26 metres for the centrifugal
impulses. We have been unable to find any description of observations upon the Cat
with regai-d to conduction, though such might easily be carried out by the use of the
graphic method, and since it was important for us to know the behaviour of the
fibres in the cord in this respect, we devoted a few preliminary experiments to the
determination of conduction time only.

These experiments may be liriefly described as follows : —

The cord was exposed in the antesthetised animal in the lower cervical and lower
dorsal regions respectively ; the rectus femoris muscle was then selected for graphic
record. The advantages offered by this muscle are —

(1.) Its anatomical relation with the pelvis, enabllngthe leg and trunk to be fixed,

and the muscle brought out at right angles to the body.
(2.) The ease with which it could be separated from the surrounding parts.
(3.) Its own structure — that of a long thin muscle, with parallel fibres.

The lower tendon of the muscle was divided and ligatured ; the ligature was then
attached by means of a pulley to the lever of Tigerstedt's break key,| so adjusted
that the smallest contraction of the muscle was sufficient to raise the lever, and thus
break an electrical contact. This contact formed part of an independent circuit,
including one of Smith's new chronographs § and three storage cells ; it was ascertained
that the movements of the chronograph armature occurred within 3/10000 second of
the break of the circuit.

The movement of the chronograph lever recorded upon the travelling glass plate of a
spring myogra})h {Federmyographion), the rate of movement being 25 centims. in
1/100 sec. The lateral column of the exposed cord was excited by a single
induction shock, which was obtained by allowing the traveller to break, at a given
point of its course, the primary circuit of a Kronecker's inductorium. As the
moment of break was always the same, the break induction shock obtained was
uniform in all cases, both as to intensity and time of occurrence ; the position of the
secondary coil was 2000.

The following measurements were obtained of the duration of the period between

* ExNER, PrLUGEK'.s ' ArcMv," 1873, vol. 7, p. 632.

t V. WiTTiCH, ' Archiv f. Pathol. Anat.,' 18G9, vol. 46, p. 476.

J Log. cit.

§ Loc. cit.

ON THE MAMMALIAN NERVOUS SYSTEM.

367

excitation and commencement of muscular response, when the former occurred at the
level of the 5tli cervical and L'nd lumbar nerves respectively : — ■

Cat (163).

Excitation at
6th cervical.

Excitation at
2nd lumbar.

Difference.

•0217
■0210
•0217
■0241

•0170
■0170
•0172
•0200

■0047
•0040
■0045
■0041

Average '0221

•0178

■0043

The distance between the 5th cervical and the 2nd lumbar was found to be
17 centims., and since the average difference of time between the muscular response
evoked by the excitation at the two regions is "0043, this difference indicates that the
cord delay is such as would be caused if the nerve impulses starting from the cervical
region travelled along the cord at 39| metres per second. This experiment, therefore,
seems to show that the time occupied by the conduction of nerve impulses in the
Mammalian cord closely resembles that occupied by their conduction in nerve trunks ;
and it confirms the view that there is an efferent path in the cord leading from the
cervical to the lumbar region in which a physiological continuity exists similar to tiiat
which forms tlie basis of nerve conduction in the fibres of mixed nerve trunks.

It will be noticed that the more exact physiological methods just x'eferred to rely upon
the observation of muscular movements. This involves serious disadvantages, since,
in the first place, the method is thereby limited to the efferent fibres, and thus the
physiological scope of any inquiry is narrowed ; whilst, in the second place, the struc-
tural connections between the efferent fibres in tlie cord and the anterior roots of the
nerves is to a great extent unknown, and certainly involves cellular elements, thus
introducing new physiological conditions.

There is, as it seems to. us, only one line of experimental enquiry which is free from
these objections— that, namely, of ascertaining the existence of the nerve impulses in
the fibres of the spinal cord itself, through careful quantitative observations of the
electrical effects by which they are accompanied.

The present chapter will be devoted to the consideration in detail of the results of
such observations in the dorsal and lumbar regions of the cord respectively. The
novelty of the method, and the important bearing which the results have upon the
physiology of the cord, we trust will warrant this extended treatment.

3G8

MESSRS. F. GOTCH AND V. IIORSLEY

Section 3. — General Chai!Actki{ ok tiik Elkothical Effects in the Cot^d.

The earliest experiments we made upon the coixl showed lliat pronounced electrical
changes always occurred in any portion of the lower dorsal and lumbar regions when
some other portion in continuity was excited electrically. Tlic simplest method of

Fig. 10.

excitation was that of inserting needles into the cord, the needles being connected
with wires attached to the secondary circuit of the inductorium. It is obvious that if
the excitation was carried out upon the cord whilst- still in connection with the
cei'ebrum, any resultant effects in the former would be the sum of the direct excitatory

ON THE MAMMATJAN NERVOTJS SYSTEM. 3f59

changes and tliose indirectly produced by the reflex discharge of the cortex, :ts
described in the preceding chapters. To avoid this and to obtain uncompUcated
eflTects tlie cord Avas always severed from its connection with the higher centres.

The pLan followed in the experiment was, therefore, as follows : two situations
were selected, in almost all cases about the level of the 8th dorsal and 2nd to
3rd lumbar vertebrae ; the cord was then first exposed for a short distance at the
upper one and divided. It was then carefully exposed in the lower region and
divided again, the upper division being made first in order to diminish, by cutting
off all connection with the higher centres, the shock which subsequent operations
necessarily involved. The portion of cord (see fig. 10), included between the upper
or 8th dorsal section and the lower or 2nd lumliar one, formed thus an isolated
fragment, and since it was upon this portion that the experiments to be detailed
were carried out, this fragment may be termed the " experimental region " of the
cord. A portion of this experimental region was always prepared for observation at
either its dorsal or lumbar end, the preparation consisting in the exposure of from
4 to 5 centims. of the cord in immediate connection with the cross section and the
di\^ision of all I'oot and other attachments. The cut end of the portion was then
ligatured, and the prepared part raised from the canal and suspended in air, care
being taken to avoid all undue pull upon the structure. As previously described,
(see Plate 29) the trunk was immovably fixed by holding the vertebrae in the ivory
jaws of a clamp, which was rigidly attached to a metal rod fixed into the experi-
mental table.

The ligatured end of the cord thus freed was attached by the before-mentioned
cables of thread soaked in "6 per cent. NaCl solution and plastered with kaolin, to the
galvanometrlc non-polarisable electrodes, one cable being tied round the ligatured end
and cross section, the other round the longitudinal surface about 1 centim. distant.

Tiie usual resting difference between the surface and cross section was found to
be present in all cases, and was always kept carefully compensated ; its characteristics
have been fully gone into in Chapter IV., and it will therefore not be further alluded
to here.

If the other exposed end of the experimental tract were now excited, either by
thrusting through it a pair of needle electrodes, or by placing upon the cut end the
points of a pair of ordinary platinum electrodes, an excitatory electrical effect was
seen in the portion connected with the galvanometric electrodes. This effect is
always of such a character as to be opposed to the I'esting difference ; it is suddenly
developed with the commencement of the stimidation, and subsides on its cessation,
being followed by the rise in the resting difterence which is described in Chapter IV.
It is evidenced both in the galvanometer and the capillary electrometer, the amount
of deflection of the needle and of movement of the mercury being dependent upon
(1) the nature, intensity, and number of the successive stimuli, (2) the condition of
the animal.

MDCCCXCr.— B. 3 b

•■^70 MESSRS. l'\ COTl'IT AND V. IIOUSLRY

(1.) lu'hdum tifflw n/fccf U> the Stimnlas.

A. Nature of Stimulus.

(a.) Single Induetioii Sliock. — The effect can be evoked by a single .stimulus, but in
that case gives only slight deflections in the galvanometer, but appreciable movements
in the electrometer. Thus, in one experiment (Cat 122) in whiuh the dorsal end of
the experimental tract was connected witli the non-polarisablc electrodes and the
lumbar end with needle electrodes, on excitation by a single break induction shock
the electrometer showed a sudden transient movement of the mercurial meniscus of
considerable size, seven divisions of the eyepiece scale ; with the single make shock,
a movement of five divisions was observed, both movements being opposed in direc-
tion to that produced by the resting difference.

{!>.) Repeated Induction Shocks. — When the cord is excited by a rapid succession of
equal and alternately directed induction shocks, more pronounced electrical changes
are produced in the cord ; the amount of the galvanometric change becomes very
appreciable, amounting in some cases to two or three hundred scale, the amount of
the effect in both the galvanometer and the electroineter varying with the intensity
and du'ection of the successive stiniTili. The best general idea of this effect is
obtained with the latter instrument. The mercur}" is seen to leap at the moment of
excitation in a direction opposed to that of the movement previously due to the
uncompensated resting difference, this sudden change of level amounting to from 10
to 15 divisions of the eyepiece; it then slowly continues to rise until the successive
stimuli cease, when a rapid subsidence to rather below its original level occurs.

(c.) Mechanical Stimulus. — If with one end of the experimental tract in connection
with the electrodes the other end receive a sudden mechanical stimulus, an electrical
effect is evidenced in both galvanometer and electrometer. The most effective
mechanical stimulus is that of sudden com2:)lete division or squeezing of the end of the
cord ; for this purpose ivory scissors were first used, but we afterwards found that if
insulating precautions were taken, sharp metal scissors could be employed with more
advantage, since the keenness of the blades ensured a clean cut and avoided the
dangers due to dragging on the cord. Such a division produces a small deflection of
20 to 30 scale in the galvanometer, and a pronounced movement of from 5 to 7 (eye-
piece divisions) of the mercurial meniscus.

After injection of strychnia, the slightest mechanical irritation of the cord or
sensory impression evokes marked electrical effects in the galvanometer and electro-
meter.

B. Intensity and Number of Stimuli.

Other things being equal, the cord electrical effect varies directlj' with the intensity
and number of the successive stimuli, a limit being reached in this respect, the change

ON THE MAMMALIAN NERVOUS SYSTEM. 371

being conditioned in the same general way as is the well-known electrical effect in the
excited Frog's nerve.

As far as the galvanometer Is concerned, since the falling time of our instrument
was 10 seconds, the effect for successive stimuli kept up for less than 10 seconds must
be obviously directly proportional to the time, hence, in all exact experiments involv-
ing this instrument, it was essential that the duration of the period of stimulation
and the number, therefore, of the successive stimuli used should be the same. This
was effected by the revolving paraffin key described in Chapter III., which ensured
a strictly uniform period.

(2.) Condition of the Animal.
A. AncBsthesia.

The character of the effect in the cord, unlike that in the nerve, is varied not only by
the intensity of the stimulus, but also by the introduction of changes in the condition
of the animal. These are connected with the awakening of the central cellular
elements of the coi-d through the intensity of the stimulation used. The condition
of the animal largely affects the limit of intensity at which any stimulus becomes
adequate to arouse these corpuscular elements. Thus if profoundly aneesthetised a
strong stimulus is necessary, but if the narcosis be but slight, a weak stimulus may
evoke the result.

Although the subject of anresthesia has been already referred to in Chapter III.,
it has such an important bearing on the present results that it must be reintro-
duced at this juncture. Our experiments abound with instances of the following
character : in a Cat (371) the lower (lumbar) end of the experimental tract was
connected with the galvanometer, and the lateral region of the cord in the upper
dorsal section was excited with a series of 500 successive equal and opposite induction
shocks (100 a second for 5 seconds). The galvanometric effect obtained was a
deflection of 230 scale, the anaesthesia though complete as regards consciousness
being of a comjiaratively slight character. The ether was now pushed and the
anaesthesia made more profound, abolishing reflex movements, when the effect of a
precisely similar excitation was indicated by I 42.

In another animal (Cat 875) electrical effects similarly produced in a state of
profound and slight ansesthesia show, when compared, the same differences, both with
minimal and maximal intensities of stimulus.

3 B 2

372

iMKSSRS. V. C.orCII AND V. IIOIISLEY

Cat (375)

Vrnfomul ana3stliesia.

Slight ansesthesia.

(Stimuhis 500 minimal) .

60 galvaiiornctric scale degrees

190

82

140

98

215

(Stimulus 1000 maximal) .

192

243

273

295

165

210

295

410

Thus the degree of anaesthesia in which the animal happens to be influences the
result very considerably, and is a convincing proof of the true physiological basis of
the cord electrical effect, viz., that it is dependent upon the number and intensity of
excitatory impulses in the observed region. It may be pointed out once more that
this influence of ansesthesia is also suflicient to show, what might perhaps be otherwise
suspected, that there is no objection to the use of the electrical method of stimulation
through its possibly involving errors due to electrical escape.

Further proof that with the method of isolation used no such escape occurs is
shown by the similarly directed efiect evoked by both the make and break induction
shocks, and by the production of the effect by mechanical stimulation. A still more
convincing proof is, however, the complete disappearance of the effect on systemic
death. (See Chapter III., section 4.)

B. Sijste)iiic death and injury.

The influence of systemic death upon the physiological condition of the spinal cord
has been already referred to in connection with the resting difference. It was there
stated that the diff'erence keeps up and increases in the exposed portion of cord as
long as it is in connection with a part which through the maintenance of an unim-
paired circulation retains its normal state of nutrition. When systemic death occurs
the difference immediately begins to fall, and in a very few minutes (2-5) the
excitatory electrical effect disappears, the loss of excitability occurring in the case of
the spinal cord with much greater rapidity than in the case of the sciatic nerve.

This disappearance occurs without systemic death if by any movement of the animal
the cord is pulled upon, or if it is bruised in preparation at a point intervening between
the seat of excitation and that of observation. Finally, the functional endurance of
the tissue is dependent upon the animal used, differing in different species and in
the different animals of the same species.

The most important general conditions by which the cord electrical effect is con-
trolled, having been thus set forth, we pass on to consider what information a detailed
examination of the characters and amounts of the electrical effects evoked under
different circumstances, furnishes as to the structure and functions of tlie spinal cord.

ON THE MAMMALIAN NERVOUS SYSTEM. 373

In our earlier experiments we made use of the capillary electroioeter, and we will,
therefore, first briefly describe those made with this instrument.

Section 4. — ExcrTAXORY Electrical Effects in the Cord Investigated by
Means of the Capillary Electrometer.

The electrometer, owing to the rapidity with which the mercury moves, furnishes
valuable information as to the alterations in character and amount of electrical
changes which follow one another in rapid succession. We have already indicated
the experimental advantages which this confers in the examination of the cord effects
evoked by cortical stimulation.* Since, however, the conditions which increase the
sensitive characters of the instrument are to a great extent those which diminish its
rapidity, it is almost impossible at present to obtain an instrument of suflicient
sensibility for our purpose, without so slowing its movement that it takes more than
Yj second for the mercury to complete its rise or fall. If, therefore, a series of transient
electrical changes similar in du'ection, following one another at intervals of less than
Ys" second and all of equal intensity, are allowed to affect the instrument, then since the
movement of the meniscus due to the first change would take ^t second for its comple-
tion, the second change will occur when the mercury is either in movement or has just
completed its excursion and not returned ; a second movement is thus super-imposed
on the first ; so a third on the second, until a limit of fusion is reached, this being
dependent upon the fact that with each additional displacement the counter-pull of
the surface tension increases, and finally the displaced mercury is maintained at a
new level without any additional displacement perceptible on either side of the level
attained. It is probable that an extremely fine vibration synchronous with the rate of
successive electrical effects exi.sts, but when the successive electromotive changes are
uniform in amount, direction, and time relations, such a vibration must be extremely
small.

A very different condition is introduced when these electrical changes are alternate
in direction, since now each displacement by one effect, whilst still in progress, is
counteracted by another in the reverse direction due to the succeeding change being of
opposite sign to its predecessor. The effect of such a series of electrical changes even
when following one another at such short intervals as yg-g second is thus clearly visible,
whatever level the mercury may have reached, as a blurring of the edge of the
meniscus. The appearance to the eye may be described as a grey border to the
otherwise black opaque column when viewed under the microscope.

This peculiarity of the electrometer at once enables us to judge whether, in a series
of brief electi'ical changes, these are similar or dissimilar in direction, and we will
first draw attention to this point in connection with the electrical effects in the
spinal cord.

* See this paper, p. 324, also ' Roy. Soc. Proc.,' Nov., 1888 (vol. 4.5, p. 18).

374 MESSRS. F. CIOTCII AND V. HOllSLEY

The present series of experiments Ix'ing planned for ])urposes of quantitative
measurement, the least complicated and most constant conditions were in all cases
selected, these being associated witli careful isolation, and with the presence of one
electrode only in contact with the natural surface. It is probable that witli these
conditions the total excitatory electrical change in the electrode circuit is chiefly that
produced by alterations affecting this one electrode.

We will now proceed to describe the character of the movements of the mercury
when, with the capillary electrometer connected by the electrodes with an isolated
portion of cord, this structure is stimulated by a series of induction shocks alternate
in direction and following one another at intervals of y^o second (Helmholtz side-wire
inductorium).

It has already been stated that one end of the " experimental tract " is prepared
for connection with the non-polarisable electrodes, the other for excitation ; so far as
the character of the effect revealed by the electrometer is concerned, it makes no
appreciable difference which end is respectively used for the purpose. In both cases
a minimal excitation evokes an electrical change in the other end of the tract which
affects the mercury of the electrometer, so that it moves rapidly up to a certain point,
and there remains steady without visible oscillation, while it falls on the cessation of
the stimulus, the mercury rapidly returning to its previous resting position.

With more intense stimuli a larger excursion of the mercury is obtained, and
although no evidence of rapid vibration is detected, the character of the movement
often becomes irregular, rising and falling at intervals in a more or less abrupt
manner.

Whilst then it was clear from its character that the movement of the mercury is in 7io
way connected with any electrotonic or other escape from the exciting circuit, since
there is no evidence of rapid rhythmical alternating effect, synchronous with the
number of stimuli, it was desirable to ascertain to what extent the irregularities just
referred to were true indications of changes in the cord.

To ascertain this, experiments were made in which (see fig. 11) the cord was divided
at one point only, the 8th dorsal, and the upper end of the lower fragment connected
with the electrometer. The rectus femoris muscle was then prejiared as indicated in
the preceding paragraphs dealing with the transmission time (p. 36G), but the attached
ligature was fixed to a strong spring (Fick's isometric myograph), the movements
of which, as recorded by a lever, were magnified 50 times. The cord was then exposed
in the lumbar region and its lateral surface excited, the muscular contractions were
observed and recorded, and at the same time the character of any displacement of the
mercury of the electrometer was ascertained as far as possible by the eye, and roughly
di'awn upon paper.

It was thus ascertained that the electrical effects in the dorsal region of the cord
evoked by lumbar excitation, and evidenced by the mercurial movements, coincided
in plan and character with the muscular effects.

ON THE MAMMALIAN NERVOUS SYSTEM.

3:5

Since the electrical changes evoked by excitation of the cord thus appeared to afford
true indications of the passage of nerve impulses along its nerve fibres, it seemed to
us quite possible to obtain for any one set of fibres a measurement of the value of the
electrical effects in it which would bear comparison with a similar measurement

Fin-, n.

obtained in the case of some other set. By piecing together the information thus
obtained, and by varying the conditions in which the cord was placed, we hoped to
determine the comparative number or at least resistance of the channels along which
excitatory impulses are propagated from the excited to the observed area. Further,
by means of carefully planned interruptions, we'hoped to ascertain how far such iuipulses

370 MESSRS. F. GOTCH AND V. IIORSLEY

were limited, presumably by the anatomical relations of the jjaths along which they
were conducted, to given tracts of fibres in tlic diflerent columns. We thus designed
our work to ascertain the scheme upon which the filnxnis structure of the cord was
arranged, and this for both ascending and descending impulses.

We found that the electrometer was not a suitable instrument for this investigation,
since with minimal stimulation its iriovements were too small to admit of sufficiently
accui-ate estimation, whilst the uncertainties connected with the conditions regulating
its sensibility rendered exact comparisons of small differences between different sets
of experiments impossible. We therefore used the galvanometer in all the remaining
experiments, obtaining with it results which could l)o better relied upon for purposes of
comparison.

Section 5. — Electrical Changes Peoditced by Localised Stimulation of the

Columns of the Cohd.

Plan of Experiments.

In order to obtain quantitative comparisons between the effects evoked by localised
excitation of different columns in the cord, the plan adopted was to divide the cord in
the mid-dorsal and lumbar regions, and prepare one end of the isolated experimental
tract lying between the sections for connection witli the galvanometer electrodes and
the other for excitation. The method of connection has been already fully described,
namely, at the cross section and the surface 1 centiiii. away, by means of cables,
under all precautions previously indicated. The preparation of the other end for
excitation was effected by removing about a centimetre from the end, and thus
exposing the cross section of the cord, in which it was possible, using the precautions
described in the chapter on operative procedure, to stimulate, with a pair of fine
platinum electrodes, the section of each column thus exposed, anterior, lateral,
posterior, or the grey matter.

The experiments showed us that with moderate intensities of stimulating currents,
electrical effects were always produced in the observed end of the experimental tract,
when at the other end the cross section of either lateral or posterior column was
excited, but that only very small and rare effects were caused by a similar excitation
of the anterior columns or of the grey matter. This suggested that no continuous
strand of fibres united the excited portion of these latter tracts with the observed
portion. We shall refer to this result again later on ; since, however, the stimulation
of the lateral and posterior columns alone gave notable effects, we limited our excita-
tion in the majority of instances to these columns.

The majority of these experiments were carried out on Cats (17 animals), but we also
made a considerable number of experiments on three large Macaque Monkeys. It will
be presently seen that the results differ in the two animals in a most interesting way.

ON THE MAMMALIAN NERVOUS SYSTEM. 377

The experiments all fall into two great groups, distinguished by the ftict that in
one the lutobar end of the isolated fragment of cord was the seat of galvanometric
observation, whilst, in the other, the dorsal end was observed and the lumbar end
excited. In the first case the electrical changes are obviously due to the discharge of
nerve impulses down the cord, in the second to their discharge up the coi'd.

We will now proceed to a detailed desciiption, first of a typical experiment, and
then of the results obtained from several experiments under these two opposite con-
ditions, and we will take first the case of impulses descending the cord, these impulses
having been produced by excitation of some one column, as displayed in the dorsal
section of the " experimental region " of cord. The galvanometric connections having
been made with the lower lumbar end of this tract, electrical changes were evidenced
in it when the descending impulses reached that part which was in connection with
the electrodes leading to the instrument.

Section 6. — Electrical Effects evoked in the Lumbar Cord by Excitation
OF the Columns in the Dorsal Region.

(a.) Typical Experiments.

Since the results in the case of Cat's cord, as mentioned in the preceding paragraph,
are diff'erent from those obtained with the Monkey's cord, it is desirable to separate
out the experiments made upon each. We will therefore describe a single typical
expei-iment and its results in the case of these animals. In order to avoid repetition,
the general procedure, which is the same in both animals, will be described more
minutely in the case of the Cat (this being taken first) than in that of the Monkey.
To make clear the relationship of the excitation to the galvanometric connections,
the actual positions in a typical case of the seat of operation and of both sets of
electrodes is given in fig. 10, p. 368, which represents the spinal cord of an animal
thus experimented upon.

The cord of the Cat (243) was divided at the level of the 7th dorsal vertebra; it was
then exposed in the lumbar region for 3 centims. and divided at the level of the
3rd lumbar vertebra, thus isolating an experimental region which extended between
these two levels. The distal (lumbar) end of this isolated fragment was now ligatured,
and all connections having been divided, was raised in air by the thread. By means
of cables the transvei'se section was connected with one non-polarisable electrode, the
longitudinal surface at 1 centim. distance by a similar cable with the other. The
electrodes were fixed, as before, on a stand, and arranged at such a height that a con-
siderable length, two or three inches, of thread cable hung loosely between them and
the cord, all error due to mechanical displacement, as in the cortical experiments,
being thus guarded against.. The upper end of the isolated fragment or experimental

MDOCCXCL — B. 3 C

378 MESSRS. F. flOTClI AND V. IIORSLEY

tract of cord was now exposed for about a centimetre, and a fresh division made so as
to expose a more excitable section for stimulation.

The portion of cord investigated showerl, when connected with the galvanometer,
the usual resting electrical difference between the surface and ligatured cut end (see
Chapter IV.).

When after compensation the galvanometer needle was steady, the exciting elec-
trodes, guarded by a short-circuiting key, were carefully held by one observer against
the cut end of one column of the cord, as described in Chapter III., every precaution
being taken to ensure that the surface should be dry. The key was now opened by
the observer at the galvanometei', and by means of the revolving paraffin key the
column was stimulated for a definite time (3| seconds in this case), by a series of weak
induction currents (100 per second) produced by the usual magnetic interruptor, and
made equal and alternate by means of the Helmholtz side-wire. The intensity of the
exciting currents had to be adapted to the varying condition of the animal, etheri-
sation, &c., but the results aimed at being to stimulate the fibres in the separate
columns rather than to arouse the reflex activity of the cord, it was thought advisable,
at any rate in the commencement of an experiment, to use currents only just sufficient
to evoke definite effects; such excitation will be termed here "minimal."

In this particular case one Daniell cell was used in the primary exciting circuit and
the secondary coil stood at 500. The excitation of either anterior column produced
no electrical change ; that of either lateral produced a very distinct effect, of the usual
excitatory character, that is, an efiect which commenced with the stimulation and did
not persist after its cessation.

The effect indicated the establishment of a transient electrical state, opposed in
direction to that of the original resting difference. The galvanometer deflection, which
recorded the change, was very distinct in its commencement and termination, so that
its amount could be determined with accuracy. The deflection was 41 when the left,
and 57 when the right lateral column was excited. The stimulation, in exactly the
same way, of the cut end of the posterior columns produced much more pronounced
effects, amounting to 96 in the case of the left, and 94 in that of the right column.
Every precaution was taken to ensure that as far as possible the degree of narcosis, &c.,
should be the same during the four stimulations, and an interval of 1 minute was
generally allowed to elapse between the successive applications of each.

When a stronger stimulus (double the strength, coil 1000), of precisely the same
character and duration was employed, larger effects of the same kind were produced,
stimulation of the anterior regions being again followed by only a mere trace of
effect, that of either lateral by well-marked deflections (left 105, right 140); and that
of the posteriors by very large deflections (left 290, right 208).

It will be seen that this relationship of magnitude of effect is one which is retained
in all the results obtained in this way by differential columnar excitation of the
divided coi-d of the Cat.

ON THE MAMMALIAN NERVOUS SYSTEM.

379

A very different relationship is, however, found to exist when the spinal cord of the
Monkey is investigated in a similar way.

The spinal cord of a Macaque Monkey (232) was divided at the 10th dorsal vertebra
for excitation, and also at the 2nd lumbar, where having been freed from its nervous

Eig. 12.

Macacus Rhesus

Adult. 9

[-fT^sTOTTScjcreo^-

and other attachments, it was ligatured near its divided end, raised as in the preceding
experiment, and by means of cables connected with the leading off uon-polarisable
electrodes at Its ligatured end and at its surface (see fig. 12).

The usual resting electrical difference was observed when the electrodes were
connected with the galvanometer, and was compensated. Each column was now

3 c 2

380

MESSRS. F. GOTCII AND V. lIORSLEr

excited where it had been exposed in the dorsal cross section, the excitation being as
ill the preceding case, viz., that of the interrupted induction currents ( I (JO per second)
with the lielmholtz side-wii-e, and the period of" excitation being 3^ seconds.

As in the Cat, the excitation of eitlier anterior column produced no effect ; that of
each lateral, however, evoked a marked electrical effect, the two laterals giving
respectively, left 75 and right 62, the deflection being in each case opposed in
direction to that of the resting ditl'erence. Excitation uf eitlier posteiior column
was followed by a much smaller deflection than the corresponding lateral, the left
giving 34, the right 40.

Fig 13

200 1 1-

[PCi

lOU

ic,n

ItU

y

leU

100 —
80 —

■■1

1=1

B

60 — -

40 —

20 —

0 —

1

SB

■

r

■

r

1

ANT I

d

.AT POST LAt|pOSTA

Cat

ow/p cord

\. U. L. L , r\. I\, MX.

NT ant|lat POSTiAf post|ant|

Monkey
down cord

Wlien a stimulus of double the strength (coil 1000) but of the same duration was
employed, very large deflections were produced, but the relationship of the results
still held good, as excitation of the left lateral gave 210, whilst that of the corre-
sponding posterior gave 165, and similarly excitation of the right lateral gave 140,
whilst that of the corresponding posterior gave 120.

If the comparative value of the electrical change produced in the cord by " minimal "
excitation of the different columns be represented graphically, as in the annexed figure,
a glance is suflBcient to show the striking contrast between the result obtained in the
Cat and that obtained in the Monkey, since in the latter case the laterals pre-
dominate over the posteriors. (See fig. 13.)

ON THE MAMMALIAN NERVOUS SYSTEM.

381

(b.) Average of Experiments.

The two experiments just described are types of all those made upon normal
animals. The following table gives the results of the readings obtained in four Cats
and one Monkey, with stimuli which were varied in one particular only, that of
intensity.

It will be noticed that the average in the Cat of the deflections produced by
excitation of the posterior columns is twice as great as that following stimulation of
the laterals.

In the Monkey, on the other hand, the average of the stimulation effects obtained
from the lateral is about half as large again as that produced by the excitation of the
postei'ior columns.

Excitation of Cut Dorsal Cord (Peripheral Surface). Electrical Effect in

Lumbar Cord.

Region
observed.

Region
excited.

Intensity of

stimulus.

Electrical effects evoked by stimulation of

Cat (243) . .

Left.

Right.

Ant.

Lat.

Post.

Ant.

Lat.

Post.

2 Lumbar

8 D.

500 min.

trace

41

'.)6

trace

57

94

„

1000 max.

105

290

140

208

, (351)

1 ,-

7 D.

250 min.

33

70

55

62

1 j»

»>

))

)>

12

39

42

52

) J)

600 max.

90

170

, ,

, ,

> >1

))

J)

H

31

70

90

105

» )1

ji

))

)J

56

110

91

115

, (23G

13 Dorsal

10 D.

500 min.

30

130

65

150

, (37',)

3 Lumbar

48

86

41

123

» )J

i»

48

92

32

105

1 »)

;»

1000 max.

39

225

62

197

) )l

))

H

General

75

175

98

207

average

608

1553

773

1418

51

129

70

129

Monkey ('i.-^a) .

2 Lumbar

10 D.

500

75

34

..

62

40

)» 1)

)>

»»

i»

, ,

98

22

. ,

65

30

)» )1

»j

))

1000
General

210

165

••

140

120

avei

"age

128

74

•

89

63

382

MESSRS. F. GOTCII AND V. HORSLEY

A further result which an analysis of these effects shows, is the difference between
the average deflection in the case of the posterior and the lateral column with weak
minimal stimulation and maximal stimulation respectively.

AvERAGK effect evoked by stimulation of one lateral column.

Minimal stimulation . . .
Maximal ., ...

Cat.

Monkey.

42

87

75
175

Average effect evoked by stimulation of one posterior column.

Minimal stimulation . .
Maximal „ ...

Cat.

Monkey.

92

170

31

142

It is seen that in the Cat both minimal and maximal stimulation evoke effects
which are twice as large in the case of the posterior columns as in that of the
laterals.

In the Monkey the lateral column effect with minimal stimulation is, on the other
hand, twice as large as that produced by the stimulation of the posterior column, but
with a stronger stimulus this relationship does not hold, owing, possibly, to the
increased reflex discharge which excitation of the posterior column now evokes.

The difference between the results of excitation of the columns in the two animals
is, therefore, best marked with the weaker intensity of stimulus. It is evident that
this must depend upon the number of fibres which form direct connections between the
excited dorsal and observed lumbar region, along which fibres as constituting paths of
least resistance the nerve impulses are almost entirely propagated from the excited
area, when the fibres this contains are aroused by a weak stimulus.

This seems to us to afford an experimental proof of the relatively larger number
of continuous lateral column-fibres which must exist in the Monkey as compared
with the Cat ; this greater proportion may, in the light of the previous results
described under cortical stimulation (Chapter V.), be ascribed to the more complete
development of the fibres forming the pyramidal tracts.

The histological investigation of the fibres in the tract in the two animals seems to
support this view, there being apparently many more fibres in the pyramidal tract of
the Monkey than in that of the Dog, as determined by the degeneration method. It
is, moreover, to be expected that the fibres in question must increase in number
with the completeness of the differentiation of those cortical structures from which

ON THE MAMMALIAN NERVOUS SYSTEM. 383

they spring ; and it need hardly be pointed out that, wliilst minute localisation of
representation of movements on the cortex of the Monkey has been clearly demon-
strated, such perfect differentiation has not been found in the Carnivora.

We will now pass on to the consideration of the results obtained when the dorsal
portion of the experimental tract of the cord is connected with the galvanometer, and
the columns excited in the lumbar region, thus evoking impulses which, to produce
effects, must pass up the cord.

Section 7. — Electrical Excitation Effects evoked in the Dorsal Region of
THE Coed by Excitation of the Columns in the Lumbar Region.

In these experiments the cord was divided in two places, as desci'ibed in the
preceding paragraphs ; but, since the electrical effects in the dorsal end of the
tract thus isolated were to be investigated, the cord was prepared for several centi-
metres at the upper dorsal level. At the lower lumbar section the cord was prepared
by excision of one centimetre for purposes of excitation. The exciting and galvano-
metric arrangements were of the same character as before, their relative disposition
being sufficiently indicated in the annexed fig. 14.

(a.) Typical Experiments.

We will again first describe the results of an experiment sel3cted as a typical one,
and carried out on the spinal cord of the Cat (244), as follows . —

The cord was divided at the 7th dorsal vertebra and pi'epared for observation as
before described : it was then divided at the 2nd lumbal' vertebra for excitation.

The upper end of this experimental tract showed the usual resting electrical
difference between surface and cross section, which was compensated.

The cut lumbar sui'face of each different column was now excited for 3^ seconds by
the interrupted current, 100 per second (Helmholtz side-wii'e), and the galvanometric
effect observed.

Although the secondary coil stood at only 500, the preparation was very excitable.
Excitation of the anterior columns was followed by a slight but distinct effect, 22 with
left side, 28 with the right; these, however, did not occur on repetition, and are
exceptional.

On exciting the left lateral column, a deflection of 120 was observed, whilst on
exciting the left posterior column, a deflection of 215 was produced ; finally excitation
of the right lateral was followed by 100, of the right posterior, by 190.

The general relationship of magnitude of electrical change evoked in the coi'd of the
Cat by lumbar excitation of the cross sectional area of its different columns is thus
the same as that obtained by dorsal excitation.

384

MESSRS. V. GOTCH AND V. HORSLET

That is to say, whether the nerve Impulses proceed down the cord from an upper
excited to a lower observed area, or up from a lower excited to an upper observed area,
the anterior column fibres give no eft'ects or only small ones, the lateral give marked
effects, but the posterior give still more pronounced effects.

^

Fig. 14

j!

^ 5 Cat.

>i Adult 9

ii.

'-^ia229!)50i)i)i)PA>fflr

|IX.D

•

•1

;J -?9i?Q2o_ Ex

IVL l!

^ Mk.

im^^

U4'r

If we turn now to experinaents upon the Monkey's cord, the same contrast with the
results in the Cat appears as was described before. Thus in a Macaque Monkey
(271) the cord was exposed and divided at the 8th dorsal vertebra and was then
freed from its attachments, ligatured on the peripheral side of the section and raised

ON THE MAMMALIAN" NERVOUS SYSTEM.

385

from the canal ; it was then connected with the electrodes exactly as in the case of
the Cat.

The lower section and exposm-e occurred at the level of the 1st lumbar, and the cut
sui'face of the columns was then stimulated as in the preceding case. The necessaiy
sti'ength for minimal excitation was only found when the coil was 2000. The
anterior colxmins evoked no definite effect, but excitation of the laterals evoked
large effects, left 190, right 225, whilst similar excitation of the posteriors evoked
smaller deflections, left 112, and right 120.

Fig. 15.

■■■■■■ n

■■■■■■ ■!

■b: : B

■■"■ ■ ■!

■i ■ ■ ■!

■ ■ Br

■ I ■!

■ ■
■
■

B

■
_ ■

BB

L. L. R.IRJ L. L. R. R.

LAT POST LAt|p05t| UT P0S1(lAT POST

Cat Monkey

up coi-d up cord

The results of these two typical experiments will be rendered still more obvious
by the annexed figure, hi which the value of the excitatory effects in the Cat and
Monkey are displayed in a graphic form. It will be seen how completely the relative
size of the effect evoked by stimulation of the two kinds of column is reversed in
the two animals. (See fig. 15.)

(b.) Average of Exijeriments.

We will now take the average of the observations in which the condition of the
animal, &c., was, as far as possible, the same throughout, and the effect evoked in the
dorsal cord by stimulating the lumbar columns accurately noted.

MDCCCXCI. — B. 3 D

88r,

MESSRS. V. COTCU AND V. irORSTiKY

Excitation (tf'Cut riUinhar (lord (Central Surface). Electrical Eit'oct in Dursal (Jord.

Cat (197) . .

Rctrion
observed.

lU'ifioii
excited.

Inlonsity of
stimulus.

Klectriciil ullects cvo

<ril by SI

iiiiulul.iiin of

i

Left.

Right.

Ant.

Lat.

Post.

Ant.

Lat.

Post.

10 dorsal

1 lumbar

2000 max.

0

G4

136

0

32

165

Cat (244) . .

8 dorsal

500 „

ti'ace

80

120

trace

96

100

22

120

215

28

100

190

)J )1

0

126

120

14

66

100

, ,

160

230

. ,

126

140

Cat (256) . .

8 dorsal

3 lumbar

800 mill.

. .

66

76

. ,

20

90

Oat (365) , .

8 dorsal

2 lumbar

1000 max.

, ,

65

80

140

82

• 80

60

142

106

i» »

110

151

115

175

Cat (375) . .

9 dorsal

2 lumbar

500

145

126

. ,

145

78

250 min.

61

46

, ,

41

64

600 „

. 60

82

81

98

1000

, ,

192

273

165

295

1000

, ,

243

295

,

210

410

ly

)»

)».

600

••

190

140

••

68

215

1761

2138

• •

1545

2357

Average .

••

117

148

103

157

Monkey (233)

7 dorsal

1 lumbar

2000
2000

••

••

••

120
76

92
20

Monkey (271)

8 „

2000
Average .

••

190

112

225

120

••

190

112

140

77

It will be noticed that the general relationship holds in the case of impulses passing
up the cord, namely, that in the Monkey the most marked effects are evoked by
excitation of the lateral, in the Cat by excitation of the posterior columns. There
are, however, differences in the precise characters of the relationship as compared with
that indicated in the preceding group of observations when the nerve impulses were
descending. These differences are exhibited by separating all the lateral and
posterior column effects, in accordance as they were obtained with the minimal and
maximal stimulation, respectively.

ON THE MAMMALIAN NERVOUS SYSTEM. 3S7

Excitation of Lumbar Cord. Electrical Effect in Dorsal Cord.

Cat. — Average effect evoked in the Cat by stimulation of one lateral column.

Mimimal stimulation 55

Maximal „ 124

Average effect evoked in the Cat by stimulation of one posterior column.

Minimal stimulation 74

Maximal ,, 164

Monkey. — A vei'age effect evoked by stimulation of one lateral column.
Maximal stimulation 133

Average effect evoked by stimulation of one posterior column.
Maximal stimulation 86

The chief differences between these results obtained when nerve iropulses proceeded
up the cord and those obtained when they travelled down, are the following : —

(1.) The maximal effect in the Cat is more than twice as great as the minimal with
stimulation of both lateral and posterior columns when the impulses are ascending,
i.e., the lumbar end excited. On the other hand, when the impulses are descending
the maximal effect was less than twice as great. There is thus a gain in the amount
of the effect, when the stimulus is on the lumbar side of the region, in proportion
as the excitation increases in intensity.

(2.) The minimal posterior column effect with ascending impulses is larger than
the lateral in the Cat in the proportion of 3 to 2, but this proportion is less than that
obtained when the dorsal cord was excited and descending impulses evoked, since this
was 2 to 1.

(3.) In the Monkey the lateral column effect is larger than the posterior in the
proportion of 3 to 2, when ascending impulses are evoked by maximal stimulation ;
with descending impulses similarly evoked the proportion is 9 to 7.

The characteristic predominance of the lateral column of the Monkey is however
marked, whether the excitation is dorsal and the descending impulses investigated,
or lumbar as in the present case, and the ascending effects observed.

The differences above described seem to indicate that the generated impulses travel
not only along direct but indirect paths, that is along paths involving cells, and that
the structural relations and physiological effects of the interposed structures are such
that these influence, iii both animals, traversing nerve impulses differently according
as their direction is centripetal or centrifugal. It is particularly as regards the
lateral column effect that this influence of direction is most marked ; it would,
therefore, seem that the lateral column comprises among its fibres some (presumably
internuncial) which offer less resistance to ascending than to descending impulses.
This view will receive confirmation, and be again referred to in the later experiments
on the relations of the cord to the nerves. (See Chapters IX. and X.) It is sup-

3 D 2

388 MESSRS. F. GOTCII AND V. IIORSLEY

ported by the series of experiments to l)e next detiiiled, in which the result of
intervening localised section of columns upon the electrical effects evoked by excita-
tion of the same is set forth.

Section 8. — The Electkical Changes in each halk of the Longitudinally

Divided Cord.

The question as to what amount of the electrical effects, and thus of the excitatory
processes underlying these, when evoked by stimulation of one coknnn, is due to
changes confined to that cohnnn only, is one of such importance in respect to the
well-known views as to conduction by the different groups of fibres in the cord that it
seemed advisable to ap[)roach this subject by particular experiments. In the case of
the changes in the cord following excitation of the cortex, we had found that it was
possible to split the cord longitudinally without destroying the excitability of the
two halves, and so to obtain evidence of unilateral localisation of effect. This seemed
the most straightforward means of getting information as to localisation when the
columns of the cord were the seat of the excitation ; at the same time we felt that
the operation of longitudinal division must in all probability seriously affect the
functional continuity of the posterior columns, whose excitation in the Cat, as the
foregoing experiments have shown, is productive of such marked results. To these
experiments attention must now be given.

We have only performed two successful experiments ; both upon the Cat.

In the first the cord was divided at the 9th dorsal and at the 4th lumbar
vertebra3. The upper dorsal end of the fragment was prepared for excitation, the
lower lumbar end was freed and split in a manner similar to that employed in the
cortical experiments. (See fig. 16.) Each half was connected with a pair of non-
polarisable electrodes, as in those experiments, and either pair could be switched into
tbe galvanometer circuit and its electrical changes observed. The general arrange-
ment of the circuit in this case is shown in fig. 1.

When the columns were excited for 5 seconds with the secondary coil at 2000,
and the right portion of cord investigated, then —

Excitation of the right lateral produced an effect of 31,
„ left „ „ no effect.

On the other hand,

Excitation of tbe right posterior produced an effect of 53,
J) ); lett ,, „ „ 15.

The left poi'tion of cord was now investigated and the same strength of stimulus
employed. It was found that —

ON THE MAMMALIAN NERVOUS SYSTEM.

Excitation of left lateral produced effect of 42 ; 56
right ,, ,, „ (i 0

„ left posterior ,, „ 87

right „ „ „ 45

Fig. 16.

389

f. Cat.
I . AduLt I I

'D0T)3^

On repeating this experiment with a stronger stimulus (coil 3000), the unilateral
character of the right lateral column became less marked, whilst the left and right
posterior columns gave now equal bilateral results ; thus —

IVM

MESSRS. F. GOTCH AND V. HORSLEY

Excitation of left lateral prorliiced effect of G8

right „ „ „ 36

„ left posterior ,, „ 96

„ I'ight „ „ „ 89

Vig. 17.

f

T*

?^"i.c

-■

Cat.

Adult 9

-

•|. ra.D

.c^r^

s^

M ^

~ — xm.D

i

t^X

r^lJ '%\^

It tlius appears that when half the split cord is observed and the columns excited
on the central side, the excitatory electrical change, produced by weak excitation of
the lateral column is entirely confined to the side stimulated, whilst that produced by
similar excitation of the posterior column is but mainly confined to the stimulated
side, while when the excitation is of sufiicient intensity it is produced equally on

ON THE MAMMALIAN NERVOUS SYSTEM. 391

both sides. This, therefore, suggests that descending impulses in the lateral column
are, to a very large extent, confined to these fibres, but those in the posterior column
cross into similarly situated fibres on the opposite half of the cord.

A second experiment of the reverse kind was made on another occasion, the excita-
tion being now in the lumbar region, and the impulses evoked thus being ascending.
The cord was divided at the 7th dorsal and 3rd lumbar, and the upper dorsal portion
was prepared for galvanometric observation, and split longitudinally, whilst the
lower lumbar portion was prepared for excitation ; the general arrangement is shown
in fig. 17,

When the right half of the split cord was examined —

Excitation of the right lateral produced an effect of 38

„ left „ „ „ 3

,, right posterior ., ,, 40

left „ „ „ 14

The cord was then again divided a little centrally to the seat of the previous
excitation, so as to expose a fresh surface for the stimulation, and now —

Excitation of right lateral produced effects of 86 ; 94

left „ ., „ 16 15

,, right posterior ,. ,, 81 112

left „ „ „ 52 60

The left half of the cord when observed gave the following results : —

Excitation of left lateral produced effect of 33
„ right „ „ „ 3

„ left posterior ,, ,, 40

» right „ ,, ,, 7

The general conclusion to be derived from these experiments is that the excitation
of the lateral column produces an effect which is limited to a very remarkable degi'ee
to the side excited. On the other hand, the excitation of one posterior column
produces with appropriate excitation bilateral electrical effects, but the effect is twice
as great on the same side as it is on the opposite one.

It must, however, be borne in mind that the operation may, by depressing the
excitabihty of the posterior columns, be a source of error wliich can interfere largely
with the above results, since the posterior columns are more liable than the lateral to
suffer by the procedure of splitting, and earlier experiments have convinced us that if
they are injured the electrical change produced by their excitation is very largely
affected.

The extent to which the effects and thus the nerve impulses are localised in the

392 MESSRS. F. GOTCH AND V. JKJKSLKY

columns ex-cited was, therefore, now ;ij)proached l)y anotlier method, thai, namely, of
intervening sections.

Section 9.— Influence of Intervening Sections upon the Electrical Change
IN the Cobd following Excitation oi- tiiio Different Columns.

The extent to wliicli the electrical change in (lie cord following excitation of the
cut columns is dependent upon direct continuity of nerve fihres between the part of
the column excited and the part of the cord observed is more clearly indicated by the
experiments now to be described.

In these the excitatory and galvanometric arrangements were similar in all details
to those employed in the two preceding groups of experiments ; but the experimental
tract had been subjected to important additional operations, some particular column
indicated being divided at a position intervening between the region of excitation and
the observed region, and the result of such division as affecting the electrical excitatory
change being then estimated.

The section, the influence of which was thus investigated, was made in most cases
at the time of the experiment. In a few cases the section was made three or four
weeks beforehand, in order not only to ensure a more striking alteration of effect by
the degeneration of the continuous nerve tracts involved, but also to provide against
any transient disturbance due to the operation, affecting the excitability ho\h of the
particular tract and of other nerve tracts than the one operated upon.

Since the influence of each separate intervening section had to be studied in the
case of excitation evoking both descendmg and ascending impulses in the experimental
tract, the experimental results of any section naturally fall into the two groups
already indicated in the preceding Sections.

(].) Electrical Effects in the Lumbar Cord Evoked by Excitation of the Dorsal Cord.

A. Injiiience of Hemisection.

The experiments were made upon three animals (Cats), in all of which the inter-
vening hemisection occurred at the level of the 12th dorsal vertebi-a; in the first on
the right side, in the second and third on the left side.

Each exper-iment was conducted precisely as those previously described ; that
is to say, the cord was divided at the 8th dorsal and 2nd lumbar vertebrae. The
lower end of the experimental tract was prepared for connection with the galvano-
metric electrodes, the upper for excitation. The different columns were first succes-
sively excited before the hemisection in the usual manner, and the electrical effects
observed. These are shown in the preceding table, see page 381, and were of the
kind indicated in Section 6 (b). The cord was now exposed at the 12th dorsal, the
hemisection made and the electrical effects of the excitation of the different columns

ON THE MAMMALIAN NERVOUS SYSTEM.

393

under these circumstances were then noted. The following is a table showing
the results obtained : —

Electrical Effects in Lumbar Cord Evoked bj' Excitation of Dorsal Cord after
Intervening Hemisection on the Right Side at 12th Dorsal Vertebra.

Cat (351)

Resrion
observed.

Electrical effects.

Roerii Ti
exfit( d.

Stimulus.

Left.

Right.

Lat.

Post.

Lat.

Post.

2nd lumbar

7 th dorsal

500 min.

„ min.

„ min.
li)00 max.
2(X10 max.
2000 max.

45

e>2

43
100
201
180

71
45
65
81
145
145

0
0
0
0
6
12

0
0
0
0
26
31

Intervening Hemisection on Left Side at 12th Dorsal.

Cat (371)
. Cat (278)

Region
observed.

Region
excited.

Stimulus.

Electrical effects.

Left.

Right.

Lat.

Post.

Lat.

Post.

1st lumbar
3rd lumbar

8th dorsal
9th dorsal

500

1000

500
1000

18
18
15
49
29
0
19

0
0
12
26
26
0
34

132
126
110
201
184
47
148

56

75

58

193

182

56

198

The average results of the excitation of the lateral and posterior columns,
when grouped, as dependent upon the intensity of the stimulus, show the following
quantitative comparison.

Average Effects obtained by Exciting Columns on Side of Section.

Lateral

Posterior ....

Minimal
stimulation.

Maximal
stimulation.

8
2

23
29

MDCCCXCI. — B.

3 E

394 MESSRS. P. GOTCH AND V. IIORSLEY

Average Effects obtained by Exciting Columns on Opposite Side to Section.

Lateral

Posterior ....

Minimal
stimulation.

Maximal

stimulation.

74
62

183

17;J

From this talile it will be seen that the electrical effects observed in tlie dorsal
region are profoundly modified by an intervening hemisection of the cord. With an
excitation of minimal intensity, the application of the stimulus to the columns
at the dorsal end situated on the side of the section evokes little or no electrical
effect in the lumbar region. The interruption of the directly continuous fibres by
the hemisection has thus reduced the effect to such an extent, that in two Cats it was
completely abolished, and only present in one.

With the same minimal intensity of stimulus, excitation of the posterior colunui
evoked no effect at all, except in one instance ; the interruption in the fibres of this
column thus abolishes the electrical effect.

With a minimal stimulus, therefore, the electrical effect in the lumbar region
obtained by excitation of the columns in the dorsal region is dependent upon an
unbroken continuity in the fibres descending along the same side as the excited
column. It is thus clear, that not only are the stimulating induction cuiTents which
traverse the excited end of the exposed column confined to that side of the cord, but,
in addition, that the nerve impulses generated by this stimulus travel down that side,
since it is the arrival of the nerve impulses at the lumbar end of the tract which
produces the electrical effect and this is practically abolished by the interruption
offered by the hemisection.

Whilst the experiments thus furnish strong evidence against any direct and con-
tinuous tract of nerve fibres passing from one side of the cord to the other, the
presence of effects on the two sides of the cord, when evoked by more intense stimuli,
shows tliat the interruption of hemisection is not a complete block. With increasing
intensity it will be noticed that excitation of both lateral and posterior columns on
the side of section evoke electrical effects in the cord below. These must be due to
the passage of nerve impulses from the excited area along paths which are in relation
with cells, i.e., indirect, since these are the only known channels across the cord.
These impulses may discharge, in a reflex manner, the cellular energy of both sides
of the cord, the opposite side in a much less degree than the same side ; a crossed
electrical change would then indicate the sum of the nerve imjjulses evoked in the
opposite side by a discharge of nerve cells. Similar eflfects would, however, be
obtained if, without any reflex influence being present, an indirect path crossed from
the columns of one side to a column of fibres on the opposite side. Both agencies may

ON THE MAMMALIAN NERVOUS SYSTEM. 395

be presumed to operate in these cases, though the results of the experimental observa-
tions set forth in the succeeding chapters all favour the supposition that the fibres
which form the indirect crossed path are almost entirely derived from the posterior
column, whilst the interniuicial fibres, by which the successive segmental groups of cells
are associated with one anothei', are mainly confined to tlie lateral column. It will be
observed that the crossed effect is only |- to ^ of the direct effect evoked by stimulation
of the side opposite to the lesion ; hence either the nerve impulses, of which the effect
is the index, are either greatly reduced in quantity by their jjassage through the grey
matter, or compai-atively few nerve impulses take this crossed path.

Finally, as regards the effects in the side opposite to the lesion, it will be noticed
that, comparing the relative value of the two effects, due to excitation of the
posterior column is now less than that of the lateral column. This is probably due
to the necessary exposure of the columns at the 12th dorsal vertebra acting injuriously
upon the excitability and conductivity of the exposed but otherwise uninjured
posterior column not included in the section.

A glance at the full details of experiments 351 and 371 will show (see table,
Section 6 (6)), that the same relation was present in most cases before the hemisection
was made ; the subject will be referred to in more detail in the next paragraj^h, which
deals with the result of section and injury to the posterior columns.

B. Influence of Section of Posterior Columns.

The inffuence of the section of both posterior columns in the lower dorsal region
upon the electrical effect evoked in the lumbar region, and thus upon descending
nerve impulses, is shown by the following experiment : —

The cord was divided in a Cat (357) at the 9th dorsal and at the 2nd lumbal
vertebrse, it was prepared for excitation at the former region, for attachment to the
galvanometric electrodes at the latter. The cord was then also exposed at the 11th
dorsal vertebra, and, before any operative lesion, an experiment of the customary kind
was made. The excitation, with stimulus 500 of the two laterals, gave deflections of
30 and G5, average, 47 ; that of the two posteriors gave 130 and 150, average, 140.
After division of both posterior columns at the 11th dorsal, similar excitation of the
lateral columns evoked effects of 60 and 105, average, 82,* whilst that of either
posterior column evoked no effect whatever ; unfortunately, the effects evoked by
stronger stimuli were not observed.

The interruption, therefore, as regards the posterior columns was complete with
this intensity of stimulus, and hence all nerve impulses generated by such "minimal"
excitation of either column in the dorsal region are propagated to the lumbar region

* It will be seen that the average effect obtained due to excitation of the lateral columns after section
of the posteiior columns is increased. A similar exaltation occurs in the experiments detailed in
Chapter IX., Section 7, C, where the causation of the phenomenon is discussed.

3 K 2

396

MESSRS. F. GOTCII AND V. IIOIISLEY

along fibres which do not pass out of these columns, and a strict localisation of nerve
impulses i.s with this intensity of stimulus possihle.

The posterior columns are particularly susceptible to injury, since they are especially
exposed to danger of drying, &c., in the pre])aration of the cord. This circumstance
must be always kept in view when experiments involving an intervening exposure of
the cord are made. We were muth puzzled by the result of two experiments in
which, after dividing the cord in tlie dorsal and luinl)ar regions, we had exposed a
portion of the experimental tract in the intervening part, about the 12th dorsal, for
subsetpient operative lesion. The following results were obtained : —

Dorsal cord e.'ccited.

Stimulus.

Lumbar coi

■d observed.

Le

ft.

Right.

Observed
cud.

Excited
end.

Cat (.-510)

Lat.

Post.

Lat.

Post.

2nd lumbar

8th dorsal

min.

90

31

103

28

max.

190

91

130

85

ma.x.

1G5

65

98

115

Cat (371)

1st lumbnr

9 til dorsal

min.

145

112

140

124

mm.

132

98

144

104

max.

170

190

240

182

max.

280

145

192

195

max.

142

179

U,-3

205

General

mm.
average .

62

42

73

30

1.53

106

141

117

In these cases therefore all lateral column efiects averacfe with minimal stimulus
111, with maximal 166; wdiilst all posterior column effects average with minimal
stimulus 70, with maximal 145. It Is seen that these results are exceptions to the
general rule (see p. 385) of the effect from the posterior being in the Cat larger than
that from the lateral. That this was really due to the prolonged exposure at the
12th dorsal injuriously affecting the posterior columns is shown by the fact that such
exceptional results were only obtained when an intervening portion had been thus
exposed. If prolonged exposure is avoided by performing the necessary operation
mmediately before the desired intervening lesions are to be made no such lowering
of the amount of the posterior column effect is observed. . In this relation it is
possible to explain the comparatively small posterior column effect (see p. 385) of the
Monkey as in some measure clue to a lo,ss of excitability in these columns, which
seem to be much more susceptible than the laterals (Chauveau).

Injluence of Section of one Posterior Column.— It is possible to interrupt one
column only, and thus to obtain conclusive evidence of the extent to which the

ON THE MAMMALIAN NERVOUS SYSTE^r.

397

electrical effects evoked by its stimulation are due to the propagation of impulses
alono' fibres which are confined to the column.

This lesion was effected in three Cats on the left side, the results are seen in the
following table : —

Electrical Effect obtained in the Lumbar Cord by Excitation of Columns in the
Dorsal Reo'ion after intervening: Section of Left Posterior Column.

StimuluF,

Left.

Right.

Cat (357)

2nd lumbar

10th dorsal

Lat.

Post.

Lat.

Post.

500 min.

68

12

82

48

„ min.

75

Lt

120

70

Cat (371)

1st lumbar

9tli dprsal

250 min.

86

0

89

34

500 min.

115

0

145

66

1000 max.

198

45

180

170

,, max.

169

26

146

130

Cat (378)

500 min.

59

3

75

26

1000 max.

lU

34

131

193

Average Effect obtained from Side of Section.

Lateral

Posterior . . . .

Minimal
stimulation.

Maximal
stimulation.

80
6

170
35

Average of Effect obtained from Side Opposite Section.

Lateral

Posterior ....

Minimal
stimulation.

Maximal
stimulation.

102
SI

1.52

164

It will be noticed that with minimal stimulation the interruption occasioned by the
unilateral section Is almost complete ; hence we conclude that —

(a.) Tlie stimulation must be localised to the column on which the electrodes are
placed ;

(b.) The electrical effect in the lumbar cord, following excitation of one posterior
column, must be due to nerve impulses proceeding down channels confined to that
column ;

398

MESSRS. F. GUl'CIl AN'l) \'. IIOIISLEY

(c.) No tract of fibres in other columns can linve direct and continuous connections
with the fibres in one posterior coUnnn.

With maximal intensity of stimulation the interruption is sufficient to lower the
posterior effect in the proportion of 1G4 (that of the uninjured posterior) to 35, that
is, 80 per cent. The effect still obtained, in spite of the interruption, must be due to
indirect paths which come into relation with cells ; these ])aths may partly cross or
be continuous with fibres contained in the uninterrupted lateral columns on the side
of the lesion. The next experiment to be detailed (C) will throw some light upon
this point.

C. Influence of Section of both Posterior Columns and one Lateral.

By making a hemlsection and then dividing the remaining posterior column, a lesion
is made (Miescheb) in which the only intact fibres in the cord are those of one lateral
column, one anterior column, and the fibres in one half of the grey matter.

Such a lesion was made in two animals at the 12th dorsal vertebra, with the
followino- result on the electrical chano'es in the lumbar cord,

Hemisection on the right side and section of the left posterior column.

Oat (357) : ■
Cat (351) . .

Coi-d
observed.

Coi-d
excited.

Stimulus.

Left.

Riglit.

Lat.

Post.

l;at.

Post. -

2nd lumbur
3rd lumbar

9th doi'sal
rtli dorsal

600 min.
2000'max.

90

75

140

135

125

0

0

32

0

0

U

0

0

16

It will be seen that with a " minimal " stimulus no electrical effect was evoked
except by excitation of ttiat lateral column (left) which remained intact. Hence, the
lateral column minimal effect is as strictly localised to its contained fibres as that
of the posterior column, and the presumption is that it contains no fibres which
are directly continuous with those which form at the level of the 9th dorsal the
other lateral and the two posterior columns.

On the other hand, with a maximal stimulus, excitation of each cohunn evokes an
effect, which is, however, much less pronounced than that obtained by the excitation
of the uninterrupted lateral. It will be noticed that the posterior column on the
side of the uninterrupted lateral gives far the most marked of these effects. If the
effects are taken to mean that there exist indirect channels by which nerve impulses
may pass from the fibres of any excited interrupted column into those of the uninter-

ox THE MAMMALIAN NERVOUS SYSTEM. 399

rupted lateral, then the preponderance of the effect evoked by the left posterior
column would show that such indirect communication between tlie fibres in the
posterior column and those in the lateral consists chiefly of posterior column fibres,
which do not cross the cord. It may indeed be doubted whether the small effects
evoked by excitation of the lateral and posterior columns of the side opposite the
uninterrupted lateral are any evidence of veritable crossing nerve impulses, since it is
probable, that with the strength of stimulus used (2000), tb.e excitation of any
column may cause reflex discharges from the cells in the grey matter of the upper
dorsal portion of the cord, and the effect transmitted by the uninterrupted lateral
would therefore be due to spread of tlie awakened activity involving the cells which
are connected with its own internuncial fibres, and which thus send forth nei've
impulses to be ti'ansmitted down the uninterrupted channels.

However this may be, there is no indication by this electrical method of any large
indirect crossed path, in connection with the lateral column, of fibres such as would
conduct nerve impulses downwards from the dorsal to tlie lumbar cord. The interest
which attaches to this negative result hangs upon the presumed connection of the
lateral^' tract with nerve fibres on both sides, and upon the circumstance that it
forms the first of a series of results all pointing conclusively in the same direction,
which will be set forth in Chapters IX. and X.

(2.) Electrical Effects Evoked in the Dorsal Cord by Excitation of the Lumbar

Cord, after intervening Sections.

A. Influence of Hemisection.

Since the characters of the normal electrical effects in the cord are not the same
when the nerve impulses are ascending, i.e., evoked by excitation of the lumbar
region, as when descending, i.e., evoked by excitation of the dorsal region, it is not
surprising- that the influence of intervening lesions upon the former group of effects
should differ from the influence already described upon the latter group.

The influence of hemisection upon ascending changes was studied in two animals
(Cats), one with minimal, the other with maximal stimuli, the experimental tract
being situated as shown in the following table, and the hemisection made on the left
side at the level of the 13th dorsal and lOth-llth dorsal vertebrce respectively.

* MiEscHEE, ' Arbeiten aus dem Pbysiologiachen Institut in Leipzig,' von C. LCDWio j WoeOschiloff, Hid.

400

MESSRS. F. nOTCIl AND V. IIOKSI.KY

Intervening Hemisection on tlio Left Side.

Cat (375)
Cat (355)

Cord cut for
observation.

Cord cut for
excitation.

Stiinulu.s.

T.eft.

Right.

Lat.

Post.

Lat.

Post.

Sth dorsal

■Jiul luinl);u'

500
1000

8
0
0
0

0

C5
2G
40

80
105
106

120

40
80
8(3
9G

"Minimal" Stimulation. — In the first Cat (375) the "minuTiar' stimulation in the
lumbar region of the lateral and posterior columns on the side of the section evoked
no appreciable effect in the dorsal region, the negative result thus confirming our
previous conclusion, that the direct fibres in these columns are completely confined
to one side of the cord. This, it may be pointed out, is strictly in accordance with
the present state of minute anatomical knowledge, as obtained by the study of the
degeneration and developmental histological changes.

• Average of Effects with Maximal Stimtdation. — With the maximal stimulus, no effect
was evoked by the stimulation of the lateral column on the side of the intei-ruption.
This contrasts with the fact that when the nerve impulses are propagated down the
cord, as in the previously deso'ibed experiments, such stimulation evoked appreciable
effects.

The remaining columns gave average effects of

44 for the posterior column on the side of the section,
110 ,, lateral ,, „ opposite to the section,

87 ,, posterior on the opposite side.

The posterior column thus contains indirect fibres which come into relation with
cells, but through these with fibres in the column on the opposite side of the cord, and
this connection is such, that apparently it offers less resistance to the passage of
ascending or afferent impulses than it does to that of descending ones, or possibly
special facilities exist for the discharge of cells in response to impulses ascending in
the posterior column.

B. Influence of Section of Posterior Columns.

The influence of the section of both posterior columns upon the effect evoked by
excitation of the lumbar cord was studied in the case of such lesions made upon three
animals (Cats), in two of which the section was carried out at the time of experiment
at the level of the 12th dorsal, whilst in the third it was done 28 days before at the
level of the 10th dorsal vertebra. The results are displayed in the following table : —

ON THE MAMMALIAN NERVOUS SYSTEM.

401

Section of both Posterior Columns.

Cat (250)
Cat (197)

Cord cut foi-
observation.

Cord cut for
excitation.

Stimulus.

Left.

Ri.Q:ht.

Lat.

Post.

Lat.

Post.

Sth dorsal
11th dorsal

3rd lumbar
4th lumbar

300

1000 max.

2000

30

240
(34

0
65
35

45

310

32

15

105

25

Previous Operation, Division of both Posterior Columns.

Cat (251)*

Cord cut for
observation.

Cord cut for
excitation.

Stimulus.

Left.

Right.

Lat.

Post.

Lat.

Post.

6th dorsal

2nd lumbar

1000

min.

max. (less

ether)

68

58
80

0
2

20

30

100

96

10
0

28

The differences between the results obtained from sections made at the time and
beforehand, are perhaps those involved by the necessity in the latter case of using a
slightly stronger stimulus. And since the different column effects are similar iu both
cases, the massed average of both lateral and both posterior effects evolced with the
minimal and maximal stimulus respectively may be taken from all three animals.

Effect evoked by lateral columns.

Minimal excitation, 55; maximal excitation, 130.

Effect evoked by posterior columns.

Minimal excitation, 4 ; maximal excitation, 4G.

It will be seen on looking at these averages, that with minimal excitation the
posterior column effect practically disappears as the consequence of the interruption
produced Ijy the section ; thus the view as to the localisation of both stimulus and
generated impulse to these columns is true of afferent as of efferent impulses, when
the former is sufficiently weak.

With a relatively more intense stimulus (or with the animal less etherised), this
localisation of the effect does not occur ; the excitation of the lower and posterior
columns now evokes effects above the interruption, the average amount of which is as
much as one-third that produced by the stimulation of the laterals, that is to say,

* For full desci-iptiiin of tins animal as regards its condition during life and (ho post morliin appear-
ances of the cord, see p. 439.

MDCCCXCI. — B. 3 F

•JO:!

MKSSRS. V. (lOTCIl AND V lIOIiSLEY

there are iiidiieel lilires in the posterior colmnn.s wliicli arc not interrupted, and
wliieh eonnect these tlirough cells wiLli tlie lateral columns. This we have already
seen in the casedrihe lateral euluinn exclusion experiment (see p. ;'.i)S), to be also true
for ellerent impulses, hence each posterior cohuiiu is indirectly coiuiected with the
lateral of its own side.

We now pass to the experiments nuule in simihii' [ireparations on one posterior
colunm only.

Injlmnce of Section of one Posterior Column. — Our experiments upon this point
were made on two animals, but in each case with a strength of stimulus which was not
minimal, since, allhough in one animal the absolute intensity of the exciting currents was
that generally used for minimal stimulation (500), yet the preparation was in a hyper-
cxcital)le state owing apjiarently to the lumbar section having hit the entrance of a
posterior root. That the preparation (Cat 375) was hyperexcitable at the moment
when the intervening section was to be made, is shown by the fact that with this
weak intensity of stimulus the excitation of the lateral columns gave deflections of
IGG and 285, and that of the posteriors of 1G8 and 202. We endeavoured to lower
the excitability by more profound etherisation, but the results evidently belong to that
class which have l)een considered as evoked by maximal stimuli. Tlie true eftect of
minimal excitation was not, therefore, observed. The left posterior column was divided
in both animals at the level of the 11th dorsal vertebra.

Electrical Effect obtained in the Dorsal Cord by Excitation of Columns in the
Lumbar Region after intervening Section of the Left Posterior Column.

Cat (375) .
Cat (U>7) .

Cord cut for

obsorvatiou

at

Cord cut for

excitation

at

Stimulus.

Left.

Right.

Lat.

Post.

Lat.

Post.

OtL dorsal
nth dorsal

2nd lumbar
^tli lumlwr

500 max.
2000 mas.

108
70

26

48

115

20

58
205

Average effect evoked by excitation of the —

(1) Lateral columns 78

(2) Posterior columns on side of lesion 37

(3) Posterior columns on side opposite lesion .... 131

The interruption in the left posterior column thus did not abolish, but only reduced,
the electrical effect present in the dor.sal cord above the interruption, when the column
was maximally stimulated below in the lumbar region. The amount of the reduced
effect is one quarter of that wiiich can be evoked by maximal stimulation of the
opposite uninterrupted posterior column. The existence of even a reduced effect is

ON THE MAMMALIAN NERVOUS SYSTEM.

403

presumably due to nerve impulses proceeding up channels which are composed of fibres
starting below the section in the posterior column, but brought, by the intervention of
cells, into relation with fibres contained in the lateral column of the same side, and in
the posterior of the opposite side.

An instructive experiment bearing upon this subject remains to be detailed. We
divided the posterior roots of the 5th, Gth, 7th lumbar, and 1st and 2nd sacral
nerves on the left side. After twenty-one days the animal (Cat 227)"' was used for
experiments on the cord. These were of the same kind as the previous ones, the
cord being divided in the dorsal and lumbar regions, and the experimental tract
prepared for attachment to the galvanometric electrodes at its dorsal end and for
stimulation at its lumbar. The excitatory electrical effects obtained with excitation
of the differtiut columns are given in the following table.

Division of Left Posterior Roots of Sciatic Nerve.

Cat (227) . .

Cord cut for

observation

at

Cortl cut for

excitation

at

Stimulus.

Left.

Rigbt.

Lat.

Post.

Lat. Post.

lOtli dorsal

3rd lumbar

1000

82
125

66
71
41
30

71
145

105

205
61

85

Average effect obtained by exciting both lateral columns lOG

Average effect obtained by exciting posterior column on side of lesiou 52
Average eflect obtained by exciting posterior column on opposite side

to lesion 114

It is clear, therefore, that the result of the division of the root was to reduce the
posterior column effect on the side of the lesion, the diminution of effect amounting
to about a third of what must be considered its full value, since the posterior column
effect on the side opposite the lesion is probably lower than it would be in the normal
animal.

If, tlierefore, the stimulus had been of sufficient intensity to evoke impulses trans-
mitted by direct fibres only, then this diminution would mean that a third of these
direct fibres had degenerated in consequence of the section. The stimulus, however,
was evidently too strong to be classed as " minimal," and all the effects must be con-
sidered as increased by the accession of impulses proceeding up indirect paths.
Hence, the experimental results suggest that from the total nunibei- of direct and
indirect fibres uniting the lumbar region of the left posterior column with the dorsal

* For full description of ibis animal see p. 'iiiG.
3 F 2

404

MKSSIIS. K. (iOTOir AN' I) V. HOIISLKY

n^gioii, iho socLioii ol' the pi).slei'ii)i' roots Ikls willulrawii I'rtiiu aclivo servico ii pro-
portion equivalent to about oue-tliird. and rendered them incapable of conducting
impulses.

It is conceivable that the cells in (he posterior h(ini, which, in some animals,* ;ire
in relation with the ])osterior root fibres, may have then' functional activity impaired
by the section and consequent de^'enerative changes, and that a small part of the
diminution may he due to the withdrawal of im[)ulses possibly generated in tlicm.
(See Chapter XI.)

C>. Infiuence of Section of both Posterior Columns and one Latend.

{\\ order to ascertain to what extent ascending impulses sta)-ted by lumbar excita-
tion of the lateral column are interrupted by its section, a similar experiment to that
described with descending impulses was carried out. An intervening hemisection on
the left side and a further division of the posterior column on the right side were
made ; there remained only the lateral of one side, the contiguous grey matter, and
the anterior columns to bridge across the interruption (Miescher).

The cord was prepared and excited as in all the previous exj)eriments, two strengths
of stimulus being used, both more than "minimal."'

Cat (35.5)

Cord cut for

observation

at

Cord cut for

excitation

at

Stimulus.

Left.

Right.

Lat.

Post.

Lat.

Post.

8th doreal

■2nd lumbar

1000
2000

0
0

0
19

110
220

19
40

It is evident from this experiment that (l) the excitation of the uninterrupted
lateral column of the right side alone j^roduces a marked effect ; (2) that the excita-
tion of the posterior columns can evoke eliects in spite of the interruption in the lateral ;

(3) that effects can be obtained from the posterior column of the same side with an
intensity of stimulus which is not effectual in the case of that of the ojDposite side ;

(4) that even with the strongest intensity of stimulus used, excitation of the inter-
rupted left lateral column evokes no result. We may, therefore, infer that afferent
tracts in the cord lying in one lateral column are brought into relation with the
posterior but not lateral column of the other side. The importance of this result will
be realised if it be remembered that it has been concluded from the result of physiolo-
gical experiments (Woroschiloff, &c.), to be detailed in the next chapter (p. 41D), that
the majority of sensory fibres entering the cord pass into the lateral column of the
opposite side, the remainder continuing in the lateral column of the side on which

* See Fkeud's paper on the cells in Petromyzon.

ON THE IklAMMALIAN NERVOUS SYSTEM.

405

they enter. It has been shown l)y tlie present experiments that, as far as the
nerve iin})ulses aroused by electrical excitation of nerve filjres in the cord ai'e con-
cerned, no such extensive crossing occurs in the dorsal or lumbar regions ; and the
experiments in Chapters IX. and X., on the relation of the cord to the lumbar nerves,
show that no such extensive crossing into the opposite lateral colunm occurs when the
impulses are generated in the entering nerves themselves.

It is difficidt to perform a satisfhctory experiment on tlie influence of section of one
lateral column, since the extensive lesion alters the excitability of the jsosterior colunm
of the same side, and this interfei-es with the sharpness of the results. To avoid this
we made the experiment upon an animal (Cat 259)* In which thirty-four days pre-
viously the lateral column had been severed on the left side at the level of the 10th
dorsal vertebra. The cord was then divided and prepared In the dorsal and lumbar
regions, and the following results obtained : —

Cat (259)

Cord cut for

observation

at

Cord cut for

excitation

at

Stimulus.

Left.

Right.

Lat.

Post.

Lat.

Post.

4th dorsal

1st lumbar

1000

0
Trace

40
50

38
37

40
73

Average of lateral on side of section Trace,

Average of lateral opposite to section .... 37,

Average of posterior columns 51.

The lateral column Interruption with this Intensity of stimulus, which from the
inexcltable condition of the cord was practically minimal, Is thus sufficient to abso-
lutely block any transmission upwards of nerve impulses generated in those fibres of
that cohmin which lie below in the excited lumbar region.

* In this auimiil, just before tlic experiment, it was found that tliere was spastic paralysis in the left
hind limb witii diminished sensibility. The knee jerks were equal and normally present. On micro-
scopical examination the lesion was found to have destroyed the lateral culumn of the left side, the
outer portion of the anterior column, and the dorsal two-tifths of the posterioi' column ; while, on the
right side, a portion of the posterior median was injured. Of descending degeneration there was
noted the posterior third of the left lateral and the median portion of the left anterior column. Of
ascending degeneration a section at the 5th dorsal showed degeneration in both postero-median
columns, extensive on the left side, but only few fibres in the centre of the right column. Left cerebellar
and antero-lateral tracts degenerated.

40G

MESSRS. F. GOTCll AM) V. IIOIISLKY

Section IO.^Summahy of Results.

It will 1)0 well now to express the general conclusions (bunded upon a coinparisim
ot'.dl these experiments, especially those involving the influence of intervening sections
upon the cord electrical effects.

I. If a portion of cord bo severed by a dorsal and lumbar section from its connec-
tions above and below, then stimulation of either lateral oi- posterior column at one
end evokes electrical effects at the other, Avhich are dependent for their amount on
the uninterrupted structural continuity of fibres in the particular column excited ;
and the electrical changes are thus the index of the arrival at one end, of nerve
impulses generated at the other, and propagated mainly along the fibres in each
column. .

II. If the stimulus used be minimal, then the interruption due to the section is in
every case sufficient to ])ractically abolish the effect ; hence, with this stimulus, the
nerve impulses aroused by localized stimulation of fibi-es, which are exposed in the
section, are entirely confined to the particular colunui excited. This is true, whether
the stimulus be central or pei'ipheral to the observed region, that Is, whether the
impulses are ascending or descending.

III. If the stimulus be maximal, then a greatly reduced effect is still obtained, in
spite of the interruption. This effect is due to nerve impulses, which must at some
period be conducted along indirect fibres outside the excited column. Its average
amount varies in the different columns. These amounts may be arranged in series as
shown in the following table, in which the nature of the excitation and of the intei-
rujation is noted, and the average eft'ect evoked by maximal stimulation is given.

Character of
impulses.

Ascending

Descending-
Ascending
Descending

Asceudiuo-

Column excited.

Lumbar cud of cut posterior . <(

I Dorsal end of cut posterior
i
l^Dorsal end of cut lateral .

Lumbar end of cut posterior
( Dorsal end of cut posterior
"1 Dorsal end of cut lateral .

Lumbar end of cut latei'al .

Column interrujjted.

Section of same posterior

Section of botli posteriors

Hemisection same side

Section of both posteriors and lateral

of opposite side

Section of same posterior

Section of both posteriors and opposite

lateral

Hemisection same side

Hemisection same side

Section of both posteriors and lateral

of same side

Ditto

Section of both posteriors and lateral of

same side

Hemisection same side

Hemisection same side and section of

posterior of opposite side . . . . '.

Section of lateral same side ....

Effect.

52
46

44

40
35

32

29
23

19
10

11

0

0
(.1

ON" THE MAMMALIAN NERVOUS SYSTEM. 407

IV. The table just given indicates that the indirect path along which impulses are
conducted from the excited end is not confined to the column, and suggests that the
size of the effect evoked indicates the amount of indirect connection through cells with
other columns.

V. The magnitude of the indirect effect evoked is greatest in the case of the
lumbar excitation of the posterior columns ; the afferent indirect connections leaduig
upwards from this column are, therefore, either very numerous or offer little resistance,
and spread largely into other regions.

VT. The effect is still comparatively large when evoked by dorsal excitation of the
interrupted posterior column. There are, therefore, indirect connections leading
downwards towards the periphery, possibly the same to some extent as those just
mentioned in V., but either less numerous or offering more resistance, or spreading to
a less extent into other regions.

VII. The excitation of the doi'sal end of the interrupted lateral column evokes
electrical effects, hence indirect connections are thereby suggested spreading from the
lateral column downwards into other regions ; these, however, are less numerous or
offer more resistance than those mentioned in VI.

VIII. The excitation of the lumbar end of the interrupted lateral column evokes
no effect, hence all the channels in this column, by which impulses could pass
upwards, are absolutely confined to the column.

IX. The striking feature of the preceding conclusions is that the connections of
the fibres of the posterior columns are framed on a different plan from those of the
laterals. The posterior column fibres spread into other columns both as they ascend
and to a less degree as they descend ; the lateral fibres, however, spread into other
columns particularly as they descend. The posterior columns offer thus special
facilities for conveyance and distribution of afferent impulses ; the lateral special
facilities for the conveyance and distribution of efferent impulses.

It wiU be remarked that the conclusions just suggested as to the relations of the
columns are in accordance wath the views of Schiff, " die Hinterstrilnge bis ins Hirn
hiuein bilden die Legislative, die Tiirck'schen Btindel erwecken die Exekutive."

In couclusion, we draw attention to the general indications as to the structure of
the cord which the whole of these experiments imply.

Since the electrical effect obtained by the " minimal " stimulation of the cut end of
one column is apparently due to the excitation of the direct fibres in that column
only, the average amounts obtained in the case of the dif!ei-ent columns is related .
to the excitability and number of the fibres contained. As regards the excitability
of the direct fibres in different columns, there is no reason for supposing that these
direct fibres, which have no connection with cells between the two ends of the experi-
mental tract, are in any marked degree more excitable in one column than in another.
The amount of the " minimal ''' effect must therefore depend upon the number of fibres

408 MRSSRS. F. GOTCU AND V. IIOUSI.KV

excited, and a comparison of the amounts in tlio two rolinnns wniild ilnis oive a
rough estimate oi' their conijiarative number.

In the Cat the eifects up and down evoked liy (he niiuiniiil excitation of tlie
posterior cohnnn arc Oi2 and 74, that is an average of 8:5 ; those sinTdarly evoked by
minimal excitation of t'le lateral column are 42 and 55, giving an average of 48.
Hence we infer that- tin' tiiret't iilncs in the Cat Ix^tween the doi'sal and lumbar
regions are nearly twice as numerous in the posterior as in the lateral columns. In
the Moid<ey similar averages are 31 for the posterior and 75 for the lateral ; hence in
this animal we infer that the direct fibres are at least twice as numerous in the lateral
as in the posterioi- column.

In both animals "minimal" excitation of the anterior columns at one end of the
experimental tract evokes a doubtful effect at th.e other end, and we infer that there
are few, if any, directly continuous fibres leading in these columns between the mid-
dorsal and the lumbar regions. Indeed, the cord may be divided in its anterior
region* without in any way interfering with the production of electrical effects, either
on the peripheral or central side of the division. Further, the application of more
intense stimulation to the anterior columns although producing local movements evokes
no constant electrical effects, the ordy results ever obtained being either with the use of
considerable intensity of stimulation or with the animal but very slightly etherised.
In these last cases, the magnitude of the local movements renders any strict localisa-
tion of the stimulus upon the deeply situated anterior columns impossible ; any results
obtained under such circumstances we therefore rejected as absolutely untrustworthy.

It must be borne in mind that in all the preceding experiments a degree of
anaesthesia was used, with which no violent movements were caused by any of the
stimuli employed. How far the relations of one column to another can be detected in
the unanaesthetised animal by the employment of similar methods to those now used,
does not seem to us a question likely to yield fruitful results by its attempted
solution. The nervous impulses generated by the direct stimulation of fibres are
undoubtedly far more intense in their character than those which form the flow of the
sensory and motor processes in the normal animal This is especially true wiien the
fibres in the posterior roots and the cord are subjected to external stimulation, jaossibly
because some mitigating Influence of the cells in the posterior root ganglion is removed.
Hence, when such external stimuli are used, it is desirable that these should be of the
weakest character consistent with obtaining definite results, and that the animal
should be in a state of narcosis sufficiently profound not merely to cause complete
insensibility to pain, but the abolition of all violent reflex movements.

The conclusions as to the conduction of nerve impulses in the cord of the Cat and
Monkey, to which the study of the foregoing experimental details has brought us, are
only true when these impulses are generated in the mode hitherto employed, viz., that of

* These experiments will be published in detail in a later communication in which wo shall specially
deal with the anterior columns.

ON THE MAMMALIAN NERVOUS SYSTEM. 409

electrical excitation of the fibres of the cord itself We have, however, used for the
sake of control, a method which did not involve such stimulation, that, namely, of inject-
ing a dose of a 1 per cent, solution of acetate of strychnia sufficient to ensure toxic effect.

Under those circumstances, whether the upper or lower end of the experimental
tract is connected with the galvanometer, marked electrical changes, evidently
excitatoiy, are produced whenever by touching the afferent nerves of the experimental
tract a strychnia convulsion is evoked. Hence the impulses thus generated in the
cells of the cord pass both up and down along the nerve fibres. As to the channels
in which they pass, we can only say that since a section of the whole ventral or
anterior portion of each half of the cord did not sensibly diminish the amount of the
electrical changes, whilst a section of the whole dorsal or posterior portion of each
half of the cord did notably diminish the amount, it would seem that these channels
along which the discharge of the cells passes are situated in the dorsal half of the
cord, that is, in the posterior or dorsal third of the lateral columns, the posterior
columns, and the posterior horns.

In another experiment of the same general kind, we divided the upper end of the
cord into two halves, dorsal and ventral, as described in the method of operative pro-
cedure. (Chapter III., Section 2.) We then connected each half with the galvano-
meter electrodes, and observed the amount of the electrical changes in each during
strychnia spasms. The changes in the dorsal half were very much more marked than
those in the ventral (anterior) half, thus not only confirming our previous conclusions
that the nerve channels by which the impulses were conducted were situated in the
dorsal half of the cord, but also suggesting novel considerations as to the source and
direction of discharges of nerve energy in the cord. (See Chapter XL)

This experiment being rather framed to control our method than to ascertain
channels of discharge, must not however be considered as conclusive.

Finally, it will be noticed that we do not attempt to differentiate between the
different columns of nerve fibres which are situated in the lateral column, direct
cerebellar, antero-lateral, pyramidal, &c.

We will now pass on to consider the electrical changes evoked in the cord, not by
stimulation of its own fibres, but those of its nerves.

3U)(_CC'XC1. — B. 3 G

10 MKSSRS. F. GOTCH AND V HORSLRY

CHAPTKR IX.— ON THE Er,ECTRrCAL EFFECTS EVOKED IN THE SPINAL CORD BY

EXCITATION OF THE LUMBAR NERVES.

Section I. — Introductoiy.

Section 2. — Results of Preliminary Experiments.

Section 3. — Present Knowledge as to the Conduction of Afferent Impulses by the Spinal Cord.

Section 4. — Anatomy of the Lumbar Plexus in the Cat and ^lonkcy.

Section 5. — Method involving Intervening Sections.

Section 6. — iTifluence of Hemisection.

Section 7. — Section of T'ostcrior Columns.

A. Both Columns.

B. Column same side as Nerve.

C. Column opposite side to Nerve.

D. Section some time previous to observation.
Section 8. — Section of the Lateral Columns.

Section 9. — Summarv of Results.

Section 1 . — Introductory.

The presence of excitatory processes in the bundles of nerve fibres which compose
the spinal cord is characterised, as the experiments detailed in the preceding section
have shown, by very definite electromotive changes. It is thus possible to extend
the field of enquiry which the estimation of the amount and character of these
changes has opened up, so as to embrace not merely the relations of the fibres in one
portion of the cord to those in another portion, but the isolations of these to the
entering and issuing nerves. The importance of obtaining some definite data with
reference to these questions will be obvious when it is remembered that there is
hardly any subject in which the evidence is at once so conflicting and so unsatis-
factory as that of the paths by which afferent and efferent impulses travel from and
to their respective nerve roots. If the present method of observing the excitatory
electromotive effects is a trustworthy guide to the presence and amount of nerve
impulses in the particular region connected with the galvanometer or electrometer,
then there is every hope that its application in experiments, designed to show the
relations now referred to, will be one from which definite results may be expected.
How far this is borne out the following experimental details will indicate, but it may
be at once stated that the procedure has not belied its promise, and that, by applying
still more accurate and delicate methods of the same kind, there is every reason to
suppose that in the future results still more definite will be obtained, and that the
actual differentiation of the afferent from the efferent fibres in their course through the
bulbo-spinal system will thus be arrived at. In this manner another method will be
secured for research into the anatomy of the central nervous system.

We must again note that there are two considerations concerning the functional
causation of any excitatory process in the cord, that of pure conduction in fibres, and

(W THE MAMMALIAN NERVOUS SYSTEM. 411

that involving the activity of the corpuscles. If, then, on exciting a mixed nevve it
be found that excitatory processes are present in a distant portion of the cord under
investigation, these may be due to —

(1.) The propagation of tliese processes from the point of stimulation along direct
and continuous nerve fibres which pass into the cord and on through the investigated
region ;

(2.) To the reflex discharge of interposed cells in the cord, brought about by the
arrival of excitatory processes which, having travelled up the nerves from the point
of stimulation, have entered upon paths ending in cells. These cells being thus
awakened, themselves start nerve impulses which travel on up the cord, and, reachhig
the investigated area, produce the obsei-ved electrical effects ;

(3.) To the mixture of effects produced by impulses both in direct paths and the
indirect ones just described.

It must, therefore, be. borne in mind throughout the present research that these
different causes may be operating in the production of the electrical effects observed ;
hence, one of our first experimental enquiries was to endeavour as far as possible to
discriminate between the effects referable to the direct fibres actually excited, and
those which were related but indirectly with these fibres, being primarily associated
wdth channels connecting the area of investigation with corpuscular mechanisms in the
cord,

This we have attempted to do by employing such a low intensity of stimulation as
should in any particular degree of narcosis be sufficient to evoke nerve impulses, as
evidenced by electrical changes, and, at the same time, not so intense as to arouse
any sign of reflex movement. The weak impulses thus generated In the excited nerve
presumably pass for the most part, if not entirely, along direct nerve fibres in the
cord which are not in relation with nerve cells before arriving at the region observed
(mid-dorsal) ; the evidence to be detailed later strengthens this presumption.

The subject matter now dealt with really comprises the whole question of the rela-
tion of the bulbo-spinal system to the lumbar nerves. It is obvious that this relation-
ship is one which may be discussed under three heads according to the particular part
of the nerve tract under observation.

First, the characters of the nerve impulses which pass along the channels of the
coi-d after entering it by its nerve roots ;

Second, the characters of the nerve impulses which, aroused by excitation of the
channels of the cord and passing out into its nerve roots, descend along fibres of
the mixed nerve ;

Third, the characters of the nerve impulses which enter and leave the cord by
virtue of its so-called central structures and their various connections.

Of these three groups, the first alone will be treated of in the present chapter, the
second and third being reserved for Chapters X. and XI. respectively.

3 G 2

412 MESSRS. F. nOTCII AND V. Um?STiKY

Section "2. — Gkxekai- Vlkn and Results or Preliminary Experiments.

The experiments to lie referred to in this chajiter were made upon botli Monkeys
and Cats, and all involved the following operative procedure. The spinal cord of the
etherised animal was exposed in the lower dorsal region ; it was then divided and
prepared on the peripheral side of the section for 4 centims., so that the upper end of
this lower fragment of cord could he raised from its canal and attached to the cables
of the non-polarisable electrodes as described in the previous sections. The galvano-
meter circuit was then brought by one electrode into relation with the cross section
of the cord, by the other with a ring of longitudinal surface at about a centimetre
distance. (See Plate 29.)

The two sciatic nerves were now exposed and divided, so that they could be I'aised
and the central end placed, when desired, upon a pair of platinum electrodes (see
p. 301 and fig. 3), precautions being observed in their preparation with respect to
maintenance of circulation, drying, &c., as indicated at length in Chapter III.

If the large and persistent resting electrical difference between the cross section and
surface of the cord be balanced, and the galvanometer needle thus brought into its zero
position, then on exciting the central end of either sciatic nerve for 5 seconds with
the interrupted induction current (Helmholtz side-wire),' an electrical change occurs in
the cord, the surface contact becoming gal vanometrically negative to the cross section.

On the cessation of the excitation, the needle rapidly returns to its original point
and, in most cases, then continues to slowly move to a position upon the opposite side
of the zero, thus indicating that the excitatory negative change only lasts as long as
the excitation, but is succeeded by a more lasting after-effect of opposite sign. This
increase in the resting difference has been already alluded to in Chapter IV. Errors
due to any electrical escape are mimimised by the nature of the excitation, and the
position of the galvanometer electrodes (see Chapter III., Section 4) ; moreover
convincing proof that the effect is truly an excitatory one is afforded by the following
facts : — ■

(1.) A similar negative change occurs when, with the sciatic nerve uncut, the skin
over the foot is pinched with ivory or touched with a hot glass rod ;

(2.) A sudden change of small amount, but similar in sign, is produced when the
nerve is mechanically excited by a ligature, cut, &c. ;

(3.) Very pronounced effects of similar sign occur when, after injection of a few
drops of a 1 per cent, solution of strychnia acetate, the nerves connected with the
lower part of the cord are mechanically excited.

The electrical change in the coi'd is thus excitatory, and its resemblance to that in
nerve warrants the presumption (see p. 277) that it is due to the passage of excitatory
impulses along nerve fibres present in that portion of the cord with which the
galvanometer is connected.

ON THE MAMMALIAN NERVOUS SYSTEM. 413

Etherisation.

These nerve impulses may, as before stated, have two direct sources of origin,
namely, the excited mixed nerve or the grey matter in the cord, which the arrival of
afferent impulses may have thrown into functional activity. They may be conducted
along direct fibres from the lumbar nerves to the dorsal cord or to central mechanisms,
and from these either along continuous fibres to the area observed or along short
internuncial fibres to other cellular mechanisms, and so again to others and thus finally
reach the area.

That the change is due to the presence in varying degrees of all these three factors
is extremely probable from anatomical considerations, since Gaule" has shown that
the number of fibres in the Frog's cord can be satisfied by such a triple arrangement.
It is, moreover, clear, from the influence of profound etherisation upon the electrical
change, that, when an adequate strength of stimulus is employed, the galvanometric
effect varies in amount with the pronounced or slight character of the visible reflex
effect. This has been already referred to (p. 372), but its importance in connection
with the subject of this section warrants the introduction of some experimental
details. We therefore give as examples the following : — In one case, on exciting
the sciatic of a profoundly etherised anunal (Cat) with a weak but adequate
stimulus for 5 seconds, the galvanometric cord effect was 76 scale ; the animal was
then less etherised and the effect was 220 scale. On again deeply etherising the effect
with the same excitation sank to 130 ; upon the ether being then more or less removed,
the excitation produced an effect of 200 ; whilst on renewed profound etherisation
this was only 80. In all these cases very feeble reflexes or none were observed in
the profoundly etherised animal, but the reflexes were well marked when with less
narcosis the larger electrical changes were noted.

These differences evidently point clearly to the fact that the electrical change may
not only be that due to the transmitted nerve impulses up direct tracts, but also to
the presence of interposed cellular mechanisms and their susceptibihty of response
to the arrival of afferent impulses. In Chapter III. we have discussed the value
of the use of ether as a means of discriminating between the functional activity of
fibres and of fibres 2>lus cells, to that chapter we refer the reader, merely remarking
now that we must not lose sight of the fact that etherisation, especially when very
profound, may affect the nature of the excitatory processes in the nerve fibres them-
selves, as in the experiments made on the nerve of the Frog by Biedermann with ether
vapour. The differences between effects obtained with unvarying intensity of stimulus
but different degrees of narcosis are practically the same as those obtained with
different strengths of excitation and unvarying but not too profound anaesthesia.

* Gacle, ' Abhandl. Math.-Physik. Classe d. Kgl. Sachs. Gesellsch. d. Wiss.,' 1889 (vol. 15, No. IX. ;
' Neurol. Centralblatt,' vol. 0, p. 8.

414 MESSHS. V. OOTCn AND \'. IIOUSLKY

Relation of Effect to Stretigfli of Stimulus.

To ascertain this lelationsliip, the practice we employed was to use an initial
strength of stimulus M'hich was just sulHcient to produce a cord electrical eflect when
applied to either nerve with the animal efficiently narcotised, then to follow witli a
stimulus of double that intensity, and observe the now increased value, and finally to
proceed by doubling the strength of stimulus until a point was reached when the effect
was only slightly altered. Such an experiment is the following one made upon the
Cat. An initial excitation for 5 seconds with the secondary coil at a distance of 500
(see Chapter III. on experimental method and apparatus) gave a cord effect indicated
by a galvanometric deflection of 65 ; similar excitation of twice the intensity, 1000,
gave a cord effect of 122, and an excitation of four times the intensity, 2000, gave a
cord effect of 152. In the Rhesus Monkey a similar result was observed; thus an
initial excitation for 5 seconds of the sciatic nerve, with the coil at 2000, gave a cord effect
of '62 (galvanometer deflection), whereas twice the intensity of stimulation, 4000,
gave 60, and three times the intensity, 6000, gave 78. These selected individual
examples furnish data which are strictly in accordance with the average effect as
deduced from all the observed instances which occur in our experimental records.
Thus if we select all the unexceptional readings of galvanometric cord effects obtained
in both Cat and Monkey with the initial strength of stimulus, the average of seventeen
is 56 ; similar readings (twelve instances) with twice this strength of stimulus give
an average of 100 ; whilst the average of sixteen instances with four times the initial
strength is 168. It is thus evident that the cord eflect increases with the intensity
of the stimulus at first in direct proportion to the strength, but afterwards in one
less directly related with this factor. A point is ultimately reached at which no
increase, but a decrease, in cord effect occurs. Tiiis, it need hardly be said, is partly
dependent upon the damaging eflect of the stronger exciting current upon the
stimulated nerve, and partly upon the fatigue of the cord cells through the arrival in
the cord of the very intense nerve impulses evoked by the strong stimulus. In these
experiments, as in all others to be referred to in this section, the number and
characters of the successive stimuli employed were the same in each case. (See
Chapter III. Section 3.)

It is remarkable how close the amounts of the different cord readings obtained in
various individuals of one species are to one another, and how often, especially in later
experiments, the initial intensity of stimulus required with different nerves and different
animals was the same (500 coil), thus showing the similarity of the narcosis. We may
here draw attention to the minimum and maximum effect obtained in the Cat, the first
with the coil at 500 and the second at 2000 ; the minimum being 20, the maximum 340.
So far then as the relation between the size of the cord effect and the strength of the
nerve stimulus is concerned, there is a strong resemblance between a nerve-to-cord
excitatory electrical change and that in a nerve ; the ether eflect before- mentioned

ON THE MAMMALIAN NERVOUS SYSTEM. 415

indicates, however, a difference, since we have not as yet observed in the Cat or
Monkey any sucli very marked change in the excitatory effects in purely nerve
preparations as definitely dependent upon the degree of etherisation.

The absolute determination of the true character of the minimal cord effect, and the
question to what extent it is to be considered as entirely due to the propagation of
nerve impulses along direct and continuous paths, unmixed with impulses generated
in nerve centres, might be settled by accurate experimental observations upon the time
relations of the cord electrical change, particularly the time of its development
relation to the seat of nerve excitation. Such an enquiry would doubtless involve the
use of Bernstein's differential rheotome, since the small size of tlie minimal eff'ect is a
serious objection to the use of the much simpler method of recording by photography
the movement of the meniscus of a capillary electrometer.

It may be mentioned incidentally at this point that the cord effect evoked by a
single electrical excitation of the sciatic nerve Avas, however, sufficiently pronounced to
cause a distinct movement in the mercury of our capillary electrometer. The move-
ment wdien magnified 300 times was just visible with a weak stimulus (coil, 500), and
was quite marked Avith a maximal stimulus (coil, 2000). It was always a single move-
ment, but appeared to the eye rather more prolonged in character than that wliich
was obtained in the case of the single nerve effect as displayed in the photographs of
our former paper. ''^

It may be asked at this juncture upon what evidence we assume that the effect in
the cord, following the excitation of the sciatic nerve, is due solely to the arrival therein
of nerve impulses which have travelled up posterior or afferent fibres. The evidence,
which is supplied in full detail (in Chapter XL) upon the relations of the nerve fibi'es
and the nerve cells, shows that no electrical changes appreciable by the galvanometer
can be produced in the cord by even strong stimulation of the anterior or efferent
roots.

It is thus clear that stimulation of the sciatic nerve generates nerve impulses which
enter the cord by the posterior roots only, although, owing to our ignorance of the
influence of the spinal ganglion, it cannot be assumed that the impulses are the same
in character as those generated by excitation of the posterior x'oots themselves. We
may now pass on to the consideration of the object of our experiments.

The object of the investigation will be best set forth by considering the present
position of our knowledge of the relations of the spinal cord to the lumbar nerves,
and particularly to the afferent fibres of those nerves. For if we ai'e justified
in assuming that the electrical effect observed in the cord is due to the passage of nerve
impulses, and their ari'ival at the portion of the cord under observation, then, since
the method of observation is one which furnishes us with quantitative data, its
employment places within the experimenter's grasp a means of ascertaining to what
extent these impulses can be interrupted by definite section of an v one tract of cord

* ■ Kov. Sui-. I'roc' I'ic. cif.

416 ]yiESSRS. V. GOTCH AND V. HORSLEY

fibres, and thus of indicating wliicli groups of fibres in the cord form tlic cliief
channels of conveyance ; and the vaUie of this method Avill be manifest wlien the
character of those previously used and tlu^ conflicting nature of their results, have
been placed before the reader.

Section 3. — Our Present Knowledge as to the Conduction in the Cord of

Afferent Impulses.

We have already referred in Chapter II. to the investigations made by previous
experimenters in order to determine the paths of conduction in the spinal cord. It
will, however, be advantageous to select the most typical of these experiments ;uid
discuss their results, in order that the scope of these methods, together with those
excellencies and deficiences which each possesses, may be clearly in the mind of the
reader, before the uses of that Avhicli we advocate and its results are entered upon.

The various methods may be grouped as anatomical and physiological.

I. Anatomical.

The posterior root fibres have been traced into the spinal cord, and a considerable
number* (according to Kolliker the majority) of these have been seen to divide and
send branches up and down the cord. The further course of these fibres has up to
the present been traced only by the method of degeneration.

(a.) It has been ascertained that when, in consequence of section of a lumbar
posterior root, the fibres upon its central side degenerate, the degeneration is con-
tinued in certain regions of the posterior column of the same side, such degeneration
up to the present only being seen in parts above the entering root. In other words,
the degenerated fibres present in the posterior root are represented in the cord above
the root by certain fibres in the posterior column of the same side. These fibres are
believed to be the direct continuation of some of the fibres in the root ; they are situated
in the postero-external column at first, but gradually shift towards the middle line as
they ascend, and ultimately occupy a definite portion of the postero-median column.
In addition to these fibres which appear to have a continuous course up the posterior
columns of the cord, there are others which show degeneration only in the immediate
neighbourhood of the entry of the root ; these are situated in the posterior root zone
and in the marginal zone, as well as in the postero-external column. They are
characterised by only occurring in that portion of the cord which lies above the
root, and by being all on the same side of the cord as the lesion. Hence these
fibres, whilst in direct continuity with fibres in the root, either end (Kolliker)

* GoLGi: ' Anat. Anzeiger,' vol. 5, pp. ] 3, 14. Ramon-t-Cajal : ' Trabajos del Laboratorio Anatoraico
de la Faculdad de Medicina,' Barcelona, Abril, 1890. Kollikee : " Ueber den feineren Baa des Riickeu-
niarks," ' Sitzungsberichte d. Wiirzburger Phys.-Med. Gesellschaft,' 8. Maiz, 1890.

ON THE MAMMALIAN NERVOUS STSTEM. 417

or come into relation with cells shortly after entering the cord, the degeneration in
consequence ceasing. This cessation of degeneration is indicated by the numerical
superiority of the degenerated fibres in the posterior roots over those in the posterior
column, and above all the postero-median column. As far as it goes the evidence
itself aflPorded by this method is fairly conclusive ; whatever other channels there may
be, undoubtedly the above-named do exist, and thus the posterior column of the same
side offers a. direct tract up which some nervous impulses undoubtedly not only
can, but do, pass.

(6.) A similar investigation applied to the cord, since it determines all the fibres
in which degeneration is seen above the level of a hemisection, carries us a step
further than the above method. In this not. only tlie fibres just indicated in
the posterior coltimn degenerate, but all the through fibres from every root below
on the side of the hemisection. Further, it marks otit all fibres which have their
nutritive or developmental centre, not in cells in the ganglion on the root, but in cells
situated in the cord below the section. Such fibres now seen as degenerated above
the lesion are grouped into two parts of the lateral column, the antero-lateral, and the
direct cerebellar. Except that the former is much more diffuse, the degenerated fibres
being scattered about among other sound ones, there is little to distinguish the
character of these two groups except their situation. They both appear to have
similar terminal connections, at any rate, on the cerebral side. They are both
chai'acterised by containing fibres of very variable length, many of which appear after
a shorter or longer distance to again come into relation with nerve cells, so that in
these the degenerative change ceases. The characteristics of these tracts are, there-
fore, twofold ; they appear to afford not only a dii'ect connection between certain cells
in the cord (Clarke's column, &c.) and the peduncle of tlie cerebellum of the same
side, but also a series of internuncial connections, by means of which cells at variable
distances are brougfht into relation with each other.

The tracts which degenerate on the peripheral side of a hemisection are foreign to
the purpose of the present chapter ; one, however, may be mentioned, the so-called
" comma-shaped " patch in the ventral part of the postero-external column, since it
is in all probability a tract of ascending fibres which are looped downwards, the
degeneration corresponding with tlie loop.*

As far as this evidence under (6) goes it is not so direct as that in (o). It is
conclusive as showing that, if thei'e be crossing of any of these tracts from one side
of the cord to the other, such crossing must occur 'by means of cells, since the
degenerative change does not cross. It is further conclusive in showing that there
are continuous nerve fibres in these two situations. It is not conclusive as regards
the physiological characteristics of the fibres, since in anatomical evidence, as in cir-
cumstantial evidence, every link of the chain must be present for the evidence to be

* Possibly these fibres ai-e the descending brandies of tlie bifurcating postciior root fibres, cf. espe-

ciall}- KoLLIKER.

MDCCCXCI. — B. 3 H

418 MESSRS. F. OOTCH AND V. ITORSLEY

complete. The essential link, that of connection witii the posterior roots, the fibres of
which convey afferent impulses, is, however, wanting. It is, notwithstanding, highly
probable that these paths (direct cerebellar and antero-lateral), are to be classed as
artbrent from the resemblance in extension of the degeneration to that present in the
fibres of the posterior colunms.

There are many considerations connected with growtli, development, &c., which
suggest that ascending degeneration is the characteristic of afferent, and descending
degeneration of efferent tracts in the spinal cord ; in other words, that the centres of
growth and nutrition are also the centres of functional activity, and, therefore, in
consequence of an interruption, that loss of functional activity and loss of nutrition
both occur in the part on the distal side of the breakdown. The fact, however, that
the cells in the ganglion on the postei'ior root exercise this nutritive function on the
nerve fibres on both sides, although explicable in consequence of developmental rela-
tions (Ills) must cause some hesitation as to the propriety of accepting, without
reserve, the principle that the direction of the degeneration coincides with the direc-
tion of the function. What has happened in the development of the posterior root
fibres may possibly have also occurred in the case of some fibres in the cord. More-
over, it is conceivable that even with this relation between nutrition and function,
ascending degenerative changes might occur in fibres which ought to be classed as
outgoing, since they would lie as much on the efferent side of the central nervous
system as do those of the pyramidal tract. Such fibres would be long internuncial
fibres, connecting the kina^sthetic system of the lower (hnnbar) centres with the
efferent side of a similar system of the upper (or cervical) region. It is conceivable
that such long internuncial fibres exist, and, if so, then impulses which subsequently
become motor are conveyed by them. These impulses, although afferent in the sense
that they are actually travelling up the cord, are analogous in character with all the
motor or outgoing impulses, since the}' occupy that relation to the centres from which
they started.

Without insisting on these hypothetical fibres, we may again point out that the
anatomical evidence becomes vague when some of the links in the chain are wanting,
hence all conclusions founded upon it must be used with the greatest caution. The
facts which it surely evidences are —

1. That there are continuous fibres in the cord.

2 That these fibres do not cross from one side of the cord to the other.

3. That a small pi'opoi'tioji of those situated in the posterior column are in direct
connection with those entering the cord by the posterior roots.

II. Physiological.

If we turn now to the method of physiological exjieriment, thei*e is hardly any
subject in the whole realm of physiology upon which such divergence of experimental

ON THE :\[Ai\IMALIAN NP^RVOUS SYSTEM. 419

result and inlevjjvetation exists as tliat of the relations between the cohimus of the
cord and the passage of afferent impulses.

The contradictor}' focts set forth in the various papers on this subject are snffioieut ■
to justify the 'belief that many of the experiments made must have involved the
presence of some factor of capricious character and uncertain action, common to most
of them, and tiiat the conclusions wliich the various experimentalists have drawn from
their results are all vitiated by its presence.

When such a fundamental matter as that of the extent to which the afferent path
lies on one or the other side of the cord is answered so ditierently that, according to
one set of investigators (Browx-Sequakd, Ferrier, &c.), it is wholly crossed ;
according to others chiefly crossed (Woroschiloff, Miescher, &c.); erpially luicrossed
and crossed (van Deen, Stilling, &c.); chiefly uncrossed (von Bkzold, Mott); wholly
uncrossed (Chauveau); it is plain that the method adopted is at fault and is quite
inefiicient for the purpose of determining the extent to which particular tracts in the
cord are concerned with the passage of impulses. The reason why the different
results of former investigators are so conflicting may, we think, be gathered by the
careful study of the experimental method in two of the most elaborate of the series
of investigations just referred to, those, namely, carried out in Ludwig's Laboratory
by Miescher''" and Woroschiloff. +

Although the experiments are well. known to physiologists, we think it necessary
to allude in a little detail to their character. Both sets of experiments were carried
out on Kabbits, and in both the method consisted in the production of an intervening
localized destruction of a portion of the spinal cord, and in then ascertaining what
changes this destruction produced in the reaction of some more central portion of the
nervous system of the animal to the stimulation of the afferent nerves of the distal
portion. The reaction observed was, however, different in the two cases, it being in
Mtercher's experiment the rise of blood pressure due to the activity of the so-called
vaso-motor centre of the medulla being awakened ; in Woroschiloff's, the movement
of the upper limbs of the animal, due, as he believed, to the awakening of a convulsive
centre in the medulla. The index used liy Miescher being the amount of blood
pressure, had the great advantage that it varied with the strength and duration of
the afferent stimulation, and thus a (piantitative comparison between the results of
two series of stimuli was possible, and it is, perhaps, due to this that the facts have
been held as aftbrding data of .so cogent a character. They show first that the
reaction of the centre, as indicated by rise of pressure, is very largely reduced when
the lateral column on the opposite side of the cord to that of the excited sciatic nerves
has been divided ; and second, tliat when a complete intervening division of the
cord, with the exception of one lateral colunm, is made, the reaction still persists,
beino- most marked when the distal nerve on the side of the complete .section is

* 'Bcr. d. Siichs. Ges. d. Wis*., Math.-Phy.sik. CI.,' 1870. Also ' Arlieitcii n. d, Pliys. Lab.,' Leipzio-.
t ' Ber. d. Siichs. Gos. d. Wiss., Math.-Physik. CI.,' vol. 26, 1874.

3 H 2

420 MKSSllS. K. CiOTCH AND V. IlUKSLKr

stinnihited. From this it was concliided that the spinal path for those affei-ent nerve
impulses which, proceeding up one sciatic nerve into the cord, subsequently reached
the vaso-motor centre, was chietiy contained in the lateral column of the side of the
cord opposite the nerve. The experiments of Wouoscuii.off exteiided these, and
showed that laoveraents of the upper litnbs could be only aroused in response to
stimulation of the skin of the lower limbs, provided that the lateral columns were
intact ; that destruction of the continuity of other portions of the cord did not notably
influence the reaction ; that when one lateral column was divided, the skin stimulation
on the side of the lesion produced a greater reaction than that on the side opposite
the lesion ; and that when the whole cord, with the exception of one lateral column,
was divided, the stimulation of the skin on the side of the intact column, produced a
much less reaction than did that of the skin on the opposite side. From this last
experiment, Woeoschiloff assumed that, since the impulses were transmitted fi'om
the lower to the upper portion of the cord by the bridge of fibres in the intact lateral
column, they travelled throughout along such fibres in one continuous path, keeping
entirely to the lateral column. The difficulties involved in this assumption, and the
unsatisfactory character of the experiments, are ably set forth in a recent edition of
Foster's physiology, to wdiich the reader is referred.* The conclusions, therefore, to
which WoROSCHiLOFF arrived were that afferent impulses were transmitted entirely
by fibres contained in the lateral columns ; and that by far the chief part were trans-
mitted by the fibres in the lateral column on the opposite side to the entering nerve.
By his first conclusion, he therefore placed both the posterior columns and grey
matter out of court, although the former must convey at any rate some impulses
since they contain a considerable number of the posterior root fibres, whilst there being
no means for the Impulses to cross over from the entering roots to the lateral columns
of the opposite side, except through the grey matter, this latter is assumed to be in
operation by the second conclusion.

Apart from the contradictions involved in this interpretation of these two sets of
experiments, the results remain as definite phenomena. In seeking to ascertain their
true meaning, it is essential to realize all the conditions of the experimental procedure.
Now, of all these conditions, the one which is the most important is that connected
with the experimental necessity of evoking definite reactions.

These reactions, whether of vaso-motor, convulsive, or other centres, are pro-
foundly influenced by antesthesia. It is certain that in the unantesthetised animal
the centres are largely affected by the functional discharge of other centres, and that
this influence, although by no means abolished, is diminished by narcosis. In all the
experiments of Miescher and Wouoschiloff no complete anaesthetic was employed,
the animals being in Miescher's investigations simply curarised, and in WoRO-
•sohiloff's fixed in a suitable holder.

* M. Foster, ' Text-book of Pliysiology,' vol. 3, p. 189. Also ' Arbeiten a. d. Physiol. lust.,' Leipzig,
1874.

ON THE l\rAMMALIAN NERVOQS SYSTEM. 421

This being the case, the whole series of experiments are in j-euUty a demonstration
of the extent to which one compUcatecl reflex act is influenced by the awakening in
the entire nervous system of a whole series of reflex discharges. The uncertainty
which must attend such a method may be illustrated by the following experience of
ourselves. It often happened in the experiments described in this paper (all of
which were performed inider an antesthetic), that after not merely complete section,
but absolute removal of a piece of cord 1 centim. long in tlie mid-dorsal region of the
Cat or Monkey, mechanical irritation of the sciatic nerve, and especially of the posterior
roots, caused not merely movement in the upper limbs, but of the head, &c., also.
This eftect was an indication that the anaesthesia had become too slight, and it was
immediately abolished on making the latter rather more profound: We convinced
ourselves that the mechanism of its causation consisted in first a reflex of the lower
limbs, and tliat this movement dragging on the trunk aroused a reflex in the upper limbs
by pulling on the nerves of the upper fragment of cord, which in the neighbourhood
of the section was in a hyperexcitable state. The afferent impulses thus started now
evoked general movements by discharging the higher centi-es. The movements of the
lower and upper limbs often followed one another so i-apidly that the eye was unable to
distinguish between their time of commencement ; in some cases, however, particularly
when the narcosis was increased, the delay between the two was quite plain, and eveiy
stage might be observed. Of what value, therefore, as regards the question of conduc-
tion in the continuous tracts of fibres in the spinal cord are the results of experiments
in which the animal is under no ansesthetic at all, for the employment of curare
by MiESCHER did not provide for the adequate exclusion of reflex centres such as can
be obtained by the use of ether. Indeed, Miescher's method depended upon the
maintenance of the functional activity of the reflexes. It seems to us, therefore, that
this is the crucial point of the whole position, and accounts for the capricious character
of the results of different observers as dependent upon different methods of investi-
gation. The index for the arrival of nerve impulses from one side of a block in the
cord to the other has always been a reflex one, and it is essential for its proper
working that the reflex centres should be in an excitable condition. Since the experi-
ments involve operating on the cord and the examination of the animal in tlie
unanpesthetised state, within a short time of the operation there is every opportunity
for abnormal conditions of the other centres (in consequence of the procedure), to
influence the result ; such influence is so certain from the state of the animal that the
results really determine, not the localisation of afferent fibres in the cord, but the
extent to which the operation has augmented or depressed the excitability of
the successive central mechanisms, and the bridge by means of which the lower
and upper parts of the cord communicate thus becomes chiefly a column of inter-
nuncial fibres. Moreover, almost all experiments done within a few hours of the time
of operation lay themselves open to our further criticism, that it is possible for the
reflex movement of the hind limbs and trunk to arouse by mechanical pull of the parts

422 MESSRS. F. GOTCll AXD V. HORSLEY

iirouiid llio seat of lesion, .siniilar inoveineiits of tlio upper limbs. That .such rellcx
movements were present in the case of Worosohiloff's own experiments, is evident
from tiie studv of the account of tho.se evoked in the ;iiilinal after the operation
iiad heen made upon tho coi-d. Thus, in one (if his experiments (VI.), the animal
(Kalihit), was subjected to an operation involving the division of the spinal cord in
the dorsal I'egion, the whole cord being cut through with the exception of the lateral
tract of the left side. On examining the animal an hour or two afterwards, the right
hind limb was found to be paralyzed, due to section of its efferent spinal tract, this
efferent paralysis being confirmed subsequently by dividing the cord below the medulla
and stimulating its distal cut section wlicn only tlie left hind limb responded ;
notwithstanding this, pressure on either foot caused movements in l)oth liiiid legs.
Now in this case the movement of the right leg must obviously be a reflex eflect,
having its central physiological seat in the lower fragment of the cord, and any
arguments based upon the amount of movement of the respective limbs, must deal
with the question as to the extent to which the section has heightened or depressed
the normal excitability of the reflex centres in this lower fragment. But it is precisely
this question which it seems to us has not been sufficiently dealt with in the text,
there being a tendency to treat the cord in an unantesthetised animal as if one of its
main jjhysiological characteristics, reflex excitability, were in abeyance in the distal,
though not in the proximal and central portions. The evidence cannot, therefore,
be considered as adequate to warrant his statement : " Meine bis dahin vorgelegten
Beobachtungen bestlltigen und erweitern die Angaben von Miescher, Nawrocki
und DiTTMAR, denn von mm an siud wir darauf liingewiesen, die Bahn, welche das
Gehirn mit alien Nervenwurzeln verbindet, in dem Seitenstrange zu suchen." .

The criticism just advanced is not a new one since it forms the basis of the
objections of Chauveau, Schiff, and others. Its cogency is increased by the discovery
made l)y Fodera of hyperfesthesia after section. The most important points in
connection with this fur our present purpose are those brought out by experiments
made by Schiff* before the observations just referred to. These were, that in many
cases of hemisection of the cord there is not merely hyperfesthesia below the lesion t)n
the side of the section, but heightened reflex excitability of the cord in the neighbour-
hood of the lesion both above and below. This hyperexcitability is less marked when
the whole cord is divided, than when a bridge joins the parts above and below the
lesion. It has been found by MARTiNOTTit that tlie hyperfesthesia is particularly
bound up with injury of a particular region in the lateral column, and It is atti'ibuted
by him not to a dii-ect increase in excitability, but to the removal of inhibitory, i.e.,
depressant, influences by the section. It may be pointed out that, as regards the
nerve trunk, the experiments of Hering show conclusively that the hicrease of

* Schiff, ' Leln-bucli d. Physiol.,' 1859.

t Maetikotti, " Hj-peraestliesie nach Verletzung des Hiilsniarkes,'' ■ Arcliiv f. Auat. mid Pliysiol.'
(Fhysiol. Abth.), 1890.

ON THE MAMMALIAN NERVOUS SYSTEM. 423

excitability therein caused is in direct causal relation with the electrotonic effects of the
resting or persistent nerve cuiTents produced between the cross section and the surface
of the nerve trunk. Since similar electrotonic alterations in excitability have been
shown to exist in the case of the spinal cord, it is most probable that a direct inci-ease
would be produced by the very large resting currents which (see Chapter IV.)
have been found by us to be produced by cross sections of the cord. In this
connection we may draw attention to the following striking phenomenon described
by ScHiFF,* and wliich in the progress of tlie present research we liave had the
opportunity of seeing ourselves. We often saw that after complete section of the
cord and the removal of a piece 1 centim. in length in the region of the 8th dorsal
vertebra, when the animal was asphyxiated on the completion of the investigation, a
sudden prolonged discharge of the nerve centres in the lower portion of the cord
occurred. This revealed itself by the rising u[) of the tail of the animal, and its main-
tenance in ail erect position for one or two minutes. The discharge and its effect then
subside to be followed in two or three minutes by a second feebler one, and this by a
third still more feeble. In addition to this it often happened that during an experi-
ment on such a preparation, rhythmical discharges of the centres in the lower portion
of the cord occurred after the section, evidenced by rhythmical contractions of the anal
muscles. This has also been noticed by several investigators.

Further, in one animal (Cat 325) in which the cord was divided at the 13th dorsal
vertebra, althougli the tail jihenomenon did not spontaneously appear, wlien the
animal was asphyxiated, it was readily evoked in full strength by slight stimulation
of the skin of the tail.

These phenomena afford a striking instance both of the hyperexcitable state of the
centres in the lower portion of the cord and of the capacity of these centres to
discharge such a series of nerve-impulses as will produce sustained co-oi'dinated
movements.

The difficulties thus involved in the experimental methods of Miescher and Woro-
SCHILOFF are to a great extent obviated by experiments in which, a given lesion of the
cord having been made with all due antiseptic precautions, the wound has been
allowed to heal, and the animal examined carefully at varying intervals, from a few
days to weeks and months, after the operation. Such an experiment is that made
by FERRlERt on the Monkey, in wliich a hemisection at the level of the 8th dorsal was
followed by complete insensibility below, on the side opposite the lesion, to every
form of sensory stimulus, whilst the sensibility on the same .side remained unimpaired.
This experiment is similar to the paralysis on the same side, and anaesthesia on the
opposite side, observed in some cases of injury or disease localised in one half of the
spinal cord. Other observers, notably Mott,J have, however, obtained results which

* ScHiFF, PFi.iJGEE's ' Archiv,' vol. 30, 1883, p. 202.

t Ferriee, ' Brain,' 1884.

J ' Proceedings of the Physiological Society.' ' Proceeding.s of the Royal Society.'

424 MESSRS. F. OOTCH AND V. HORSLEY

vary between partial accordance and complete o])po.sitiou to the above. It is not our
object to endeavour to reconcile this apparently conflicting evidence, but we may
point out tliat great difficulties are involved in the application of the tests necessary
to determine the sensibility of animals. In our own experiments, wliicli liuve been
made upon Cats, and which involved hemisection of the cord, the application of the
water and other tests to the skin indicated that the aiferent stimulus was conveyed
from both hind legs up the cord, but with much greater certainty, and, judging by
the quickness in the evoked movement, with much greater intensity, on the uninjured
side than on that corresponding with the lesion.

It is evident that the flaw in all these methods is the fact that, owinir to the index
of the arrival of impulses above the lesion being some movement, the central struc-
tures of the nervous system must be sufiiciently excitable to be capable of responding,
and the stimulus used must be intense enough to arouse discharges of sufficient
magnitude to evoke definite muscular movements.

Wliilst the same flaw in the classical experiment is also present in the case of
efferent tracts, its presence here does not produce such a blurring of the result, since
the method can always be checked by a concluding experiment involving division of
the cord below the medulla, and excitation of its distal jaortion, with observation of
the amount and character of the movements of the lower limbs. AlthouMi acknow-
ledging to the utmost the value and precision of the experiments carried out in
Ludwtg's laboratory, we must exjwess our conviction that, as regards the conduction
of afferent impulses, they do not warrant the inference either that the lateral columns
contain the fibres of the afferent tract to the exclusion of the grey matter and the
other columns, or that the fibres which they do contain form a continuous afferent
path, it being quite possible that those fibres which undoubtedly conducted at the
point of section were merely acting as internuncial links between distal and jiroximal
portions of the grey matter. This latter was, by the exigencies of the method of
observation, often, if not always, thrown into a state of abnormal excitability ; and in
the neighbourhood of the section tliis state was that of hyperexcitability.

It will be gathered from what has been said that the question of the relations of
the fibres in the cord to the posterior roots can, in our opinion, only be arrived at by
the use of a method which allows of discrimination between events taking place in
the cord itself when, by means of complete anaesthesia, the activity of the reflex
mechanisms is subdued. Such a method is that of ascertaining the electromotive
changes in a portion of cord following stimulation of the afferent nerves, since the
animal may be so profoundly anaesthetised as to give little or no reflex muscular
contraction, and yet distinct effects can be detected in the cord. It is because the
conclusions to be drawn from our results clash so markedly with those of Mieschee,
WoROSCHiLOFF, &c , that we have thought it necessary to preface their introduction
into this section by this detailed criticism. This divergence in the conclusions and the
importance of the effective application of a method which besides the above advau-

ON THE MAMMALIAN NERVOUS SYSTEM. 425

tages has the greater one of furnishing us with the results of cord changes expressed
quantitatively, and, therefore, capable of strict comparison with one another, have
combined to cause us to devote a very large proportion of our experimental investi-
gations to the subject of the present and the succeeding chapters. The subject
matter of the present chapter becomes, thei-efore, an answer to the question, to what
extent are the electrical changes in the cord, and hence the nerve impulses, of which
these are the index, interrupted when between the stimulated nerve and the observed
region of the cord difterent intervening localised sections are made. The principle is,
therefore, that of the experiments carried out in LuDWici's laboratory, it is the index
alone which is novel.

Section 4. — Anatomical Relations of the Lumbar Neeve.

Since in all the experiments which bear upon the relations ot the spinal cord to its
nerves the sciatic was chosen, it is desirable at this stage to refer to the anatomical
relations which exist between this and the roots of the lumbar and sacral nerves.

These differ in the two sets of animals used. Cat and Khesus Monkey, and as we
are unable to find that the comparison of their anatomical and physiological relations
has been anywhere* set forth, we must proceed to examine the case of each species
of animal in a little detail.

In the Cat the photogi-aph of the plexus (Plate 35) shows that of the two branches
of the sciatic nerve [I.P. and E.P.), internal and external popliteal, the former derives
most of its fibres from the 6th, the latter from the 7th lumbar roots. Since the
portion of sciatic nerve which we used was that in the thigh, the fibres involved were
almost entirely those derived from the 6th and 7th lumbar roots, a few fibres coming
from the 1st sacral, and still fewer from the 5th lumbar. The obturator nerve derives
its fibres from branches of the 5th and 6th lumbar nerves, the anterior crural mainly
from the 5th, but to some slight extent from the 4th.

In the Monkey the relations are somewhat difi'erent, as will be seen by referring to
the photograph of the lumbar plexus of the Macacus rhesm (Plate 34). It will be
then foujid that the two branches of the sciatic nerve {LP. and E.P.) derive their
fibres from a more widespread origin. The majority of fibres come from the 5th, 6tli,
and 7th lumbar ; a few fibres from the 1st sacral join to form the trunk of the nerve,
but these appear to pass entirely into that branch which leaves the trunk high up and
supplies the hamstring muscles [lis.). The obturator nerve derives the majority of
its fibres from the 4th lumbar, receiving a few fibres also from the 5th, whilst the
anterior crural derives its fibres from the 4th, 3rd, and (?) 2nd in decreasing amount.

Roughly speaking, therefore, the relations of the lumbar plexus in the two species
of animals difler in this respect, that with the same number of lumbar nerves (7) the

* While this is passing through the press a very valuable paper, giving the anatomy of the plexus in
the Cat, has appeared, viz., LANCiLKY, 'Journal of Physiology,' vol. 12, No. 4, 1891.
MDCCCXCI. — B. 3 I

426 MESSRS. F. GOTCH AND V. HORSLEY

nuiiu supply of each nerve is, in Llie case of t lie Cat, one root lower than in that of the
Monkey.

The })hysiological connections of the roots were of importance to \is only as far as
the fibres in the sciatic nerve were concerned. In attempting- to ascertain these con-
nections it occurred to ns that the electrical method would furnish a valuable means
of ascertaining to what extent the fibres in any nerve trunk are derived from [)articular
roots, whether anterior or posterior. All that is necessary for the carrying out of such
investigation upon the sciatic nerve is to expose the cauda equina of an anajsthetised
animal, then to prepare the nerve, divide it, and connect its isolated central end with
the galvanometer. The various roots of the lower lumbar nerves are now divided at
their termination in the spinal cord, and their peripheral ends excited with the inter-
rupted induction current for a given time (5 seconds). If electrical effects are evoked
in the central end of the divided sciatic, then, obviously, the I'oot excited contains
nerve fibres which pass down the nerve as far as the region observed. With the
animal in a state of uniform narcosis, the comparative amounts of the deflection pro-
duced by applying to the different roots a stimulus of unvarying intensity and duration
indicate the relative amount of fibres, provided that the excitability of the various
roots is approximately the same. In this way we ascertained that the efl:ect evoked
in the popliteal region by exciting the peripheral ends of the cut anterior or posterior
roots of the 1st sacral nerve was, in the Cat, comparatively small, being only one-
sixth of that evoked by exciting those of the 7th or 6th lumbar, whilst excitation of
the 5th and 4th lumbar was followed in the Cat by very little effect at all in the
nerve. Phj^siological experiment thus bears out the anatomical details displayed by
dissection. One obvious use of the method which the few experiments which we have
carried out on this line suggests is, that it not only secures an analysis of the different
fibres which pass from a complicated plexus into a nerve, but places within the expe-
rimenter's grasp a method of determining to what extent the i-elations of the afferent
fibres of a nerve with the nerve roots are similar to those of the efferent fibres.

We will now turn to our experimental results made upon the cord itself.

Section 5. — The Operative Procedure involved in the Method of Sections

INTERVENING BETWEEN THE OBSERVED CoRD AND THE ExciTED NeRVES.

The experiments were made upon twenty-two Cats and three Monkeys, and were
all conducted in the following way. The cord of the anaesthetised animal was
exposed and divided in the lower dorsal region, the particular locality varying from
the levels of the 8th to that of the 11th dorsal vertebrse. It was then prepared for
4 centims. on the distal side of the section and connected at its cut and longitudinal
surface with the non-polarisable electrodes by cables, in the same manner as that
already described. (See Chapter III., Section 3.) The sciatic nerve was then
exposed, freed for some distance, ligatured, and divided. It was left in the muscles

ON THE MAMMALIAN NERVOUS SYSTEM. 427

in order to avoid cooling, &c., and only taken out of its muscular bed for purposes of
excitation. The large resting electromotive difference between the surface and cross
section of the cord having been balanced, an excitatory electrical effect in the opposite
direction to the difference was observed to accompany the stimulation of the nerve.
The amount of this effect, as indicated by the amount of the galvanometric excursion,
varied, as pi'eviously stated, not merely with the intensity and the duration of the
stimulus, but with the condition of anassthesia, &c., of the animal. If care be taken
to keep these factors as far as possible unchanged, the excitatory electrical effect thus
produced at each repetition of the experiment keeps very fairly uniform, as much so
in point of fact as in similar experiments iipon Mammalian nerve trunks. When we
had thus obtained a constant effect in the cord, an intervening section was made in
the lower dorsal or upper lumbar region by the method described in Chapter III.,
Section 2. Its position coincided, in some cases, with the lowest portion of the part
of the dorsal cord exposed for connection with the galvanometric electrodes, in the
majority of cases with a still lower level obtained by a fresh exposure of a small
portion of cord. It was found that it was not desirable to make this second exposure
when the cord was first prepared, since during the time lost in the preliminary experi-
ment, &c., the intervening cord at this point is apt to suffer.

The intervening section having been thus made, the nerve was excited under
precisely the same conditions as before, and the alteration, if any, in the amount of
the cord effect noted. In all such experiments the extent of the section of the cord,
and thus the part involved in the interruption, is capable of great variation, and the
results may with advantage be grouped in relation to the particular regions which
have been involved in it, thus following the exact and strictly logical method which,
under Ludwig's guidance, was such a characteristic feature in Woroschiloff's
treatise.

Section 6. — The Influence of Hemisection.

The results of a hemisection made between the jDortion investigated and the entry
of the stimulated nerve as regards its influence upon the electrical changes in tht
cord, which are evoked by the nerve stimulation, will be best seen by a glance
at the subjoined table, which gives the result of the comparative observations made in
the manner previou.sly indicated in two animals before and after the section. The
animal was in all cases carefully anaesthetised and kept well under the Influence of
the anaesthetic ; the stimulus was always in any two comparative results adjusted to
the same degree of intensity, and was of the same dui-ation, whilst, as far as possible,
errors due to variations incidental to the application of the stimulus, the condition of
the animal, and the galvanometric condition of the observed region of the cord, were
excluded, and observations in which contemporaneous changes in either of these three
conditions could be detected were rejected.

3 I 2

428

MESSRS. F. GOTCH AND V. HORSLBY

Result of Hemisection.
Excitation on same side as Section.

Cat (<0 (.S49) ....
Cat (b) (827) ....

Total ....
Ratio ....

Deflection.

Before.

After.

185
105
132

29
45
24

422

98

100 to 23

Excitation on opposite side to Section

Cat (a) (349) ....
Cat (b) (327) ....

Total ....
Ratio ....

Deflection.

Before.

After.

182

82

142

165
90
91

406

346

100 to 85

[n Cat (fl) the cord was divided and prepared at the lower border of the 9th dorsal
vertebra, and the section made at the 13th dorsal. In Cat (b) the cord was divided
and prepared at the 9 th dorsal. Both sciatic nerves were prepared, and first one and
then the other e.xcited in the manner previously described. The stimulus was in the
above cases that of the Helmholtz side- wire, 100 stimuli per second, continued for
5 seconds with an intensity just sufficient to evoke slight reflex effects in the lower
part of the cord, i.e., the secondary coil was placed at 500 or 1000. The ratio of the
sum of the two different columns indicates that an intervening hemisection, when on
the same side as the excited nerve, will diminish the cord effect 77 per cent., and
that when it is on the opposite side it also diminishes the effect, but to a very much
less extent, namely, only 15 per cent.

As far then as the nerve-to-cord effect is concerned, its production is very largely
dependent upon the unbroken integrity of the fibres and grey matter along that side
ot the spinal cord with which the stimulated nerve is connected.

ON THE MAMMALIAN NERVODS SYSTEM. 429

It is thus clear that in the etherised animal the excitatory electrical change is con-
ducted along the structures on the same side of the cord as the excited nerve, and if
the effect is a true indication of the passage of nei've impulses along nerve fibres in
that portion of the cord connected with the galvanometer, then these nerve impulses
themselves must be for the most part confined to that side of the cord which they
enter. Further, if we assume that the readings furnish us with data for a quanti-
tative comparison of the amounts of the variously awakened physiological activities
on the two sides of the cord, and thus of the number or intensity of the afferent
nerve impulses, then this preliminary set of observations would seem to indicate
that from four to five times as great a volume of excitatory change flows up the cord
on the side of entry of the nerves as on the other.

This result is fully in accoi'dance with the anatomical evidence previously referred
to, and with the results of spread obtained in the cord by excitation of its own
columns detailed in Chapter VIII. It is in absolute opposition with the before
mentioned results of MiESCHER and Woroschiloff.

The indications of these experiments ai-e confirmed by a large amount of collateral
evidence dependent upon the influence of section of the various columns. This will be
seen as we detail in succession the results of our different lesions.

Previous HemisectioiK

In order to ascertain whether this loss on the side of the lesion was independent of
all functional changes due to tbe sliock of the operation, &c., the hemisection was made,
in two instances, a considerable time (thirty days and more) before the experimental
investigation.

On the first animal. Cat (283), the hemisection had been made four months pre-
viously at the level of the 12th dorsal vertebra on the left side. Before commencing
the experiment the animal was carefully examined both in its normal condition and
under an anaesthetic. The left hind limb was dragged in walking, and showed
marked loss of muscular power. On allowing the feet to touch cold water move-
ments were started, the right hind leg being smartly drawn up, the left only after
a considerable delay. This and other tests applied to each foot indicated no impan-
ment of sensibility on the right side (that opposite to the lesion), but considerable
impairment of both tactile sensation and movements on the left, the same side. On
placing the animal under an anaesthetic the left knee jei'k was seen to be much
exaggerated, and it was possible to start clonic spasms in this limb.

For the actual experiment the cord was divided at the level of the 7th dorsal
vertebra and the upper end of tbe lower fragment prepared for connection with the
galvanometric electrodes. The electrical effect produced by exciting each of the two
prepared sciatic nerves was now observed, the stimulus used being of minimal and
maximal intensity respectively and consisting of the usual series of induction currents,
100 per second for 5 seconds.

430

MESSRS. F. UOTCH AXU V. IIOUSLHY

Secondary Coil 500 (minimal).

Cord efFect ....

Excitation of right

nerve.
Side opposite lesion.

Excitation of loft

nerve.

Side of lesion.

90
90
95

20

28
45

Average 92

Average 31

The electrical change is thus three time.s as large when evoked by stimulation of
the nerve on the side opposite to the lesion, that is, the right side, as on the side (left)
of the division. Hence, the major part of the change must be cut off' on the left side
by the lesion,* that is to say, the nerve impulses, of which the change is an index, are
unable to ascend the cord on the left side to the observed region in the neighbour-
hood of the 7th dorsal, owing to the hemisection at the 13th dorsal interrupting their
path. If we suppose that the actual effect observed by stimidating the left nerve is
due to impulses which normally cross the cord, whilst that observed by stimulating
the right is due to impulses which do not cross, then the interruption has cut down
the effect which the noi-mal excitation of the left sciatic nerve would produce from
92 + 31 =123 to 31, i.e., 75 per cent.

When the stimulus was four times the intensity of the above, that is, sufficient to
evoke sti'ong reflex contractions, a .similar disproportion was found to exist although
it was not so marked.

This is shown by the following table : —

Secondary Coil 2000.

Excitation of right
nerve.

Excitation of the left
nerve.

340
215
220

170
160
135

Average 258

Average 155

* In this animal microscopical examination of the cord showed thai the lesion involved all the left
lateral and posterior columns, the whole grey matter, and, except its median edge, the anterior column.
Of ascending degeneration there was beautifully marked destruction of the left postero-median column, as
also of the direct cerebellar and antero-lateral tracts, and a few degenerated fibres in the right postero-
median column. Of descending degeneration there was, immediately below the lesion, degeneration in
the left lateral pyramidal tract and also- in left the anterior column.

ON THE MAMMALIAN NERVOUS SYSTEM. 431

The effect even with this strong stiniuhis is nearly twice as large when ev'oked by
the nerve opposite to the side of the lesion. Since, however, a considerable effect
is evoked by stimulation of the nei've on the same side, the use of the stronger
stimulus has either produced more impulses strong enough to cross the cord, or it has
caused a more complete discharge of the central mechanisms in the grey matter.

The other animal. Cat (259), had a section made upon it at the 10th dorsal vertebra
thirty days before the observations were to be made. The section involved the left
lateral and the majority of the left posterior column. The subsequent examination
of the cord showed that the section involved on the left side the lateral column,
the ventral portion of the anterior column of the grey matter, and the dorsal half of
the posterior column ; on the right side the posterior median column. The ascending
degenerative changes were at the 5th dorsal traceable on the left side to a marked
degree in the direct cerebellar and antero-lateral tracts, and in the posterior root zone
and posterior median columns. A much less extensive degeneration was found to
have occurred in the posterior median column of the right side.

The animal, when examined before the experimental observations, showed spastic
paralysis of the left hind limb, and marked diminution in the tactile sensibility of both
hind limb and trunk on the left side. The knee jerks were equally present on the two
sides. The experimental investigation, as far as it relates to the present question,
involved the section of the cord at the 4th dorsal (this was accompanied hy prolonged
tetanus of lower limb muscles and those of the tail), the preparation of the upper end
of the lower fragment and the excitation of the two sciatic nerves. The excitation of
the right nerve evoked effects of 43 and 56, the left of 15 and 22. The sum of these
readings is 99 for the right, and 37 for the left side. If, as in the previous experi-
ment, we suppose that the total sum of the two effects roughly represents the full
responsive change of either nerve excitation, then the effect on the side of the chief
lesion (hemisection) is reduced from that in the ratio of 136 to 37, i.e., 100 to 27, that
is, it has been reduced 73 per cent, through the hemisection.

The result of these experiments is to show that the excitatory changes in the cord,
evoked by stimulation of the sciatic, are largely limited to the same half of the
cord as the excited nerve, the amount of limitation being such that three times the
effect, and presumably three times the number of nerve impulses, proceed up the cord
on the side of entry than that which is evoked by impulses either crossing in the cord,
or freshly generated by cells in response to the arrival of impulses in the grey matter
on the opposite side.

Our method, however, enables us to determine not merely that the excitatory
effects in the cord are to a great extent limited to the side of the entering nerves, but
also that the effects are mainly dependent upon the unbroken integrity of particular
columns of fibres on this side. To observations on this point we now pass.

432 MESSRS. F. GOTCH AND V. HORSLKY

Section 7.- — Tiik Influence of Section of both Postkiuor Columns.
A. Section of both Posterior Coluvnis,

The aiiiitoniical fxiilciice clearly shows that there is an unbroken path between
some of the entering afferent fibres and the fibres in the posterior column, and that
these subsequently pass more or less completely into the posterior median column from
the posterior root zone. It was therefore essential to ascertain to what extent the
cord effect was de])endent upon the presence of these fibres. This we have
endeavoured to do by first ascertaining to what extent the effect is due to the
unbroken integrity of both posterior columns. A remarkable incident in the course of
our experiments led us early to suspect that by far the greatest part of the effect,
whether crossed or direct, is conducted along these tracts. In the course of the
operation for one experiment, whilst opening up tlie neural canal, the point of the
bone forceps slipped, and so slightly bruised the postei'ior columns that the injury
could not be detected at the time. The preparation was proceeded with, and
the cord divided at the 8th dorsal. The upper end of the lower fragment was then
connected in the usual manner with the galvanometric electrodes, and the sciatic
nerves exposed for excitation. Only extremely small cord effects were, however,
evoked by the excitation. We then discovered the bruise and determined to make a
fresh section below the injury, exposing for this purpose more cord ; on connecting
this fresh portion with the galvanometric electrodes and exciting the nerve, an effect
of 175 scale was obtained.

We now made a direct experiment along the lines thus indicated.

The cord was exposed in a Cat, and divided at the 10th dorsal vertebra,
prepared for 4 centims. and connected with the electrodes by its surface and
cross section. Excitation of the left sciatic nerve, with the secondary coil at
2000 (1 Daniell in primary) for 5 seconds (500 stimuli), produced cord effects indi-
cated by a deflection of 75 scale. The posterior columns were then divided at the
level of the upper border of the 12th dorsal vertebra, and a similar excitation was
accompanied by a deflection of oidy 12 scale. As a control, the cord was now
completely divided a.t this latter level, and the part distant from the new section was
connected with the galvanometer, when a similar excitation produced a deflection in
one case of 64, in another of 75 scale, thus indicating that the great diminution of the
effect was due to the local operative interference with the posterior columns, and not to
any general alteration in the condition of the whole cord.

A similar result was obtained in another animal (Cat), in which the cord was
prepared and connected with the galvanometer at the level of the 10th dorsal
vertebra, and the cauda equina having been exposed, the central end of a cut
posterior root was excited, instead of the sciatic nerve. In this case the cord effect,

ON THE MAMMALIAN NERVOUS SYSTEM.

433

which was accompanied by reflex movements, was 210 with the cohnnns intact, and
after tlieir division at the level of the 12th dorsal vertebra, sank to 36.

In the Monkey this marked diminution was also found. Thus in a Rhesus the cord
was divided at the lower border of the 6th dorsal vertebra, and prepared as before.
Excitation of the right sciatic nerve evoked a cord effect of 82 ; excitation of
the left sciatic an effect of 60 ; after division of both posterior columns at the level
of the lower border of the 11th dorsal vertebra, excitation of the right sciatic
produced an effect of only 18, and that of the left sciatic an effect so small as to
be indicated by a mere trace, viz., 2.

It need scarcely be pointed out that the presence or absence of concomitant reflex
movements as an index of the awakened corpuscular activity of the cord is a most
important factor to be borne in mind, but the diminution occurs even when a more
pronounced stimulus of greater duration is used, as is shown by the following
experiment on a Rhesus Monkey. The cord was divided at the 8th dorsal vertebra,
and prepared for galvanometric observation. The right sciatic was excited and
evoked a reflex movement and large eftects, indicated in repeated observations by
deflections of 210, 300, and 308. The left sciatic gave effects of 230, 250. After
section of both posterior columns, the effect sank to 55 and 58 following excitation of
the right nerve, and to 90 and 60 following excitation of the left nerve.

If now we sum up the various readings (15) obtained in two Cats and two Monkeys,
we find, as expressed in the following table, that the intervening section reduces the
massed results in the ratio of 100 to 22, that is, both the posterior columns are
directly concerned with the production of 78 per cent, of the total effect, and the
transmission of the same percentage of impulses.

Table of Readings. Section of Posterior Columns.

Cat . . .
„ . . .

»> • • •

Monkey

Cord iuvestigated at

Deflection
before section.

Deflection
after section.

lOth dorsal vertebra

10th "to 11 til dorsal

7tli dorsal (average of six readings) .
8th dorsal

76
64
210
90
210
300
308
230
250

12

12
39

18

65-

68

57

90

60

1743

389

Ratio 100 to 22.

This relationship is confirmed by a control expei'iment of the reverse kind in which a

MDCCCXCT. — B. 3 K

434

MESSRS. F. GOTCH AND V. IIORSLEY

section of thf part of the cord ventral to tlie ])osterior columns was made. The cord
was prcpaivd In a Cat by division at the 8th dorsal vertebra ; the upper end of the
lower fragment connected witli the galvanometer, and the left sciatic stimulated with
three diflex'cnt strengths of stimulus, and the effect noted in each case. An inter-
vening section was then made at the 10th dorsal, so as to divide the \vhole region of
the cord lying ventrally to the posterior commissure and the commissure itself, the
result being that the minimal effect remained unaltered, whereas the other effects
were diminished in the ratio of 100 to 76, as indicated in the subjoined table.

Strcnn-th of
stimulus.

Effect before
section.

Effect'aftcr
section.

500
1000
2000

35

35

65/ ^■''

Ratio 100 to 76.

This experiment, whilst it establishes the importance of the posterior colunuis b}''
furnishing the control experiment of leaving these tracts and excluding the others, is
at the same time a remarkable instance of the quantitative precision of the method,
since in independent experiments lesions effected quite differently show in the one
case that the posterior columns are concerned with 78 per cent., in the other with
76 per cent, of the effect. As the experimental details accumulate, this steady
quantitative relation will be found to vary wonderfully little, considering the nature
of the structure investigated and the difBculties connected with the method.

The experiment just alluded to also indicates a very important fact which will be
refen-ed to in detail in the concluding section of this work, namely, that when the
stimulus is minimal the nerve impulses are almost entirely confined to the direct
path ; in other words, this path is that of least resistance. It is necessary for the
stimulus to be of a certain intensity before nerve impulses can be started of sufficient
volume to break through corpuscular structures and so enter on indirect paths. This
is indicated by the fact that no change in the deflection could be observed in the case
of the weak initial stimulus to follow the section of the grey matter ; presumably,
therefore, the cord effect was due, in this case, to afferent impulses passing up. the
direct path.

A further extension of the experiment just described may be now referred to,
namely, the influence of dividing in addition one posterior column. After the section
of the anterior or ventral portion of the cord the left posterior column was divided at
the 10th dorsal vertebra ; on now stimulating the left sciatic nerve an effect of 8 only
was obtained, this observation being repeated several times. The importance of the
posterior column on the side of the nerve is thus forcibly brought into view.

ox THE ilAMMALIAN NERVOUS SYSTEM.

435

B. Section of the Posterior Column of the same side as the Nerve Excited.

The marked diminution in tlie eflect obtained by hemisection on the same side as
the nei've in the one case, and by section of both posterior columns in the other'
would seem to imply that the posterior column on the same side, which is common to
both these lesions, is the fundamental structure concerned. We shall see that there
is every reason to suppose that it is concerned with 60 per cent, of the total eftect,
that is, with that proportion of the number of nerve impulses, the remainder being
pretty equally divided between the lateral column of the same side and the posterior
of the opposite side. Tt will be seen that the direct evidence of the result of section
of one posterior column is somewhat complicated by changes in excitability which
appear to affect the remaining columns.

The results were observed in two Monkeys and five Cats, as shown in the following
Tables :—

Section of Posterior Column on Same Side as Nerve.

Monkey (Macacus
rhesus)

IMonkey (Macacus

rhestis)

Cord observed at

Before section.

Section at

After section.

Deflection.

Deflection.

8th dorsal vertebra

A dorsal vertebra
(tlie 8tli probably)

130
210

304
32
60
78

lOtb dorsal vertebra
11th dorsal vertebra

38
68
100
12
21
14

814

252

Ratio

of 100 to 31.

In these Monkeys section of the posterior column on the same side reduced the
cord effect by 69 per cent., a very notable amount. Such a large reduction was not
often obtained in the Cat ; and occasionally in this animal comparativel}^ insignificant
reductions were seen. In such instances, however, the nerve stimulation was
observed to evoke violent reflex movements, and, as will lie seen in considering the
result of section of the opposite posterior column, the opposite uninjured side of the
cord became evidently hyperexcitable.

3 K 2

43(5

MESSRS. F. GOTCH AND V. IIORSLEY

Si':i'TioN of posterior columu on same side as nerve at 11th to I2tli dorsal vertehra,

Cord obsorved at
8tli dorsal.

Deflection buforo
section.

Deflection after
section.

I. Cat

II. Cat

HI. Cat

iV. Cat

V. Cat

90

100

75

7(i

a:?

210
185
185

35
38
34

45
24

115
90

121

1004

508

Ratio of 100 to 50.

The above experiments made upon five Cats give results, which, as is seen, differ
considerably in several instances, the deflection before section being in some cases
four times as great as that after section ; in others not twice as large. The latter is,
however, the case when owing to some circumstance the stimulus evoked very large
effects (see Cats IV. and V.) including strong reflex movements. If, however, we mass
together all the results obtained in the two sets of animals, the average reduction of
eflects due to section of one posterior column would amount to about 60 per cent,
of the total effect obtained in the normal condition. Hence it would seem that one
posterior columu is the main channel by which the entering nervous impulses proceed
from the lumbar to the dorsal cord.

The difliculties involved in this series of experiments will be still more evident
when we turn to the influence of the section of only the opposite posterior column
upon the cord effect.

C Section of the Posterior Column of the Opposite Side to the Nerve Excited.

In these experiments we are for the first time confronted with the remarkable fact
that in consequence of an intei'vening section the stimulation of the nerve is some-
times followed by an increased effect. (See also Chapter VIIL, Section 9, B., p. 395.)
Since the section has cut off at least some channels of communication between the
lower and upper parts of the cord, and presumably, therefore, has blocked the paths
of some, even though few, nerve impulses, siich a rise must be attributed to the
greater intensity of the effect evoked in the remaining channels, whether nerve fibres
or nerve cells, and the cord must therefore be considered to be hyperexcitable.

This alteration in excitability is probably associated with the phenomenon of hyper-
sesthesia observed by Fodkra and Schiff to follow operative interfence with the

ox THE MAMMALIAN NERVOUS SYSTEM. 437

posterior columns. That it affects the cord on the side of the lesiou is probable from
the results just given, but the direct proof is wanting since the increased intensity of
the effect is masked by the large diminution caused by the severance of the dhect
tract in the posterior column of the same side. There is, however, no such sever-
ance in the case of the opposite side, and the increased excitability is plainly seen>
as evidenced by the cases marked with an asterisk, in which the section was followed
by actual increase iu the cord efiect. Whatever may be the meaning of this it
undoubtedly is a most important factor to be taken into consideration in all experi-
ments the general plan of which consists in first causing definite lesions, and then
examining the influence of such. It will be obvious too, that in proportion as the
antesthetic is removed so much the more marked must this change in excitability
become. In the experiment upon conduction already referred to, made by Miescheb,
in which the reflex effect on the blood pressure was taken as the index of the passage
of afferent impulses through a block in the cord, any change of this kind could not
be allowed for as owing to the curarisation of the animal, the observation of con-
comitant reflex movements was impossible. We venture to think that the increased
effect obtained by Miescher on the side of the complete section in his division
of every pai't of the cord except one latei-al column, and taken by him to mean
that the afferent fibres crossed into the lateral of the opposite side, was due to some
such increased excitability in the lower fragment of cord on the side of the main
lesion, the fibres of the lateral tract connecting this fragment with that above as
internuncial fibres.

It is obviously impossible to put this factor completely out of court. The cord
effect with which we are dealing is a measure at once of the intensity and the number
of all transmitted excitatory processes, and the influence of any particular section can
therefore only be judged of by taking the average of all the readings, high and low.
It might be thought that by pushing the anaesthesia and by decreasing the strength
of the stimulus the effect would always become uncomplicated. Although anaesthesia
lessens the chance of its occurrence it would appear that this increase in the cord
effect will sometimes occur even under these conditions, and it is then evidently
dependent upon the character of the particular preparation used. Its presence or
absence in these cases may, however, be not merely due to the idiosyncracy of the
animal, but to minute differences in tlie extent of the lesion in certain instances.

The general result of experiments on four Cats and two Monkeys is as follows ; —

438

MESSRS. F. GOTCH AND V. IIORSLEY

Ekfioct of Sectitm ol" Posterior Column at the 10th to the 12th Dorsal Vcrtchra on
the Opposite Side to the Nerve Excited.

Cat (196) ....

Cat (250) ....
Cat (377) ....

Cat (349) .....
Total . . .

Cord observed at

Nerve excited before
section.

Nerve excited after
section.

lOtli dorsal

8th dorsal
9th dorsal

9th dorsal

105

128 111

107

78

78

125

182

62

87*
90*
91

180 1^"

674

470

Ratio 100 to 81

Monkey (281) . .
Monkey ....

Total . . .

Cord observed at

Nerve excited before
section.

Nerve excited after
section.

8th dorsal
7th dorsal

155

230

260

66

82

m

150
170
199
105*

87*
120*

878

831

Ratio 100 to 94

From these results it is seen that the cord effect is reduced by the section, and
that this reduction in the observed cases was more marked in the Cat than in the
Monkey. If we take the two sets of experiments together, as in the previous
results of the division of the posterior column of the same side, then the reduction is
found to be from 100 to 87, that is, 13 per cent.

This evidence therefore points to a crossing of effects from one side of the cord to
the other as probable, such crossing and the ultimate conveyance of the impulses
being moi-e or less bound up with the posterior column on the opposite side to the
stimulated nerve, to the extent of 13 per cent, in the animals we have as yet
examined.

ON THE MAMMALIAN NERVOUS SYSTEM. 439

D. Previous Section of Posterior Columns some time Previous to Observation.

In one Cat (251) we divided the posteiior columns at the level of the 10th dorsal
vertebra four weeks before the experiments. When the animal (Cat) was examined
it was noticed that the water test showed diminished sensibility in both hind feet,
since the animal could be ])laced with its hind paws in water without exhibiting any
movements of withdrawal. There were, however, vagiie symptoms that the two sides
were not equal in the degree of sensibility they still possessed. This inequality
showed itself iu the knee jerks, the left being much feebler than the right. On
placing the annual under ether, however, the left knee jerk was found to be then very
much exaggerated, and the left hind limb was easily sent into a condition of clonic
spasm.

The animal wns first experimented upon liy exposing and exciting the cord, and
observing the changes in the nerves ; the result of this experiment will be given in
Chapter X. It was then used for the present series of ex[)eriments, the cord being
originally divided at the 5th dorsal vertebra, and the upper end of the lower fragment
prepared for galvanometric observation at about the 7th dorsal vertebra.

The left sciatic, when excited, with a minimal and a stronger stimulation, evoked
effects of 10 and 20 ; the right, when similarly excited, evoked cord effects of
8 and 70.

The cord, after the death of the animal, was examined histologically, when it was
found that the section had not been complete on tlie right side, a good many fibres
having been left intact. On the left side it was complete, and at the level of the 7th
dorsal, on microscopical examination, a sickle-shaped patch of degeneration could be
seen in the postero-external column on each side, that on the left being very well
marked, whilst that on the right was small. The same diiference between the
degeneration on the two sides could be seen at the level of the 4 th cervical, where
the degenerated area occupied, on the left side, a conspicuous portion in the middle
third of the posterior median colunm. The discrepancy between the eftects obtained
between the two nerves is thus cleared up by the minute examination of the extent
ot the lesion. It furnishes striking evidence of the accuracy with which the electrical
method can gauge the integrity of through tracts of fibres, but brings out a disadvan-
tage, namely, that it involves the death of the animal ; hence, when lesions, &c.,
are made at the time of the experiment, it is impossible to be perfectly certain of
their extent, since the degeneration method, which alone would give absolute indica-
tions, cannot then be used.

Section 8. — Influence of Section of the Lateral Column.

Of the remaining columns in the cord, the lateral and anterior, it is only the former
which appear to bear any relation to the production of the electrical effect in the

440 MESSRS. F. GOTCH AND V. HORSLET

cord wlieu the .sciatic iicive is excited. This is shown by tlic fact that the cord effect
is in no way modified by an interruption in the anterior columns which involves their
complete severance, caused by a section between tlie entering nerve and the observed
reo-ion. Such a section is most easily made In the isolated fragment of cord near
its central attachment, since this portion is raised from its bed. This is in strict
accordance with the fact alluded to in Chapter VIIL, that excitation of the one end
of the anterior column evokes no electrical effects in a distant portion of the cord.

It is otherwise with the lateral columns which have distinct relations with
electrical effects in the cord, since, as shown in Chapter VIIL, their stimulation
evokes marked cord effects. It does not follow, however, that this relation is one
which comprises the afferent nerves, for we know that the lateral column contains the
main path for the efferent cord tracts, as is shown by the results of cortical excitation,
as set forth in Chapter V., &c. The conclusion at which so many observers have
arrived, that the main afferent path, whether crossed or uncrossed, is situated in the
lateral column, is clearly not substantiated by any evidence offered by the employment
of the present method, since the integrity of the two posterior columns is evidently
essential for the production of at least 70 per cent, of the total effect. We shall see
that the extent to which the two lateral columns are related to the afferent effect is
small, being approximately only a fourth part of this amount, and that this is almost
entirely confined to the lateral column on the same side as the afferent nerves excited.
The results may be best displayed as follows : —

A. Effect of Section of tJie Lateral Column oti the same side as the Nerve Excited.

The following experiments made upon four Cats show that the lateral column of
the same side as the excited nerve, If divided, reduces the cord electrical effect evoked
by excitation of the nerve to an extent the average of which is 20 per cent., that is,
from 100 to 80.

ON THE MAMMALIAN NERVOUS SYSTEM.

441

Section of Lateral Column on the same Side as the Nerve Stimulated.

Cat ri96) .
Cat (327) .

Cat (34G) .

Cord observed in
region of

Effect before
section.

Section at

Effect after
section.

lOtli dorsal vertebra .
9th „

8th „

105
105
132
90
110

1st lumbar
1st lumbal-

1st lumbar

115*
76
83
(50

105

542

439

Ratio 100 to 80.

The inference is that the integrity of the lateral column of the same side is
connected with 20 per cent, of the cord effect. Additional ground for this conclusion
is afforded by the opposite sectional method practised after the manner of Miescher
and WoEOSCHiLOFF, which, excluding the lateral column, divides everything else:
the knife being placed perpendicularly through the cord from dorsal to ventral surface
at the junction of the posterior and lateral columns, and then the cut made obliquely
inwards towards the centre, and continued so as to come out on the opposite side.
All the structures, with the exception of one latei'al column, ami the contiguous
portion of the lateral horn are thereby severed. In some instances this operation was
done piecemeal, first one and then another structure being divided.

The result of such severance of both posterior and one lateral colunm is to sliow
that an electrical effect in the cord can still be obtained by stimulating the nerve
on the same side as the uninjured lateral column. Its amount, as compared with
the total change obtained before the section, is, however, greatly reduced, as is
shown in the following Table, comprising experiments made on five Cats.

* It will be noted that in Cat 19G there ensued an increased effect after tlie section due to a rise in
excitability.

MDCCCXCl. — B. 8 L

442 MESSRS. F. GOTCII ANJ) V. IIORSLEY

Section of butli Posterior Coluinns and Lateral on opposite Side to Nerve.

(195) Cat

(196) Cat
(256) Cat
(3-27) Cat
(344) Cat

9th doi'sal vtrlobia
8th

Effect hpforo
section.

210
86
78

142
6

122

152

856

Intensity of
stimulus.

4000

4000
1000
1000
500
1000
2000

Effect after
sfection.

23
35

28
35
27

162

Katiu 100 to 19.

That is to say, this remaining effect represents 19/100 of the whole. This result
tallies with that obtained with the other method, which showed that the section of this
lateral only reduced the effect by about 20 per cent. This reduction in the cord
electrical change is thus a reduction in the number or intensity of the afferent Impulses
Avhich po.ss up the cord when the nerve is excited. Hence, as far as the method goes,
it would appear that the lateral column on the side of the stimulated nerve furnishes
a path, whether internuncial or not, we cannot say, but, at any rate, an indir'ect one
for the transmission of 20 per cent, of the nerve impulses.

B. Effect of Section of the Lateral Column on the opposite side to the Nerve Excited.

The contradiction which exists between the results of the present research and the
interpretation given to their experiments by Miescher and Woroschiloff is brought
into the strongest prominence by the consideration of the present group of results. If
the interpretation of these physiologists is wan-anted by their data, and is of general
application, then we should expect that the integrity or otherwise of the lateral column
opposite the entry of a stimulated nerve would have the preponderating influence u])on
the passage. of those nerve impulses whicli have entered the cord by the nerve, and
consequently upon the amount of the electrical change in the observed portion of cord.
We have, however, already accounted for 95 per cent, of the effect by showing that
the amount is dependent upon the integrity of the other columns, and this fact alone
is sufficient to show what a small share the integrity of the opposite lateral column
has in providing channels for the conduction of ascending impulses.

We have made expei-iments upon five different animals, the results being separated
into those in which the lateral column was divided on the opposite side to the
nerve, and those in which all structures, except the opposite lateral column, were
cut

ON THE MAMMALIAN NERVOUS SYSTEM.

44J

Influence of Section of the Lateral Column on the opposite Side to the Excited

Nerve on the Cord Effect.

1

(196) Cat .
(327) Cat .

(346) Cat .

Cord divided and
observed at

Effect before
sectiou.

Stimulus.

Effect alter
section.

lOtli dorsal \ei'tebra
9tU

9th

■

86

82

142

110

110

4000
500
1000
1000
2000

100

7;^

12.5

63

112

5.30

478

Ratio 100 to 90.

It will be noticed that in only one case (346) was there any marked reduction, and
that this did not occur when the intensity of the stimulus was increased. Instead of
reducing the effect the section in two cases increased it. Tlie average reduction of
1 0 per cent, is probably too high, owing to some unknown factor being present to
cause the exceptional low reading. This presumption m rendered more probable by
the next set of results, the converse experiments.

Section of all Columns except Lateral on Opposite Side to Excited Nerve.

(196) Cat .
(2.56) Cat .
(327) Cat .

(344) Cat .

Cord divided aud
observed at

Effect before
section.

Stimulus.

Effect after
section.

lOth dorsal vertebra .
Bth „ „
9th „

8th „

105
90

142

105
74
93

135

4000
1000
1000
500
500
1000
2000

10
0
5
5
3

15
3

734

41

Ratio 100 to 6.

These results show that at any rate under the conditions which were present in
these experiments, those, namely, of moderate narcosis, the lateral column of the cord
opposite to the nerve excited can transmit only 6 per cent, of the nerve impulses which
produce the electrical change in the cord when all the other structures are divided.
To what extent it can act as a bridge when the ether is removed, as in Mieschek

•6 i.

4 I I MESSRS. V. GOTCH AND V. HORST.KY

aiul WokoschilOFf's experiment \vc luive not yet liad :iii <i|)|)nrtniiity of ol)serving.
The experiments woulil have to he done under curare to avoid (lie <h'agging on the
cord by the convulsive movements of the animal and without an;estliesia, analgesia
being provided for by section of the peduncles. They are, therefore, of a very special
Ivind, and involve much consideration before embarking upon them.

Skction 9. — Summary.

We will now sum up tlie results of the foregoing experiments, pointing out at the
same time the deficiencies as well as the advantages of the method employed.

I. Electrical changes in the lower dorsal region of the cord are readily evoked by
excitation of the sciatic nerve, or the posterior roots of the lumbar plexus.

II. These changes are dependent in the anoesthetised animal on the integrity of
particular columns of nerve fibres stretching between the region observed and the
neighbourhood of the entering nerves.

III. The total amount of electrical change is an indication of the amount of nerve
eneigy flowing up the cord from the stimulated nerve. If the total amount be repre-
sented as 100, then the following numbers represent approximately the amount which
flows up each column and which is thus interrupted by its section : —

Posterior column same side GO

Posterior column opposite side . . . . 15

Lateral column same side 20

Lateral column opposite side 5

IV. The flow of nerve energy up the cord is thus mainly unilateral, 80 per cent.
being transmitted in the channels on the saine side as the entering nerve.

V. The comparatively small amount of nerve energy which crosses into the opposite
side of the cord is almost entirely localised in the posterior column of that side.

These conclusions are. it will be observed, simply confirmatory of those already
indicated in the preceding chapter, in which it was pointed out that no electrical
evidence existed of any crossing of ascending impulses in the cord from the fibres
in one lateral column to those of other columns ; and that the peculiarity of the
ascending impulses in the cord is (a) the direct path they take up the posterior
column of the same side, and (6) the indirect path by spread from the posterior
column, first into the lateral column of its own side, and then across into the
posterior column of the ojiposite side.

Finally, our experiments do not show the path taken by sensory impulses ; it
is possible, though hardly conceivable, that these are essentially different in quality
to the ones we have been studying. All we can say is that when nei've impulses
are evoked in the afferent fibres of the sciatic nerve, and proceeding up these reach
the cord, they apparently find several paths open to them. These paths, howev^er,

0\ THE MAMMALTAX NERVOUS SYSTEM. 445

are either not equally numerous, or are not equally easy of passage ; the posterior
cohimn of the same side offers special facilities for the passage of the inapulses, the
lateral column of the same side offers facilities greater than the posterior column
of the opposite side, but both, whilst far inferior in this respect to the posterior
column of tlie same side, are far superior to the remaining lateral column of the
opposite side. This last, at least in the narcotised animal, offers practically no
facilities for such passage.

It is, perhaps, unnecessary to draw attention to the circumstance that the above
results are only verified in the case of the lower dorsal region of the cord in the Cat,
and to some extent in the Monkey.

The principal deficiencies of the present method have been indicated in tliis and
previous chapters, but may now be summarised. Tliey are connected —

(a.) With tlie character of the nerve impulses, which, being due to electrical
excitation, are more intense, and possibly different in quality, to those which are
generated in peripheral sensory end organs.

(6.) With the necessity of insulating the observed portion of cord, to ensure the
observation of localised effects in it, this being accomplished by a severe operation
which entails the death of the animal at the close of the experiments.

(c.) With the limited anatomical scope of the method as at px-esent used, the highest
point which we have reached being the mid-dorsal region. The shock of the exposure
above this has rendered our experiments, carried out higher than this, unsuccessful.

The chief merits of the method are —

(«.) That the changes investigated are the electrical excitatory processes in the cord
itself severed from the encephalon, and are free, therefore, from admixture with
cerebral effects, and are independent of the reflected outcome of such effects in the
muscles.

(h.) That the changes are so definite as to admit of comparison as to their quantity,
and thus of a quantitative estimate of the nerve energy transmitted in the cord
under different conditions.

(c.) That this quantitative character of the results enables a comparison to be made
of the effects of nerve energy which are dependent respectively on the integrity of the
different parts of the cord.

The conclusions to which we have thus arrived will receive additional confirmation
from the results of experiments in which the impulses in the nerves are made the
subject of investigation by the electrical method, such impulses being aroused by
excitation of the different columns in the cord.

These will be detailed in tlie next chapter, at the end of which a general review
of the whole question of conduction in the cord, as elucidated by our method, will
be given.

44G MKSSRS, V. COTCIl AND V IIOKSMIY

CHAPTKR X.— ELECTRICAL CHANGES IN THE NERVES FOLLOWING EXCITATION OF

THE SPINAL CORD.

Section 1. — Introductory,

Section 2. — -Plan of experiments and preliminary observations.

Section 3. — Electrical effects in sciatic nerve following excitation of different columns of the cord.

Section 4.— Electi'ical effects in afferent nerves followini;- cxc-itation of the cord :

(1.) Posterior roots.

(2.) Sciatic nerve after section of anterioi' roots.
Section 5. — Electrical effects in efferent nerves following excilution of tlie cord.
Section 6. — Influence upon tlio effects in the nerve of intervening sections of the cord.

A. Influence of hemiseetiou.

B. Influence of section of the posterior coluiun on the sumo side.

C. Influence of section of the posterior column on the opposite side.

D. Influence of section of both columns.

E. Influence of section of the lateral columns.
Section 7.^Sammary and conclusions.

-Section 1. — Inteoductory.

lu the foregoing cliaptei- the re.sults have been given of the observation of a new
index of cord activity, that of noting tlie accompanying electrical eflrect.s when the
afferent nerves were excited together with the influence upon these of a series of
intervening sections.

It will be seen that the plan of the whole of that research was upon the lines of
previous investigations, except as regards the index employed ; in the experiments to
be detailed in this section both the method and plan of experiment are entirely novel,
since conclusions can only be arrived at by the evidence afforded by the electrical
effects as indicative of the presence or absence of excitatory changes. Tlie novelty
consists in this, that we are able to detect excitatory electrical effects in the
issuing nerves when the cord is stimulated; not merely in the nerves them-
selves, but in their roots. Now as far as the efferent motor roots and motor portion
of the mixed nerves are concerned, the method whilst giving us valuable information
has obvious relations with the data which have already been obtained by the graphic
method of recording the muscular contractions ; but as far as the afferent sensory
nerves are concerned, it introduces us to previously unknown relationships, for there
Is no indication other than an electrical one at present known, which can detect the
passage of nerve impulses from the cord into afferent tracts, since such passage is opposed
in direction to that of the normal tiansmission as usually understood. It has been
already stated that one of the most valuable results of the discovery of the excitatory
electrical change was that set forth by du Bois-Eeymond, as proving the propagation
of excitatory effects in both directions along a continuous tract of either afferent or
efferent nerve fibres, this being evidenced by the presence of the electrical change at

ON THE J^IAMMALIAN NERVOUS SYSTEM. 447

both ends of a nerve when the middle portion is stimulated. We have seen that
similar evidence in the case of the nerve fibres in the cord shows that in them also the
eifect is propagated from the stimulated region indifferently either towards the centre
or the periphery, and it follows from the combination of these two, that if any fibres
in a nerve are directly continuous with those in the cord, the excitatory changes
following stimulation of these must be propagated along the whole length of the
continuous strand and may thus travel out into the fibres of the nerve. Direct
continuity has been shown to exist by the degeneration method between the
fibres of the posterior root and those of certain portions of the posterior
columns, hence it is not surprising that the stimulation of the posterior columns in
the cord should cause electrical clumges in the ]josterior root due to the arrival in the
fibres of this structure of excitatory processes transmitted from the cord down tracts
which, owing to their terminal relations, are usually believed to be solely ascending
ones. There is nothing to warrant the belief that these descending impulses in
afferent nerves, evoked by excitation of the columns in the spinal cord, are in
themselves different from those which are evoked by similar methods of excitation
when applied to the peripheral parts of the afferent nerves, and which are propagated
in the natural direction, since it seems to be fundamental as regards nerve conduction
that fibres can conduct equally in either direction.

The fact that in one group of nerve fibres (afferent) the starting platforms are at the
peripheral end, and the receiving termini at the central, whilst in the other (effei'ent) the
positions are reversed, and that, in consequence, what are called normal nerve
impulses proceed in the afferent direction in tlie one group and in the efferent in the
other, apparently has not, physiologically or structurally, altered the fibres (polarised
them in any way) so as to make conduction in the 0T\e direction more difiicult than in
the other. As far, then, as the directly continuous afferent fibres are concerned,
there is no difficulty either in obtaining electrical results in the posterior roots on
exciting their direct prolongations in the cord, or in interpreting these as indicative
of the passage of nerve impulses.

A much more complicated condition must, however, be now referred to, that,
namely, of the nerve fibres which are only indirectly continuous with others in the
cord, there being interposed in their path corpuscular elements and unknown
channels.

The simplest of these are the efferent (motor) nerves, and it has already been
stated that, as regards these, Chapter IX., Section 2, the corpuscular connection is of
such a character that, whilst allowing ihe passage of impulses from the cord to the
nerves, it appears to completely block the passage of impulses from the nerves
into the cord. (See also Chapter XL, Section 2 (1).) The stimulation of a mixed
nerve (the posterior roots being divided) or an anterior root thus evokes impulses which
travel up and break npon the shore of the nerve corpuscles, and either remain on

448 MESSRS. F. GOTCMl AND V. IIUIISLKY

their iieriplieral side, or issue in such broken and disorderly array that their
cliaracter is completely lost, and all evidence of their presence disappears.

To what extent is this true of the aflerent ro(jt iibres ? The answer to this question
is one which can only be appreciated when the results of the experiments of both this
chapter and the succeeding one, ii])on the reflex functions of the cord., have been
analysed ; but it may be well here to state at once that there is no evidence of such
pronounced block to impulses which may be caused to descend the indirect paths by
which the posterior root fibres are brought into connection with the cord.

The difficulties of interpreting the experimental results are increased by the gap
which exists in our anatomical knowledge as to the connection between these indirect
afferent fibres and the cord. Tliis connection is one upon which the degeneration
method throws but little light, although it would ajjpear probable from the trophic
changes observed, and from morphological considerations, that there are cells in the
postei-ior horn, and possibly in Clarke's column, which are connected with fibres
in the posterior roots (Mott). The recent reseai'ches of Kolliker seem to indicate
that the termination of some of" these indirect fibres in the cord is largely that of
a fine plexus with free ends.

Although the central connections of the fibres are as yet to a great extent
unknown, yet the description indirect as distinguished from direct is warranted by
its wide-spread use. It is advisable, however, to emphasise what the term indirect
as used by us in the present case is understood to connote.

It is descriptive of all afferent fibres in the cord, which mny be supposed to be
connected with those of the posterior root, but which do not show any of the
degenerative and developmental differences which stamp some fibres (posterior
median) as being in direct continuity with the root. There is, however, nothing in
itself to show that this term indirect is a strictly logical one ; since the same line
of argument might be applied to those fibres which pass through the ganglion on the
posterior root, with regard to which the degeneration method furnishes no evidence
of du-ect continuity (Joseph).

Impulses are conducted through the ganglion in either direction (du Bois-
Beymond), and apparently without any modification in their time relations (Exner)
if we may suppose that this is due to the fact that, apart from the few fibres which
appear not to come into relation with cells at all, the relationship of the majority of
fibres to the ganglion cells is of the T-piece kind, as shown by Eanvier, in which
case the cell does not interrupt the continuity of the connected fibre. There is no
reason why some such sort of bypath (GoLGl) may not be the basis of the con-
nection of the fibres with the corpuscular elements in the spinal cord, as it is in
certain parts of the cerebrum (Flechsig).

There are, therefore, no anatomical data which are necessarily implied in the term
indirect as applied to these fibres, beyond the fact that such fibres come into

ox THK ^[AJBIALIAN XERVOUS SYSTEM. 449

connection in some way with some element in the cord, wliicli serves ns a common
centre both for nutrition and for growth.

The researches on tlie conduction of afferent impulses from the nerves into the cord
detailed in the preceding chapter, as well as our other experimental investigations,
show that such paths undoubtedly exist. The present experiments will prove that
these, although indirect, are capable of conducting impulses from the centre towards
the periphery ; the only distinction between the indirect and the direct path, in this
respect, being the greater intensity of the stimulus necessary to produce the evidence
of such effects in the case of the former, and the comparatively small amount of the
nerve energy which can l)e thus evoked, through the indirect path, in the afferent
nerves. In other words, the direct path is that of least resistance to impulses when
these pass backwards from the cord into the afferent nerves.

We now pass to the consideration of the detailed plan of the experiments as a
necessary prelude to the nnalvsis of the results.

Section 2. — Plan of Exteriments and Prelimtnary Observattons.

Our first experiments upon the nerve effects following cord excitation were crude in
design. They were made by first dividing the coi-d in the dorsal region, then
preparing one sciatic nerve in the back of the thigh, ligaturing it, dividing it, and
connecting its central end and the adjoining longitudinal surface by means of cable
electrodes with the galvanometer, and finally stimulating the cord by means of
needles wliicii pierced it and acf-ed as electrodes to the secondary coil of the
inductorium. We satisfied ourselves in this way that the excitation of the cord was
followed by excitatory electrical changes in the nerve, and then proceeded to more
methodical experiments. In tliese the cord was carefully exposed by dissection and
a piece removed, so that the cross section of its surface could be easily seen. The
various cut ends of the columns were then excited as desired bv a series of interrupted
induction currents (Hehnholtz side-wire) for a period which was controlled by the
revolving mercurial key, and was carefully kept of the same duration during any set
of observations. A pair of well-insulated fine platinum-pointed electrodes were used
for the stimulating current, these being applied to the particular region of the cross
section it was desired to excite, in the manner and under all the precautions already
described in Chapters III. and VIII. The sciatic nerve when raised in air and con-
nected with the galvanometer displayed the usiial resting electromotive difference
between its two points of contact. This difference has been referred to at length in
Chapter IV. and was compensated in all cases. On exciting the cord for five
seconds an electiical effect was produced in the nerve which was always opposed
in direction to the resting current, and which passed away on the cessation of the
stimulus. The amount of the deflecticm was in most cases considerably less than
was obtained bv npi)lying the same stimulus (unaltered in intensity and duration)

Mt)CCCXCI. — B. 3 M

t50

MKSSKS. F. CiOTtUl AND V. IIUKSLKV

to the trunk of the nerve. It was notably affected by several conditions, of which
the most important are those affecti)i<^ the state of the animal antl tliose connected
with the intensity of the stimulus. The influence of these two conditions demands
closer examination.

The state of tiie animal has the greatest influence on the amount of tlu^ change.
To take the most }H)werful factor first, the systemic death of the animal, (liis is at
once shown in the diminishing size of the effect, nntil, in alx)ut ten minutes, no
electrical chaiiire can be evoked in the nerve when the cord is excited.

Tiie following experiment may be quoted in illustration of tliis point ; it is one In
which the posterior root instead of the nerve was observed.

Tiie cord of a Cat was divided at the 10th dorsal and the cauda equina exposed. The
7tli luiul)ar ])ostcrior root on the left side was then ligatured and divided near the
ganglion, raised in air ami its central end and surfiice connected with the galvano-
metric electrodes.

Excitation for five seconds of a given tract in the cord (the left posterior column)
evoked an electrical cliange indicated by a galvanometric effect of 25.3. The animal
suddenly died from collapse, the heart failing, and the experiment was then repeated
about four minutes after death, when the deflection was found to be less, viz., 1G5.

The time of this observation was 12.1. A series of su.cii observations were then
made as follows : —

Time.

Deflection.

12. 1

16.5

12. 2

142 1

12. 3

139

12. 5

75

12. 7

.50

12.10

8

12.12

nil

Excitation of the root itself still evoked consideral)le effects at 12.15. We have
often had the opportunity of noting that the nerve electrical effect wlien evoked by
excitation of the cord, disappears on death earlier than when evoked by excitation of
the nerve trunk. This we imagine to be due partly to the circumstance already
mentioned that the latter excitation produces normally a more marked effect than
the former, and partly to the changes in excitability which, in accordance with the
Ritter-Valli law, pi-oceed from the centre towards the peripheral attachment of a
nerve.

A more important influence than this of death, since it is one present throughout
all the experiments, is that dependent upon the varying degree of anresthesia. There
are many experiments which furnish illustrations of this point. A .stinudus of the
same character, intensity, and duration, applied to the same region of the cord, evokes

UN THE MAMMALIAN NHllVOUS SrSTK:\r. 451

now a larger and now a smaller effect, dependent upon marked lessening and deepening
of the narcosis. The importance of bearing this in mind is sufficiently obvious ; it is,
however, capable of control, since the character of the movements awakened by the
stimulus furnishes a fair index of the condition of anaesthesia, and was in all cases
closely observed.

Finally, the influence of temperature is one to which attention has already been
directed in Chapter III., and although the necessity of keeping the exposed cord
covered with Avarm sponges, except during the actual stimulation, has been already
dwelt upon, it may be insisted on again at this juncture.

Alterations in the duration and intensity of tlie stimulus modify the ertect ; the
longer the duration and the greater the intensity, the moi-e pronounced is the nerve
change. The modifications are so marked, that in analysing and comparing the
quantitative value of the nerve effects obtained by stimulation of different regions of
the cord, it will be necessary to divide the indicative galvanometric deflections into
two classes, as evoked by " minimal " and maximal stimuli respectively. It will be
understood, however, that a rigid separation is impossible, there being every grade
between an imdoubtedly " minimal " effect with no accompanying reflex movements
and an undoubtedly maximal effect with vigorous movements.

Tlie experiments were made upon twenty-one Cats and six Monkeys, and may be
divided into four groups, the flrat three of which differ as regards the nerve fibres in
\vhich the electrical change was observed, the fourth differing as regards the
condition of the cord, through the stimulation of which the nerve effect is evoked.
We have, therefore, to consider in succession : —

The electrical effect in the mixed sciatic nerve ;

,, „ afferent nerve roots ;

,, „ efferent nerve roots ;

The modifications produced by intervening section of tracts in the cord.

Each group will form the subject of a succeeding section.

Section 3. — The Electrical Effects in the Sciatic Nkrve following
Excitation of the Different Columns of the Cord.

The experiments upon this subject will be best displayed by first selecting and
describing the results of a particular experiment in the Cat and Monkey respectively,
and then giving a table which will show the averages of ;ill the comparaljle
observations.

The spinal cord was exposed and divided in an anaesthetised Cat (331), at a level
between the 1 0th and 11th dorsal vertebrae; both the sciatic nerves were then care-
fully exposed, ligatured in the lower part of the thigh, and divided on the peripheral
side of the lio-ature. Thev were then freed l'n)m tlioir attachments and raised in the

3 .M 2

452

.MKS.SKS. F. (iOTCli AND V. IIOUSLEY

air by the ligature. A separate pair of iioii-[)olarisable electrodes was ])laceil in
connection with the cross section and the surface respectively of each nerve. A
])iece of cord 1 centim. long was now removed from the upper end of the lower
Iragment of the cord, thus exposing a tresh cross section of all the columns.

The left sciatic nerve showed a resting electrical diffei-ence between the surface and
cross section, which was balanced by a difference of Dl Daniell; tlie right nerve
showed a rather less marked difterence balanced by '008 Daniell.

Each pair of nerve contacts was alternately connected with the galvanometer, a
fold's reverser without cross wires being used to switch either set of contacts into
connection. The following detlectioiis, o})posed in direction to the resting ditierence,
were obtained from the nerve wlien the cut ends of the various columns designated
l)clow were excited for 5 seconds : —

Left Nerve connected.

Sti-ength of stimulus.

Region e.vcited.

Effect.

Secondary coil.
500

Left posterior coluimi .
Left lateral ,, ...
Right posterior ,, ...
Right lateral ,, ...
Anterior columns ....

140

Nil
15

Nil
Nil

KiGHT Nerve connected.

Strength of stimulus, i Region excited.

1

Effect.

Secondary coil.

600

Left posterior column . . .
Left lateral ,, ...
Right posterior „ ...
Right lateral „ ...
Anterior columns ....

5 1
Nil 1
(J.J
Nil
Nil

The intensity of the stiumlus was now doubled.

Left Nerve connected.

Strength of stimulus.

Region excited. Effect.

Secondarv coil.
1000

Left posterior column . . . 190
Left lateral ,, ... o
Right posterior ,. ... oO
Right lateral „ ... Nil.

ON THE MAMMALIAN NERVOUS SYSTEM.

453

lliGHT Nerve connected.

Strength of stimulus, j Region e.scited.

Effect.

Secondary coil.

lOUO Left posterior column . . .
Loft ateral ., ...
Right posterioi- ,, ...
Right lateral ,, ...
Auteriox' cohmius

5
Nil
132

Nil
Nil

luteusity of stimulus ugaiu doubled.

Left Nerve connected.

Strength of stimulus.

Region excited.

Effect.

Secondary coil.
2000

Left posterior coluniu . . .
Left lateral „ ...
Right po.sterior ,, ...

Right lateral

Anterior columns

240
16

40

Nil
Nil

liiGHT Nerve connected.

Strength of stimulus.

Region excited.

Effect.

Secondary coil.
2000

Left posterior column .
Left lateral ,, ...
Right posterior ,, ...
Right ateral ,. ...
Anterior columns

15

Nil

172

5

Nil

Tlie above experiment shows pl&inly that, with a weak stimulus, an eifect is only
produced in the nerve by stimulating the posterior columns, and that this effect is
very marked when the directly continuous fibres in the jjosterior column of the same
side as the observed nerve are excited, and is small when the fibies of the posterior
column of the opposite side are excited. A very small effect is produced by stimu-
lating the lateral columns until the stimulus is strong, and even then an appreciable
effect is obtained only from the excitation of the lateral on the same side as the nerve.

A similar experiment may now be quoted in detail upon the Monkey (333)
(Macacus sinicus). The cord was divided in the aniesthetised animal at the 10th
dorsal vertebra, so that the cut ends of the various columns could be excited as in

454

.MI';SSi{S. F. (iOTCll AM) \'. llOUSLKY

tlif •;icvjoiis (_'.\|H'riniciit. IJotli nerves were then prepared and connected alternately
with the galvanometer in ti:c manner already descril)ed. The excitatory electrical
ertects in each nerve following;- the stimulation of the dillei-ent cohnnns of the cord
were ohserved with dilFeivnt intensities of stimulus, and tlie le.sults, which are
analoo-ous to those ohtained in the Cat, are .shown in tlir tnliowin-;- Table. The left
nerve was found to he less excitaljle than the right, but the resting diiference oi' Ijdth
was -005 I).

Lki-'t Nerve connected.

Intensity of
stimulus.

Region of cord excited at
cross section.

Deflection of
galvanometer.

•MOO

Left posterior column . . .

■ Left lateral ,, ...

Right posterior ,. ...

Right lateral ,, ...

14

Nil

3

Nil

Right Nerve connected.

Intensity of
stimulus.

Region of cord excited at
cross section.

Deflection of
galvanometer.

•2000

Left posterior column . . .
Left lateral „ ...
Right posterior ,, ...
Right lateral ., ...

12

Nil
68
10

Intensity of stimulus doubled.

Left Nerve connected.

Intensity of
stimulus.

Region of cord excited at
cross section.

Deflection of
galvanometer.

4000

Lett posterior column .
Left lateral ,, . .
Right posterior „
Right lateral „ ...
Anterior cohimns

106
25
39

Nil

Nil

ox TIIK MAMMALIAN NERVOUS SYSTEM.

455

lliuHT Nerve connected.

Intensity of Hcgiou of cord excited at
stimulus. 1 cross section.

Deflection of
galvanometer.

4000 1 Left postoi'ior column

Left atci'al ,, ...
Riglit posterior „ ...
Right latei'al ,, ...
Anterior columns

o2

Trace

112.

6

Nil

The effect in tlie mixed nerve of the Monkey is thus shown to be most strongly
evoked when the excitation is that of the posterior column on the same side as the
observed nerve. This preponderance over those effects which are evoked in a minor
degree from stimulation of the lateral of the same side and the posterior of the
opposite side is very marked when the stimulus is minimal.

The two experiments thus quoted form two in a series similarly made upon six Cats
and five Monkeys, and the average results of all galvanometric deflections obtained in
these form the most reliable data from which to draw any conclusions as to the relations
between the amount of electrical change in the mixed nerve and the special regions of
the cord which were stimulated. We must, however, again discriminate between
effects evoked by stimuli of only just sufficient intensity to ensure their presence,
which we term minimal {i.e., 500 ti) 1000), and those which more strongly arouse
the region excited, maximal {I.e., 2000, occasionally 4000) ; hence, for every column
excited, we have the result of a minimal and a maximal stimulus.

Moreover, it is obviously useless to designate which nerve is connected with the
galvanometer, if we place togetlier —

A. All effects evoked by excitation of the posterior column which is on the same
side as tlie nerve.

B. All effects due to excitation of the posterior column on the opposite side to the
nerve.

C. Those due to stimulation of the lateral column on the same side.

D. Those due to stimulation of the lateral column on the opposite side.

Any alterations in the stimulation as to intensity and duration necessary in
different animals do not affect the value of any comparison between tlie quantitative
results of the members of eacli different group, since the stimulus was always the
same during any one set of observations, and each set included stimulation in regular
succession of all the different columns indicated. The total result of su(;h observations
in the Cat is as follows : —

450

>[RSSnS. V. COTCIl AND V

i()i;si,i';v

Cat.

Anini:il.

Cord flividcd nt

I'llTcrt ill sciatic

iiorvo !'(

ll<i\ving

A. Excitation

B. Kxeitation

C. Excitation

D. Excitivtion

of postei'ior

of posterior

of lateral

of lateral

colunui of snmc

poluiiin of

cohunii of sanir

column of

side.

ojiposito side.

side.

opi)osito aide.

Mill. Max.

Mill.

Max.

Min.

Max.

Min. Max.

Cat (254) . .

Sth dorsal .

H5
80

20
8

0
0

82

18

12

0

Cat (265) . .

11 til dorsal .

36

86

0

0

32

0
0

7

0 0

0 1

Cat (311) . .

8tli dorsiil . .

49

130
152

108

2U

■2r>
36

1

0

Cat (320) . .

9tli doi-sal . .

50

85

5

25

0

10

0 0

28

89

5

22

9

Cat (363) . .

Otli dorsal . .

42

122

8

10

0

15

0

65

145

5

28

0

10

0

138

28

Cat (331) . .

10th dorsal . .

65 140

15

0

0

190

30

5

0

132

5

172

1 15

5

0

24n

40

16

0

Aggregate
Avei-age

sum

335

2178

28

378

0

89

0 ' 0

48

128

6

22

0

10

0 ; 0

From this It will be seen that with the minimal stimulus effects were obtained only
on exciting the posterior columns ; that the effect, doe to excitation of the posterior
column on the same side as the nerve, was eight times as large as that due to excita-
tion of the posterior column on the opposite side ; and that no effect was produced by
stimulation of either lateral column. With a stronger stimulus, it is seen that the
effect evoked from the posterior cohunn of the same side was five times as large as that
obtained from the posterior of the opposite side, and further an effect was obtained
from the excitation of the lateral column on the same side as the nerve.

Turning now to the Monkey, the results similarly grouped, ai'e as follows : —

ON THE MAMMALIAN NERVOUS SYSTEM.

457

Monkey.

Animal.

Cord divided at

Effect in sciatic

nerve following

Excitation of

Excitation of

Excitation of

lateral columu

of same side.

Excitation of

posterior
column of same

posterior
column of

lateral column
of opposite

side.

opposite side.

side.

Min.

Max.

Min.

Max.

Min.

Max.

Min.

Max.

Monkey (270) .

Sth dor.sa] . .

59

0

0

0

40

70

0

42

0

6

0

0

Monkey (280) .

7th dorsal . .

190
192
135

10

2

33

2
2

80

0

Monkey (221) .

10th dorsal . .

98
60
57

26
15
24.

45

6

40

8

12

2

102

10

20

3

Monkey (333. .

lOtb dorsal . .

58

68
106
112

18

12

37

52

0

10

25

6

0

0
0
0

Monkey (368) .

lOth dorsal . .

55
58

38
37
20

16
2

6
14

8

4
0

30

9

23

0
0

0
5
0

Aggregate
Average

sum

270

1275

36

287

4

304

0

30

54

91

7

21

—

22

0

2-5

It will be seen that with "minimal" excitation the average effect obtained in the
sciatic nerve of the Monkey by the excitation of the posterior column of the same
side was nearly eight times as great as that of the posterior column of the opposite
side. A small effect was obtained once after excitation of the lateral column of the
same side, otherwise these results are in complete harmony with those in the Cat.

With the stronger stimulus the average effects following excitation of the two
posterior columns have the relation of nearly five to one, the larger being due to
that of the posterior of the same side ; and, as in the Cat, a marked effect follows the
excitation of the lateral of the same side, this being equal to that evoked by stimula-
tion of the posterior column of the opjjosite side.

Finally, if all the results in both animals are blended and we compare both the
average effect and the highest deflection obtained in response to excitation of each of
the columns, the comparison shows the following results.

MDCCCXCI. — B.

3 N

458

MKSSRS. V. GOTCII AND V. HORSLET

Average Nerve Effect produced by Cord Excitation (Cat and Monkey).

Posterior column, same side

,, „ oj>]iosite side ....

Lateral cohiinn, same side

„ ,, opposite side ....

Total

Average.

Highest.

80

15

9

4

192
68
80
12

110

The sum of" the average effects produced by all the columns is 110; of this the
posterior column on the side of the nerve is capable of evoking about 73 per cent.,
the posterior column of the opposite side 15 per cent., the lateral column of the same
side 9 per cent., and the lateral column of the opposite side 3 per cent.

There is a great similarity, at least as regards the relation of the crossed to the
uncrossed side, between these different quantities and those which were referred to at
the end of the nerve-to-cord experiments, detailed in Chapter TX. An exact similarity
could not be expected, since we are dealing in the present case with mixed nerves,
hence efferent as well as afferent fibres are the subject of observation.

The presence of impulses which may be supposed to emerge from the cord by the
anterior roots might account for an increase in the effect evoked in the nerves, since
the excitation of such columns as the posterior may awaken reflex discharges from
the cord down the motor roots.

This explanation, however, cannot be a2523lied to the effects evoked by excitation of
the lateral column on the side of the nerve, since we see that the effect evoked by
this column Is smaller than that which the results of Chapter IX. would lead us to
expect as probable from the indirect connections of its fibres with those of the
posterior root only. There is no evidence of any accession of nerve impulses through
this excitation, but rather of a resistance to the passage of imj)ulses from the cord
through the lateral indirect path into the fibres of the mixed nerve.

It is, however, essential to ascertain the amount of the effects in tlie afferent and
efferent nerve tracts before proceeding to discuss in more detail what the above
experiments seem to indicate as to the relations of the cord to its nerves.

Section 4. — The Electrical Effects in Afferent Nerves following Excitation
OF the Spinal Cord above their Point of Entry.

(1.) Tlie Posterior Hoots.

The simplest mode of obtaining the effects in afferent nerves is that of exposing a
posterior root, dividing it near the ganglion and connecting its cut central end and

ON THE MAMMALIAN NERVOUS SYSTEM.

459

the surface immediately above with the galvanometer circuit in the manner described.
The objections to this experiment consist in the great shock to the animal which the
necessarily prolonged exposure of the roots necessitated by the procedure seems to
cause, and in the fiict that the posterior root does not hold out as long as the sciatic
nerve, its excitability being more readily influenced by falling temperature, drying, &c.

We have however succeeded in three different animals (Cat) in obtaining a series
of readings of the value of the root effect, the details of the experiments being as
follows : —

In all three animals, the left 7th lumbar posterior root was selected, this being

the largest of the posterior roots in the Cat which receives afferent fibres from the

sciatic nerve, as shown in the accompanying reproduction of a pliotograph of the

plexus. (See fig. 18.)

Pio-. 18.

The cord was first exposed in the lower dorsal region and prepared for excitation at
the level indicated in the table, and then the cauda equina, and thus the roots, laid
bare. The 7th lumbar root was then exposed from origin to ganglion, ligatured near
the ganglion and divided. It was raised in air by the ligature, and cables placed
round its ligatured cut end and its surface 1 centim. above. The roots, as stated in
Chapter IV., were remarkable in exhibiting a comparatively large resting electro-
motive difference between the surface and the cut section.

The different columns of the cord were then excited, first with "minimal" and then
with maximal stimulation ; it was found that the electrical effect in the root resembled

3 N 2

4 GO

MESSRS. F. GOTCH AND V. HORSLEY

that of the nerve in character, Imt owing to tlie difficnlty of keeping the root protected,
and at the same time insulated, the amount of the eflPect, and, presumably, the
excitability of the root, declined, especially in one Cat (341); the i-esults, therefore,
are all modified by the fact, that as the •' minimal" stimulus was applied first, the
deflection caused by the excitatory electrical root cliange is much larger in the case of
the " minimal " stimulus relatively to that of the " maximal " stimulus than would
otherwise be the case.

The following table gives the results of the galvanometric observations ; each
deflection being produced by electrical changes in the nerve root, corresponding to the
localised excitation, for five seconds, with the interrupted induction current of the
sectional area of a special tract of the cord as indicated below.

Effect in Left 7th Lumbar Posterior Root.

Excitation of

Excitation of

Excitation of

Excitation of

left posterior

right posterior

left lateral

right lateral

Auimal.

Cord excited
at level of

column.

column.

column.

column.

Min.

Max.

Min.

Max.

Min.

Max.

Min.

Max.

Cat (3-il) . . .

12tli dorsal .

190
135

110

108

90

90

9
6

32
19

50
45

0
0

0

0

0

10

0
0

0

0
0
0

Cat (348) . . .

11th. dorsal

192
170

200

2

2

70

0
0

0

0
0

0

Cat (362) . . .

10th dorsal .

50

240

2

45

0

0

0

0

Total sum . .
Average . .

253

15

0

0

735

1091

21

276

0

10

0

0

147

1.56

4

39

0

—

0

0

It is seen that with the minimal stimulus an effect is practically only evoked by
excitation of the posterior column of the same side as the root ; with a stimulus of
greater intensity an effect is also obtained with stimulation of the posterior column of
the opposite side, this being one-fourth the amount of that evoked by stimulation of
the column of the same side. As regards the lateral columns, only once was any effect
in the root obtained, and that was with stimulation of the column of the same side.

If it be remembered that the minimal effect is exaggerated in the case of the
posterior columns by the fact that their stimulation coincided with the fresh condition
of the root, it will be seen that as far as the relations between the effects due to

ON THE MAMMALIAN NERVOUS SYSTEM.

461

excitation of the two posterior columns are concerned, the above experiments give
results which are fundamentally the same as those obtained in the mixed nerve. On
the other hand, the lateral column effect is almost entirely wanting in these
experiments. The alterations in the excitability of the posterior root itself may be
to some extent resj^onsible for this, and in any case should, if possible, be excluded.
In order to carry out this exclusion, experiments were performed in which, whilst
investigating the effect in the mixed sciatic nerve, the (motor) efferent paths were
annulled by section of all the anterior roots of the lumbar plexus.

(2.) Effects in the Sciatic Nerve after Section of Anterior Roots.

This experiment we have performed on two animals (Cats), in each of which
the cord was divided and exposed for excitation at the level of the 11th dorsal
vertebra, and the left sciatic nerve prepared in the usual way for connection with the
electrodes. The lower lumbar cord and the Cauda equina were then exposed in each
animal by opening the canal for about 5 centims. The anterior roots of the left 5th,
6th, and 7th lumbar, and 1st and 2nd sacral nerves, were cut within the canal.
All other connections were then divided, so as to leave the left sciatic nerve in con-
nection with the cord by the posterior roots only.

The excitation of the cord proceeded in the usual manner, but a " maximal " intensity
of stimulus alone was employed, and the time of stimulation was 7 seconds.

Effect in Left Sciatic Nerve after Division of its Anterior Roots.

Animal.

Cord divided
at

Excitation

of left posterior

column.

Excitation

of left lateral

column.

Excitation of

right posterior

column.

Excitation

of right lateral

column.

Cat (209). .

11th Dorsal

125
130

85
92

not observed

50
not observed

9
35
15
not observed

0
0
not observed

Sum

Average

432

108

60
50

59
20

0
0

From these figures it will be seen that the effect evoked in the sciatic nerve by
excitation of the posterior column on the same side was more than twice as large as
that resulting from excitation of the posterior on the opposite side and more than
four times as lai'ge as that due to the lateral on the same side.

In another animal the same anterior roots wei-e divided, and in addition all the
posterior roots on the left side and all nervous connections except the 7th lumbar
posterior root, thus leaving the left sciatic nerve as a mere continuation of this

462 MESSRS. F. GOTCH AND V. HORSLEY

7tli lumbar posterior root. Effects of 35 and 20 were obtained from excitation of the
posterior column on the same side, and an effect of 10 witli excitation of the posterior
column of the opposite side ; no effect followed the excitation of the laterals. This
experiment is very difficult in execution, owing to the depressing effect which the
severe character of the operation produces in the cord, both directly and indirectly,
through shock to the animal. We have tried the experiment several times without
success. As far as they go, the results of the second experiment would seem to
indicate that the lateral column effect is not present when only one channel of
influence, that of the 7th lumbar posterior root, is present, and this result is in
accordance with the observations made upon the posterior root itself. It is not
impossible that the results of the first experiments with the sciatic nerve given in
Section 3, in which an effect was always obtained from the lateral column of the same
side, are due to the fact that the nerve investigated was left in connection with the
cord by fibres emerging from at least three posterior roots, beside the 7th lumbar,
but as we have not been able to carry the investigation of this point further at
present, we would merely emphasize the fact that the anatomical connections with
the cord were different in the two cases.

(3.) Summary of the Facts in (1) and (2).

If now we sum up the results of all these experiments when the nerve investigated
is connected with the cord by afferent tracts only, we find that the average effect
evoked by excitation of the posterior column of the same side

147 + 156 + 108 + 28 _ 439 _ ..^

the posterior column of the opposite side

4 + .39 + .50 + 10 103 „„

= i = x = 2^'

the lateral column of the same side

10 + 20 30

2 2

15,

the lateral column of the opposite side =i 0.

That is to say, as far as these readings go, of the sum of all the effects obtained,
the posterior column of the same side was concerned with 110/151 := 72 per cent, of
the effect in the nerve ;

The posterior column of the opposite side was concerned vdth 26/151 = 17 per
cent, of such effect ;

The lateral column of the same side with 15/151 = 10 per cent, of the effect ;

ON THE MAMMALIAN NERVOUS SYSTEM. 4G3

The lateral column of the opposite side with no part of the effect.

It is most remarkable how nearly these average figures resemble those obtained
by exactly similar average methods when the whole mixed nerve was the part
investigated.

This similarity suggested to us the possibility that the fibres in the anterior
roots are but slightly concerned in the production of the electi'ical effect in the mixed
nerve when the cord is excited by svich strengths of stimulus as have been used in the
foregoing experiments (500-2000, very rarely 4000).

It wdll be found on referring to Chapter VI., that whereas the electrical effect
observed in the cord to follow excitation of the cortex is very considerable, it is very
small in the sciatic nerve, and we there suggest that this difference involves a change
in the amount, intensity, or quality of the nerve impulses in their passage through the
unknown endings of the pyramidal tracts, and the known origins of the efferent
nerves, the anterior cornual corpuscles. (See fig. 22, p. 495, Chapter XI.)

To what extent, and under what circumstances, electrical effects can be detected in
the sciatic nerve, when all the posterior I'oots are divided and the nerve is connected
with the cord by efferent fibres only in the anterior roots, becomes therefore a most
interesting question.

To the consideration of experiments upon this point we will now turn.

Section 5. — The Electrical Effects in Efferent Nerves following
Excitation of the Spinal Cord.

The least complicated mode of experimentation for determining these effects would
be that of directly observing the changes in the central end of a divided anterior
root. This experiment, however, we have not yet successfully accomplished ; the
difficulties in the way of obtaining satisfactory connections with a divided anterior
root are augmented by its anatomical relations, and by the fact that, in order, as we
shall see, to obtain any changes in the root an intensity of stimulus has to be applied
to the cord which evokes general movements, thus dragging on the short root. The
dangers of inadequate isolation have been already dwelt upon in Chapter IV. Our
attempts in this direction w^ere so unsuccessful that we determined to employ the
more laborious method of division of all the posterior roots, and examination of the
electrical changes in the sciatic nerve.

The plan of experiment, therefoi-e, consisted in exposing the lumbar cord and
Cauda equina (see Plates 34 and 35, and fig. 18), and then dividing the posterior roots
of the 4th, 5th, 6th, 7th lumbar, 1st and 2nd sacral nerves, so as to leave the sciatic
connected with the cord by the efferent fibres only.

We made experiments upon seven animals (Cats) on these lines ; but in two of these
we unfortunately did not divide the 5th lumbar posterior root and the connection

464

MESSRS. F. GOTCH AND V. HORSLEY

with the 4th, which, as the figure of the plexvis shows, possibly furnish some,
thougli a very small proportion, of the efferent fibres of the sciatic nerve.

The results of these two experiments are, therefore, not completely to the point,
but they are interesting as showing the increase in the relative size of the effects
obtained with excitation of the posterior column on the opposite side of the nerve,
and of the lateral column on the same side, in comparison with those evoked from the
posterior column of the same side.

The section of these posterior roots has thus cut down the preponderating effect
obtained with excitation of the direct fibres in the posterior column.

Electrical Changes in the Sciatic Nerve after Section of the 6th and 7th Lumbar
and 1st Sacral Posterior Roots on the Left Side.

Cat (204) .
Cat (207) .

Intensity of
stimulation.

Effect in nerve following excitation of

Posterior

columns same

side.

Posterior

columns

opposite side.

Lateral

columns same

side.

Lateral

columns

opposite side.

1000

2000
4000

Average .

58
85
30
110
11

70

120

6

43

25

40

46

2

12

OOOO

294

263

99

59

53

25

The excitation of the anterior columns produced no effect.

The results are, however, probably mixed, as will be seen by reference to the
experiment on the remaining animals.

In these all connections of the sciatic nerve with the cord, except the anterior
roots, were divided ; the interesting fact then came to light that, even with a strength
of stimulus above that employed in the previous experiments, very small electrical
effects were evoked in the nerve by cord excitation, these being evoked by stimula-
tion of the posterior column of the same side and the lateral of the same side. When,
however, the strength of stimulus was increased very markedly, effects were produced
which were more marked in the case of the lateral. It is not, therefore, until the
intensity of the stimulus is far beyond the limits hitherto used that any marked nerve
effect is produced by the passage of impulses from the cord down efferent nerve
fibres.

This is illustrated by the following experiment upon a Cat (209), the cord being
cut at the 11th dorsal, and all the posterior roots on the right side divided as

ox THE MAMMALIAN NERVOUS SY.STEM.

465

described. With a stimulus of 2000, effects in the riglit sciatic nerve were observed
only from excitation of the posterior and lateral columns of the same side, the deflec-
tions amounting to an average of 10 only. On increasing the stimulus to a considerable
strength (4000) a deflection of 70 was produced by the excitation of the lateral column
of the same side.

Still more striking experiments are the following made upon two animals (Cats),
which may be set out in detail, the necessary strength of the stimulus employed being
noteworthy.

Effect in Efferent Nerves following Excitation of Spinal Cord.

Exoitati.- n

Excitation

Excitation

1
Excitation

Animal.

Cord cut at

Strpiigth of
slimulns.

of pos(ei-inr
column of

of posterior
column of

of lateral
column of

of latei'al
Column of

same side.

opposite side.

same side.

opposite side.

Cat (195)

13th dorsal vertebra

4000

0

0

6

0

8000

10

2

17

50
32

2

Cat (191)

lOtli dorsal vertebra

4000

12

6

0
0

18
85

5000

14

0

40

COOO

12

HO

60
40

Fresh section

5000

15

0

36

11th dorsal

(30C0

28
11

35

45

9

6

56

0

In the preceding experiments with afferent fibres, the maximal intensity of the
stimulus was that represented liy 2000 ; the great increase in the intensity of the
cord stimulus necessary to evoke effects in the nerve thi-ough efferent fibres is, there-
fore, very striking.

Another animal (Cat) which was experimented upon in the same way, must be con-
sidered as furnishing rather doubtful evidence, since we had previously not only exposed
and excited the cut ends of the various divided posterior roots on the side of the nerve
In order to obtain reflex effects, but had divided all the anterior roots on tlie opposite
side of the cord. These operations must, from their severity and extent, have altered
the excitability of the cotd itself As far as they go, however, the results are in
accordance with those jii.st set forth, since a stimulus of considerable intensity (3000)
was necessary to evoke any effect in the nerve which, in the case of excitation of both
posterior columns amounted to 30, and in that of the lateral column of the same side
to 36.

Finally, in order to avoid the changes in excitability following immediate section of
the posterior roots, those belonging to the 5th, 6th, 7th lumbar, and 1st sacral nerves,
were divided on the left s'de in a Cat (227) 26 days before the experiment. When

MDCCCXCT. — B. 3 O

466 MESSRS; F. GOTCII AN11 V. IIORSLEY

examined at this later date, the animal was found to uionc in an ataxic manner, but
was not paretic ; the sensibility of the left hind limb was very much diminished, and
the left knee jerk was absent.

The histological examination of the cord showed at the lesion on the left side
defeneration of {a) fibres entering the posterior cornu, {b) fibres in the posterior cornu,
(c) fibres in tlie posterior root zone, (d) fibres in the posterior external column.
Iligiier u]), at the llth dorsal vertebra, there was no degeneration in the entering
fibres, but a large cornu-shaped patch of degeneration in the left posterior external
coluani. The left posterior median column showed no deliiiite degeneration until the
lower cervical region was reached.

The experimental results were obtained by dividing the cord at the level of the lOth
dorsal vertebra, and preparing both sciatic nerves for connection with the galvano-
meter. Each nerve when excited evoked electrical changes in the observed region,
the left one (on side of lesion) more than the right. On stimulating the columns of
the cord, and observing the effect on the left nerve, it was found that, with the
ordinary strength of stimulus, very slight effects of 4 and 8 followed excitation of the
posterior column of the same side, and no effects were obtained with this strength of
stimulus from either the lateral of the same side or the columns of the opposite side.

On the other hand, in the right nerve, effects of 35 and 56 were evoked by
stimulation of the right posterior, and varied effects from 'J, to 65 on stimulation of
the right lateral.

The observation could not be repeated with stronger stimulus owing to the failure
of the animal.

An examination of all these results and comparison with those of Sections 3 and 4,
will show that the effects aroused in the mixed nerve by stimulation of the spinal
cord must be mainly due to nerve impulses travelling from the cord down the afferent
(sensory) fibres ; since the effect due to impulses travelling down exclusively efferent
(motor) fibres, in the first place, is very small in amount and, in the second place, is only
produced by a strength of stimulus in excess of that employed in the experiments on
the mixed nerve, and finally, is then mainly evoked by stimulation of the lateral tract
of the same side as the nerve under observation.

Section 6. — The Influence upon the Electrical Effect in the Nerve of

Intervening Sections in the Cord.

The experimental results to be studied under this heading throw more light upon
the relations of the sciatic nerves to the spinal cord.

In the foregoing three groups of experiments, whilst excitatory electrical effects
were evoked in the afferent, efferent, and mixed nerves by stimulation of the
different columns in the cord, no direct evidence was afforded of the nature of the
path in the cord along which the impulses, starting from the excited cross section

ON THE MAMMALIAN NERVOUS SYSTEM. 467

of any one particular column, pass to reach the nerve roots. It cannot be assumed
that the result of the excitation of any particular column is to evoke nerve impul.ses
which, in their passage to the issuing nerve roots, are Imiited to fibres in this
particular column, or that any particular excitation, however strictly localised, may
not excite by commissural fibres a neighbouring column. Indeed, as far as the cord
itself is concerned, the evidence afforded by the experiments (Chapter VIII.) sho^\■s
there is no absolute limitation of descending impulses to one column.

It is, therefore, a matter of great importance in connection with the 23i'esent
investigation to ascertain to what extent in the foregoing experiments the path of
such impulses as issue by the nerve roots is limited in the cord to the particular
column stimulated. We endeavoured to obtain information as to this in the
following manner : we first made experiments with the mixed nerve precisely
similar to those already described (Section 3), that is, we connected the nerve
with the galvanometer circuit, and excited the cut ends of the various columns
of the cord, as displayed in a transverse section. We then made a section of one
column between the seat of excitation and the lumbar roots of the sciatic nerve, and
repeated the first experiment under these conditions. We were thus enabled to
ascertain how far the nerve effect due to the excitation of any particular column was
reduced when the fibres of that column were all cut through. Any remaining eflfect
might be due to —

(«.) The presence of fibres which passed from the stimulated into other columns;

(h.) The presence of fibres connected with the corpuscular portion of the cord, and
the awakening of the same by their means, the nervous impulses thus reflexly
discharged proceeding either along fibres in other columns or along fibres in the same
column below the interruption.

To what extent these various factors come into play will be made evident by a
careful analysis of all the results. These are best grouped in accordance with the
particular part of the cord which was the seat of the intervening section.

A. Influence on the Effect in the Nerve of Jlenusection of the Cord.

The first group of results show that an intervening hemisection of the cord
completely abolishes the nerve effect produced by excitation of the columns on the
side of section, as well as that evoked by excitation of the posterior column of the
opposite side.

Thus the spinal cord was exposed and divided in a Cat (329) at the level of the
9th dorsal vertebra ; both sciatic nerves were now prepared for galvanometric observa-
tion, and the electrical effects obtained by exciting the cut section of the various
columns at the 9th dorsal vertebra observed, tlie results being given as under.

3 o 2

468

MESSRS. F. GOTCH AND V. HOKSLEY

Nerve observed.

Left sciatic nerve observed,
diff. = -012 Daniell . .

Right sciatic nerve observed,
diff. = -Oil Dauiell . .

Excitation of
left posterior.

Deflection 85
22

Column of cord excited.

Excitation of
left lateral.

10
0

Excitation of
right posterior.

25
89

Hemisectioii of cord on left side at lovol of 1st lumbar vcrtul)!-;

Left nerve observed .
Right sciatic nerve observed

Deflection 0
5

0
0

0
130

Excitation of
rijiht lateral.

The above experiment thus shows that with such an intensity of stimulus as that
used (1000) the result of an intervening hemisection on the same side as the nerve
is to abolish the nerve efi'ect. If it is on the o2:)posite side of the nerve, then the only
effect interfered with is that evoked by the stimulation of the posterior column on the
side of the lesion. It would therefore appear that with this strength of stimulus the
nerve impulses, which subsequently cause the electrical effects, are localised to one
side of the cord in both the area of stimulation and the subsequent path through th§
cord from that area to the issuing nerves.

There is, however, one point to which it is desirable to draw attention before
proceeding to the next experiments. Tliis is the absence of any effect in the left
nerve when, with a hemisection limited to the left side, the right posterior column is
stimulated. We do not interpret this as implying that there is no crossed path frona
the opposite posterior column to the roots of the lumbar nerves, but tliat either such
crossing has to a great extent occurred at a higher level than the 1st lumbar vertebra,
the level of the section, or that the hemisection had either directly (by injury) or
indirectly lowered the excitability of the neighbouring posterior column, and that
thus the intensity of stimulus used was inadequate to evoke nerve impulses which
could pass down the affierent fibres of such an indirect path as connects this column
■with the nerves on the o])posite side. This latter supposition is rendered not improb-
able by an experiment made upon an animal (Cat, 283)* in which the hemisection liad
been performed four months before the experiment. The cord of this animal was
exposed and divided for excitation at the 10th dorsal (the hemisection had been made
on the left side at the 12th dorsal). The electrical effects produced in the two nerves
by excitation of the difterent columns (coil 2000) were as follows : —

For full description of this auimal daring life and after death see pp. 429—430.

ON THE MAMMALIAN NERVOUS SYSTEM.

4G9

Cat (283).

Left iici-ve

Right nerve ....

Excitation of
left posterior.

Left lateral.

Right posterior.

Right lateral.

0
0

0
0

X5

82

1

0
15

In tliis case the previous hemisection had rendered excitation of the cokunns on the
side of the lesion quite ineffectual, as far as the generation and propagation of nerve
impulses into either nerve were concerned. The excitation of the posterior cohunn on
the opposite side to the lesion, however, evoked effects in both nerves. It would,
therefore, appear that the crossing through grey matter from the one posterior column
to the other, and so to the nerves, was in this case below the level of the hemisection ,
viz., 12th dorsal vertebra. This experiment apparently is contradictory in its results,
as regards the ci'ossing from the o])posite posterior column, to that juct given ; but
putting aside the presence and absence of immediate shock in the two cases respec-
tively, it must be remembered that the lesion in this case was two vertebras higher
up. The localisation to the particular column excited of the generated impulses at
the area of stimulation is very clearly brought out by this experiment, as well as the
failure of production of any crossed effect from the sound (right) latenJ column into
the left nerve.

B. Infltience on the Nerve EJfeet of Section of the Posterior Column on the Same

Side as the Nerve Observed.

The influence of the section of one of the posterior columns will be here considered
before that due to section of both, for although its effects are more complicated, the
experimental procedure by which the changes were produced involved the section first
of one and then of the other column. It will be found that the result is extremely
definite as regards the particular column opei'ated upon, as the following experi-
ments upon two animals. Cat and Monkey, show. It will be noticed tb;it a con-
siderable intensity of stimulus was necessary to produce effects in the Monkey.

470

MESSRS. F. GOTCH AND V. HORSLEY

Section of one Posterior Column on the same side as tlie Nerve. Loft Sciatic Nerve

observed.

Column of cord excited.

Excitation of

E.xcitation of

Excitation of

Excitaiion of

Cat (311)

Cord divided at 9th dorsal

left posterior.

left lateral.

right posterior.

right lateral.

vertebra

Stimulus 500

130
132

0
0

20
26

0
0

Loft posterior column cut at

0

0

10

0

1st lumbar vertebra

Stimulus oOO

0

0

5

0

Monkey (333)

lOth dorsal

106

25

39

2

Left posterior column cut at

12tli doi-sal vertebra

Stimulus 4000 ....

3
6

18
20

12
10

2
0

In these two cases an intervening section of the posterior column on the side of the
nerve observed had the effect of entirely abolishing, or diminishing to a mere trace,
the large nerve electrical change which was formerly produced by the excitation of that
column It may therefore be inferred, that the fibres which connect the posterior
column at the level of the 10th dorsal vertebra with the posterior roots of the lumbar
nerves on the same side, run wholly in that column, and that the stimulation and the
path are strictly localised therein.

There is, however, a further result, that, namely, shown by a diminution in the
nerve change evoked by stimulation of other columns. This may be due either to
the cutting off of crossing fibres, or to a depressed condition in the excitability of
the remaining part of the cord.

It is improbable, however, that both the latter causes operated in these cases,
since when we group together in the next table tbe influence of the lesion upon the
electrical changes in the nerve of the opposite side, although we obtain in both
animals evidence of diminution, yet the condition of the Monkey after the section
was evidently one of more and not less excitability than it was before, since the
uninjured right posterior column evoked larger effects than it did in the normal state.
(See Hyperexcitability after Section, Chapter IX., Section 7, C.)

ON THE MAMMALIAN NERVOUS SYSTEM.

471

C. Injluence of Section of the Posterior Column on the oj)posite side to the

Nerve observed.

Right Nerve observed.

Column of cord excited.

Excitation of

Excitation of

Excitation of

Excitation of

Cat (311)

left posterior.

left lateral.

right posterior.

right lateral.

Cord divicUcl ;it 9tli dorsal

:?6

0

108

0

vertebiM

Left posterior co]unui cut at

4

0

48

0

1st luuiliar vertebra

2

37

Monkey (S::!;i)

Cord divided at 10th . . .

62

4

112

15

Left posterior column divided

30

0

150

15

at 12th dorsal vertebra

It Is seen that the effect evoked in the nerve by stimulation of the posterior cohunn
of the opposite side is diminished by section. The diminution cannot be attributed
to general lowering of excitability since it is present in both cases, and it will be seen
that in the Monkey the excitability of other columns had increased. It must be due
to the inten'uption by the lesion of fibres crossing through indirect channels from
that posterior column. As this interruption occurs in the posterior column, it would
imply that there are fibres which cross over from one posterior column to the other, which
have a wide distribution aloncr the cord. The diminution is due to the cuttinof oif of such
fibres as descend to cross below the level of the respective sections ; the efiect still
obtained after the section may be due to those fibres which, having crossed above the
level of the same, are not interrupted by the lesion. Reference to the spinal cord
experiments (Chapter VIII.) will show that according to our experiments the fibres
in the two posterior columns have very extensive indu'ect connections with each
othei".

D. Influence of Section of both Posterior Columns on the Electrical Effects evoked in

the Nerve.

When both posterior columns were divided, mere traces of nerve effect were
evoked by excitation of either column, even when the stimulus was sufiiciently
intense to produce marked effects from stimulation of the lateral column on the same
side as the nerve.

This is shown by the following experiments made upon the same two animals, the
nerve effect due to excitation of each different column being compared before and
after the intervening section of the jjosterlors.

472

MKSSRS. F. CiOTClT AND V. HORSLKY

Nerve observed.

Column of cord excited.

Excitation of

Excitation of

Kx citation of

Excitation of

left posterior.

left lateral.

right posterior.

right lateral.

Cat (311)

Sei-lion lit' '.Uli (lorsiil

Before

26

0

Left nerve (.stimulus 500) .

1.52

0

20

130

81

8

Risflit nerve

10

0

Section of both jxisteric rs at

1st lumbar

After

0

Left nerve (stimulus luOO)

0

30

13

5

Rin-bt nerve

0

0

0

0

Left nerve (stimulus 200U)

0

65

Monkey (333)

Seetion at lOtli dorsal

Bcfoi'e

112

6

Eight nerve (stinmlns 4000)

52

4

Section of Ixth ]iosterIors at

12th dorsal

After

G

21

Right nerve (stimulus -1000)

6

0

Cat (251)*

Sect'on, 1 month pi'evions, at

lOtii dorsal

Right nerve (stimulus 4000)

0

10

0

9,%

It is evident that the interruption, as far as the posterior column is concerned, is
now practically complete for both nerves, hence all the fibres by which these columns
are connected witli the lumbar nerve roots are now severed, and thus any crossing of
nerve impulses out of either posterior column into the lateral columns, or into any
structure except the other posterior at a level higher than that of the 12th dorsal, is
negatived.

It remains now to see the influence of section of the lateral column upon the
effect.

E. Influence of Section of the Lateral Columns on the Electrical Effects evoked

in the Nerve.

The limitation of the area of stimulation, and tlie localisation of the descending
nerve impulses to the fibres of the excited column, are apparently very complete in the
case of the lateral column, so faf as the records of mu.scular movements can show.
{Cf. ScHiFP, see p. 347.)

It remains to be seen to what extent this is true of the particular impulses now
studied which, as our previous remarks have shown, must be considered as conveyed
into the nei've by the posterior as well as the anterior root fibi'es.

For full description of th's animal, i.e., ajipearances during life and death, see <*. 439.

ON THK MAMMALIAN NERVOUS SYSTEM.

473

The experiments upon this point were made upon two Cats, in one of which (259)*
the lateral column had been divided 34 days previous to the experiment, whilst in
the other an intervening section was made of all structures except one lateral
column at the time of the experiment.

The results of the experiment in the first case are given in the adjoining table: —

Left Lateral Column divided 34 days before Experiment, at ] 0th Dorsal Vertebra.

Nerve observed.

Column of cord excited.

Excitation of

Excitation of

Excitation of

Excitation of

left posterior.

left lateral.

right posterior.

right lateral.

Cat (259).

Section at 4 th dorsal vertebra

Left sciatic observed —

Stimulus 1000 ....

35

0

0

0

48

0

12

0

2000 ....

22

0

52

0

Rigbt sciatic observed^

Stimulus 1000 ....

20

0

40

0

2000 ....

42

0

78

28

It will be noticed that no change at all in the left nerve followed excitation of the
left lateral column, althougli, in the right nerve, a change was evoked from the
right lateral column. Since, however, nerve etfects are not easily evoked from the
lateral, column by moderate stimuli, and when these are employed strong violent
reflex ettects are produced by stimulating other columns, it was desirable to adopt a
diflTerent procedure and make an intei'vening section of such character as would
divide all fibres except those in one lateral column, which could then be used as a
standard of comparison.

The results before and after such division are shown in the following table : —

* For full description of this animal as regards its appearances during life and after death, see p. 405.

MDCCCXCI. — B.

3 P

474

MESSRS. F. UOTCH AND V. llOilSLEy

Nerve obscrvL'tl.

Column of coi'd excited.

Excitation of

Excitation of

Excitation of

Excitatif)n of

left posterior.

left lateral.

right posterior.

right lateral.

Cat (803)

Section at tlio 9th dorsal

vertebra.

Left nerve.

Stimulus 500 ....

42

0

H

0

1000 ....

122

0

10

0

2000 ....

188

0

28

0

Section of all columns e.xcept

right lateral at 12tli dorsal.

Stimulus 2U0() . . . .

0

0

0

0

4000 ....

0

0

0

14

Right nerve. Before section.

Stimulus 500 ....

5

0

95

0

2000 ....

28

0

145

0

Section of all columns except

right lateral at 12th dorsal.

Stimulus 2000 ....

0

0

0

0

4000 ....

0

0

0

28

In these experiments, it is seen that the right lateral column was excluded from
the section, and that stimulation of no other column above the section, except the right
lateral, produced any nerve eti'ect.

It has, however, been shown in Chapter IX. that when impulses proceed up the
posterior roots into the cord they are conveyed to some extent (20 per cent.) by the
lateral columns of the same side. A connection between the posterior root fibres and
the lateral column, presumably by means of fibres in the posterior column or cornu,
must therefore exist. This connection is, however, of such a special kind that it
offers an evident resistance to the passage of impulses backwards from the cord into
the root. The fact that no effects in the above experiment could be obtained when,
with the lateral column as the sole bridge between the excited upper portion of
tlie cord and the nerves, the posterior columns were excited, shows that any spread
of path from the posterior to the lateral columns, if it exists, must either be below
the level of the section (12th dorsal) or must ofter a resistance to the passage of
impulses in the downward direction which it does not offer to their passage in the
upward direction, that is, in their ascent towards higher centres.

The relations which the cord columns appear to have with one another, as given in
the summary of Chapter VIIL, are in this respect of great interest and importance.

Two results are clearly brought out by the consideration of the series of experiments
in the wliole of this section ; the first that the analogous character of the general
electrical results in cord and nerve, without being pushed too far, may be extended to
this point, that the stimulation of any one column in the cord is a localised stimulation
of the nerve fibres in that column only, and the secon^i, that the resulting nerve

ON THE MAMMALIAN NERVOUS SYSTEM.

475

impulses generated at the seat of stimulus are propagated to the root of the nerves
along fibres constituting direct and indirect tracts, which are for the most part con-
fined to the particular column which has been excited ; the sole exception being that
of the posterior column on the opposite side to that from which the nerve roots

Section 7. — Summary and Conclusions.

The experimental results detailed in this chapter form a sequel to those of the
preceding Chapter IX., as to the relations which exist between the lumbar nerves and
the spinal cord. Complete harmony between the results of the two series of experi-
ments is not to be exj^ected from the very different physiological and anatomical
conditions which they involve. It will, however, be seen, if we compare the present
results with those given in the summary of Chapter IX., that as regards the main
points the one set form a corollary to the other.

In order that the salient features of the two should be clearly expressed, the
following Table, in which all the results of the present chapter are massed under
different heads, is laid before the reader.

The Table shows the average percentage of the total amount of the effect which
has been observed in the sciatic nerve, or its roots, when each lateral and posterior
column of the cut cord is respectively stimulated.

The same average percentage is also shown, when the stimulus is minimal, in a
separate appended table.

I. — The Amount per cent, of the Total Nerve Effect evoked by adequate Cord
Stimulation due to Localised Excitation of Particular Columns.

Character of nerve con-
nected with galvanometer
circuit.

1. Excitation of

posterior column

of same .side.

2. Excitation of

lateral column

of same side.

3. Excitation of
posterior column
of opposite side.

4. Excitation of
lateral column
of opposite side.

Posterior root ....
Mixed nei've sciatic . .
Sciatic Nerve with an-
terior roots cut . . .

Per cent.

85
73

73

Per cent.
3
9

10

Per cent.
12

15
17

Per cent.
0
3

0

Average ....

77

7

15

1

II. — The same when the Cord Stimulation is Minimal.

Character of nerve con-
nected with galvanometer
cii-cnit.

1. Excitation of

posterior column

of same side.-^

2. Excitation of

lateral column

of same side.

3. Excitation of
posterior column
of opposite side.

4. Excitation of
lateral column
of opposite side.

Posterior root ....
Mixed nerve

Per cent.
97

84

Per cent.
0
0

Per cent.

3
IG

Per cent.
0
0

3 P 2

i7G MESSRS. F. nOTCir AND V. UORSLEY

If these results are to be relied upon as indicating not merely different amounts of
electrical change, but the presence of different amounts of nerve impulses propagated
from the stinndated area along the cord and into the observed nerve, then their miun
features indicate the following relations between the dorsal cord and the sciatic
nerve in the case of cord-to-nevve experiments: —

(I.) The path of least resistance, as shown by "minimal " stimulation, between the
cord and the afferent fibres of the sciatic nerve, is that of the posterior column of
the same side ; this path being that of the directly continuous afferent fibres.

(2.) Of the indirect nerve paths between the cord and the afferent fibres, the path
of least resistance is that of the posterior column of the opposite side, as is shown
also by Table II.

(3.) The afterent paths, direct and indirect, along which nerve impulses can be made
to travel (backward) from the stimulated cord to the sciatic nerve, have the following
relations to the posterior roots of the sciatic nerve, as shown in Table I: — ■

(a.) The posterior coliniin of the same side conveys 77 per cent, of the amount of
nerve impulses ;

(/>.) The lateral column of the same side conveys 7 per cent. ;

(c.) The posterior column of the opposite side conveys 15 per cent. ;

((/.) The lateral column of the opposite side conveys a mere trace.

We will now compare these deductions with those of the pi'eceding chapter from
iien\'-to-oord experiments. These latter were briefly as follows : —

(1.) The spinal path by which afierent impulses proceeding up the nerve travel in the
cord is, when these impulses are generated by " minimal " excitation, almost entirely
that of the posterior column of the same side, i.e., the directly continuous one.

(2.) Of the indirect spinal paths along which such ascending impulses travel, that in
the lateral column is the path of least resistance of the same side.

(3.) All the afferent spinal paths have the following relations to the posterior roots of
the sciatic nerves as regards the transmission of afferent imjjulses generated in the
fibres of the sciatic nerve: —

(a.) The posterior column of the same side conveys an average of 60 per cent, of
the total amount (intensity and quantity) of nerve impulses ;

(h.) The lateral column of the same side conveys 20 per cent. ;

(c.) The posterior column of the opposite side conveys 15 per cent. ;

{d.) The lateral column of the opposite side conveys 5 per cent.

The following features of these two sets of deductions are similar : —

i. That the channel of connection with the posterior roots is jjar excellence that o''
the fibres in the posterior column of the same side ;

ii. That what crossing dt)es exist is almost entirely due to indirect continuity with
the fibre? in the posterior column of the opposite side ;

iii. That the crossed path represents only about 15 per cent, of the available path ;

iv. That the fibres in the lateral cohunn have very much closer indirect relations

ox THE MAMMALIAN NERVOUS SYSTEM. 477

with the fibres of the posterior roots on the same side than they have with those on
the opposite side, these last being scarcely represented.

The most striking differences between all the connections of the posterior roots with
the cord, as dependent upon whether impulses are made to ascend or descend
them, are —

i. For ascending impulses the lateral column of the same side affords indirect
channels of greater efficienc}' than the crossed path in the opposite posterior column,
or, a fortiori, the opposite lateral column.

ii. For impulses descending along afterent paths the lateral column affords indirect
channels of less efficiency than the crossed path in the opposite posterior column.

iii. The ease with which the ascending impulses evoked by "mirmnoT' excitation
cross from the posterior root into the posterior column of the opposite side is con-
siderable as compared with that which characterises the passage backwards from a
posterior column into the opposite nerve roots when descending impulses are evoked
by such " minimal " excitation.

These differences tend to show that although tliere is no absolute block to the
backward passage of impulses descending afferent channels into afferent fibres of the
nerve, such as exists in the anterior roots to the passage upwards of ascending impulses
in efferent channels, yet there is a physiological difference between the facilities
which afferent indirect paths offer to the passage of impulses, this passage being
much easier when it occurs in the normal ascending direction.

We now consider ourselves warranted in concluding that, as far as the afferent
tracts are concerned, the deductions of Woroschiloff, Miescher, &c., referred to
on p. 420, are founded on experimental results which, from their nature, readily
admitted of misinterpretation, and that both physiological experiment and anatomical
investigation point with great distinctness to what is the true afferent path. Our
observations confirm the views as to the physiological properties of tlie posterior
columns set forth in 1847 by Longet and reasserted by Schiff, witli this modifica-
'tion, that although the posterior columns form the main path of connection between
higher portions of the cord and the posterior roots of the nerves, yet there is a path
in the lateral column which is strictly confined to the side of the entering roots.

It may be urged that there are pathological cases (Buown-Sequard) which
undoubtedly exist, of motor paralysis on one side and hemiansesthesia on the other,
following a local lesion of the cord. In reply, we can only state that we believe much
more definite evidence than that afforded by the existing clinical data must be forth-
coming in order to shake the solidity of the foundations for our propositions, a solidity
derived from tlie welding together of all the foregoing quantitative results of exact
experiments. We tlierefore are compelled to regard the above clinical experiences as
either in no way indicative of the normal relations of the cord to the nerves, or as
capable of explanation on the supposition that the lesion lias affected at the same time

478 MRRRRS. F. OOTCII ANM) V TTORSTyEY

the motor patli (latornl c-olumn) on one side iuid the inincipal afferent path (posterior
coUinm) of the opposite side.

Filially, the results of the foregoinp; sections have brought to liy-ht some interesting
details as regards the relation of the cord to the elferent nerves.

In the Table it will be noticed that the results, obtained with "minimal" stimuli,
as regards the sciatic nerve, are not affected by section of its anterior roots ; hence,
as far as the cortl stimulation in these experiments extended, no nerve impulses passed
down the motor roots capable of causing perceptible electrical changes in the nerve
except with strong stinnili.

As far, then, as the localisation of the path of eft'erent fibres in the cord is concerned,
the present investigation has at present gained no further information ; the lateral
column on the same side as the issuing nerve is the main path for outgoing nerve
impulses in the lumbar nerves. But the method has brought to light a remarkable
characteristic of this path, namely, that at one portion, that which connects in the
spinal nerve-centres the pyramidal nerve fibres with the origins of the anterior root-
fibres, it exercises such a modifying influence upon the ti-aversing impulses that these
issue so altered in number, intensity, or quality, as to cause but very slight electrical
effects in the nerves, and that it is not until a comparatively strong stimulus is used
that the issuing impulses are of such kind as to cause really appreciable nerve effects.

This brings us to the subject of the next chapter, which deals with the physiological
relations between the nerve-corpuscles and the nerve-fibres.

CHAPTER XL— ON THE FUNCTIONAL ACTIVITY OF NERVE CENTRES AS EVIDENCED

BY THE PRESENT METHOD.

Sectiou 1.^ — The present state of knowledge of the relations of the nerve centres in the cord.

(1). Anatomical. (2). Physiological.
Section 2. — Experiments by the galvanometric method directly bearing on the spinal nerve centres.

1. Resistance offered by the efferent side of the centre to the passage of impulses.

2. Resistance offered by the afferent side to the passage of impulses.
Section 3. — Character of the impulses dischai'ged by the spinal centres.

Section 4. — The influence exerted upon the electrical changes in a -directly-excited nerve by its attach-
ment to the spinal cord.
Section 5. — The spread of reflex discharges up and down the cord (iuternuncial fibres).
Section 6. — On the electrical changes in the cortical nerve centimes.
Section 7. — Summary.

Section 1. — The Present State of Knowledge as to the Relations of the

Nerve Centres in the Spinal Cord.

The employment of the galvanometric method to estimate quantitatively the

ON THE MAMMALIAN NERVOUS SYSTEM. 479

excitatory state of the nerve channels or fibres in the cord led us to attempt to
apply it as a gauge of the conditions prevailing in and between the spinal centres.

Even a careful review of the history of the subject fails to show absolutely more
than a few facts from which knowledge of the actually Intracentral processes may be
drawn. We will now present in a brief sketch what is understood from anatomical
and physiological evidence to fairly represent the structure and functional activity
of simple nerve centres, such as those in the spinal cord, in order that the bearing
of our own observations may be more comprehensible.

1. Anatomical Relations.

Tiie most recent anatomical investigations by Golgi, Held, Flechsig, and others,
chiefly by means of the methods of staining devised by Golgi, have established the
justification for regarding a nerve centre in the spinal cord as constructed as
follows : —

(o.) An afferent side ("sensory"), to which run afferent channels from the

posterior root.
[h.) Field of conjunction between the afferent and efferent sides,
(c.) An efferent side consisting of large nerve corpuscles, from which issue the

efferent channels.

(«.) Examined in further detail it appears that, of the afferent channels running into
the spinal cord from the posterior roots, some ascend directly in the posterior and
postero-median columns, others enter the posterior cornu of grey matter, and are lost
therein, others pass through the grey matter far into the anterior cornu of the same.

Further, the afferent side is composed of a ground basis, fibres, and corpuscles, the
branches of which cannot be traced soon after their subdivision.

(b.) Tlie field or area in which connection exists between the two sides of a nerve
centre is differently accounted for by various observers. All, however, are agreed
that the central branches of the large efferent so-called motor corpuscles repeatedly
subdivide towards the afferent side, and tend to form what was formerly described as
a network (Gerlach), but the anastomosis of the elements of which appears to be
more and more doubtful with further knowledge. (The actual mode of conjunction
between the afferent side and these branches of the efferent corpuscles is, therefore,
unknown ; but it is easily conceivable that as much as is known, i.e., subdivision, &c.,
may very properly be regarded as the seat of higher resistance, and we believe that
this field of conjunction is the region in which should be localised what we subse-
quently exprefss as the "block" to the passage of functional impulses.)

(c.) The elements on the efferent side have been long known as the " motor " part of
the nerve centres. The large corpuscles, recognised by many authors to be distinctly

•480 WKS.SRy. K. (;OT('U ANJ) V. llOliSLKY

fibrilluted in tlieir Intimate structure,* shew a jieculiar arrangement of one inedullated
outyoini; branch, and from tlie opposite side numerous suhdividing branches.

Spiital Ganglia. — A word nnist here be interpolated on the structure of tlie
ganglion of the posterior root in its character as a simple nerve centre.

All researcht on the anatomical structure of these ganglia goes to show that their
cor])uscles pass through a change in the course of tlu-ir development, being at first
bipolar and afterward unipolar, and that any axis cylinder in relation to any such
corpuscle does not directly enter this but gives off a branch at light angles to the
ganglion cell, thus forming what has been termed a T-shaped junction.

In addition, it has been shown | that there are fibres (as "determined" by the
degeneration method) which pass through the ganglion (spinal) into the posterior
root without coming into any relation with its cells. To sum up shortly, therefore,
it is evident that there are passing througli a spinal ganglion numerous direct proto-
plasmic nerve channels.

We have now to sketch further the afferent and efferent relations of the spinal
nerve centres with the encephalon, with the other centres in the spinal cord, and
with the periphery.

Afferent Clianneh of Communication. — The afferent paths may advantageously be
considered in reverse order to that just given.

The afferent fibres in the posterior root are now known to have the following
destinations as regards the spinal nerve centres : —

1. Some pass the centres and, without communicating with them, enter the postero-
external column and finally course up the postero-median column to the nucleus
gracilis.

2. Others entei- Clarke's column where that is present.

3. Others enter the posterior horn, and are lost in it.

4. Others enter the posterior horn, but pass through it to enter the anterior horn.

5. Some fibres bifurcate on their entry into the cord, one branch ascending and one
descending ; the branches end in a fine plexus in the grey matter (Kolliker).

The different fibres leaving the S2:)inal nerve centre to connect with its neighbour
have never received absolute anatomical demonstration, but from the results of
physiological investigation their existence must be postulated.

The afterent fibres connecting the spinal nerve centres with the encephalon have
been determined by the degeneration method to pass up the direct cerebellar and
antero-lateral tracts and have been traced as far as the cerebellum and pons, but no
further.

* See particularly Max Schui/ize, and H. Schultze, ' Arcliiv f. Auat. u. Physiol.' Fleuming.
Hexle's ' Festscbrift.'

t See particularly Eantier, ' Traite Teclinique d'Histologie.' His, ' Tageblatt der Naturforscher
Versaramlung za Berlin, 1«86.' Fritsch, ' Archiv f. Mikrosk. Anat.,' vol. 27, 1886.

X Max Joseph, 'Archiv f. Anat. u. Physiol.' (du Bois-Reymonu), 1887, p. 307.

ON THE MAMMALIAN NERVOUS SYSTEM. 481

Efferent Channels of Communication. — The efferent pnth fiom the eiicej)halon, r.c,
cei'ebrum, is well known as the pyramidal tract, and the fibres composing it nre
commonly supposed to end in the corpuscles of the anterior liorns. As this
connection is of the utmost importance, we ma}' be permitted to discuss the point at
a little fiu'ther length.

The evidence by the degeneration method shows that the principal mass of the
pyramidal tract fibres do not run towards the corpuscular elements in the anterior or
ventral horn, but towards the posterior or dorsal portion of the nerve centres, thus
suggesting that the fibres join rather the field of conjunction than the efferent side.
Again, the degeneration method shows clearly that no pyramidal fibres pass directly
out into the anterior root.

Anatomically therefore there is evidence of some structural change in the path
where the pyramidal fibre comes into relation with the spinal centres.

The distribution of the pyramidal fibres in the spinal cord is still so much a matter
stib judice, especially after recent work by the histological* methods, that the complete
destination of the fibres, when yet at the upper part of the cord, is unknown.

The existence of efferent fibres (internuncial) connecting each nerve centre with the
next and others below it have, like the afferent internuncial system, been surmised to
exist from physiological evidence, all anatomical facts being wanting.

We believe that our previous experiments (see bilaterality, &c.) go to show that
such efferent internuncial fibres must be few in number and but feebly differentiated.

The efferent channels from the spinal nerve centres are the fibres composing the
anterior roots.

Commissural Channels of Communication. — Of commissural connection between
spinal nerve centres of opposite sides of the cord no certain anatomical proof exists,
but the presence of such connection is surmised from the physiological evidence, and
is morphologically siiggested by many researches, especially those of Lockhart
Clarke and Golgi.

2. PJtysioIogi/.

While the spinal nerve centres have been investigated in numerous ways since
WuNDxt estimated the delay in the passage of an impulse through them, it is
remarkable how little is known with certainty of their functional activity.

By the galvanometric method we have been able to establish some fundamental
considerations on this matter, and, to indicate their bearing, we will first enumerate
categorically the facts that have been previously ascertained by various investigators.

The facts may be conveniently divided under the headings of —

* See particularly C. GoLGi, ' Anatoniisclier Anzeiger,' 1890, Nos. 13 and 14, and C. Sherringtom,
' Journal of Physiology,' &c.

t WuXDT, ' Untersnchungen zur Mechanik der Nerven und Nervencentren,' 187ri.
MDCCCXCI. — B. 3 Q

482 MESSRS. F. OOTCIT AN'I) V. IK^USI^KY

A. Spinal ganglion.

B. Jutrular fjanrrlion of vaj^ns.

C. Spinal C(inl nerve centres.

D. Ciiannc-ls of" eonirnunication.

A. Owing to the fact that the spinal ganglion has been made the object of some
important experimental observations we think it better to state these first, although
they are not, in our opinion, to be placed on the same ground as the plionomena
relating to the spinal nerve centres.

This latter position is justified on consideration of the anatomical structure of a
spinal ganglion jn-eviously given, whicli shows that it is traversed by " through"
protoplasmic channels.

(a.) Transmission of the Excitatory Electrical Change.

Du Bois-Reymond observed"' that when excitation was applied on the proximal side
of a spinal ganglion the " negative variation " was observable in tlie nerve trunk on
f,he distal side.

Uj.) Delay in the Transmission of a Nerve Impulse through a Spinal Ganglion.

ExNERt considered that his observations on a spinal ganglion in the Frog, warranted
the opinion that the delay in the transmission of the excitatory condition (negative
variation) through a ganglion was the same as would be pi-oduced by its passage along
a nerve fibre ; that, in short, there was no special delay. As these observations were
made with Bernstein's differential Rheotome, we agree with Gad that they may be'
possibly conditioned by summation,

(c.) Truj^hism.

As is well known since Waller's;}, classical experiments, the ganglion exercises a
trophic influence on the afi'erent nerve channels^ running in the posterior root.

B. Before passing to the observations on the spinal cord nerve centres, we must
specially allude to the valuable work by Gad and Joseph|| on the subject of delay in a
ganglion. The ganglion they chose was not a spinal ganglion, but its homologue, the
jugalar ganglion of the vagus in the Rabbit. By recording the instant of change in the
respiratory movements when the vagus was excited, respectively on the proximal and

* ' Uutcrsvicluingea iiber tluerlsclie Elektricitat,' 18-19, vol. 2, 5, 601.

t ' Arcliiv fill- Aiiatomie uud Physiologie.— Pliys. Abth.,' 1877, p. 567.

J ' Comptes Rendus de rAcademie des Sciences,' 1851, &c.

§ Excepting the fibres described by Max Joseph.

II ' Archiv fiir Anatomic u. Physiologie,' Do Bois-Retmond, 1888.

ON THE MAMMALIAN NERVOUS SYSTEM. 483

distal sides of the ganglion, they established reason for believing that tlie impulses
which affected respiration were delayed in transmission through the ganglion to the
extent of '036 second. As Gad and Joseph point out, these experiments are open to
the criticism that the summation of subminimal stinuili may be an important factor
in producing the result.

C. The conditions under which the functional activity of the nerve centres of the
spinal cord, as indicated by muscular contraction, is evoked necessarily form the
major part of the facts at our disposal.

These conditions may be arranged as follows : —

(1.) Adjuvants. — Warmth. Preliminary influence of cold. After effects of section
of spinal cord* a little distance above the nerve centre. Some drugs, e.g.,
strychnia, &c.

(2.) Depressants. — Prolonged fall of temperature. Shock after division of the spinal
cord (greater in proportion to proximity of section to nerve centre). t Antemia.
Narcotic drugs (anajsthetics, &c.).

(3.) Delaij. — A series of measurements have been taken of the time lost during the
passage of a nerve impulse from one posterior root through a spinal nerve centre to
the corresponding motor nerve of the same side. This delay, or time-loss, has been
estimated by the majority of observers^ to be about "01 in the Frog. (Exner, in
man, bulbar reflex '04 second). This refers only to the "direct reflex," i.e., of the
same side as the excitation.

The delay with the "crossed reflex" (see later paragraphs) is "004 second longer
than the direct.

(4.) Mode of Discharge. — A spinal cord nerve centre can be excited by a single
induction shock if it be applied to an aflerent channel. The eft'ect pi'oduced as
recorded by the contraction of a muscle is apparently that of a single twitch. If an
interrupted cui'rent be applied to the spinal cord or an aflerent channel, the efl:ect
similarly recorded is a continuous contraction.

An eft'ect in the muscles of an intermediate form has been frequently recorded, viz.,
a rhythm, the rate of which has been determined to be, on tlie average, eight to ten
per second.

Occasionally an after effect has been noted, i.e., a few muscular responses after the
spinal excitation has ceased, and in some cases these are continuous.

(5.) Trophism. — It is well known that lesions of the spinal nerve centres which
involve the ett'erent nerves are attended by wasting and other evidences of failure of
nutrition in the [)arts which are in relation with the centre in question.

* See FRAXrois Fhanck, he. cit. ; v. Bezold, loc. cit., &c., &c.

t De Boeck, ' Ai'chiv fiird. ges. Physiologie ' (confirmed also by our observations).
X Pavticalarly Helmholtz, Wundt, Feanck, Cyon, and others. Loc. cit.

3 Q 2

484 MESSRS. F. GOTCIl AND V. HOIISLKY'

(6.) Siiiiiuiiition of Stimuli — A most characteristic feature of the rniiclioii.il .-ictivity
of a nerve centre is its property of sum'nating (Stirling) sulnuiiiiinal stiinuli will}
the effect of |ini(luciiii;- an a]iparently complete discharjjje.

I). The cliunnel of cominunication between a spinal and other centres may be
briefly alluded to as follows : —

(c(.) Relation of one spinal nerve centre to another. — It is not possible on this point
to improve upoii the classical "Gesetze" of Pflugkr,'"' viz., that —

(1.) In a simple reflex the muscular response is always on the same side as the
excitation.

(2.) If a reflex is bilateral, the analogous nauscles of the opposite side are thrown
into action.

(3.) If a reflex is bilateral, the movements on the side opposite to that stimulated
are much weaker than those on the same side. {Cf our results on p. 494.)

(4.) If associated reflex centres are excited, the association is found to have an
ascending arrangement in the spinal cord, culminating in the medulla oblongata ;
Avhereas of the encephalic reflexes the association is of a descending chai'acter, and also
culminating in the bulb.

{h.) Conduction. — ScHiFFf showed, by section of the crossed pyramidal tract, that
the fibres running in it to the lowest spinal centres were undoubtedly the channels of
conduction from the encephalon.

ExNEll+ came to the conclusion, by time measurements, that there was a delay
experienced by the excitatory condition or impulse traversing the spinal centres,
when, the descending tracts in the coi'd having l^een excited, muscular contractions
followed ; but he admits that his results do not permit of a demonstration of the
amount of time lost.

Section 2. — Experiments by the Galvanometric Method directly bearing
UPON THK Structure of Nerve Centres.

(1.) Resistance offered hij the Effereiit Side of the Centre to the Passage of hnpidses.

The results wliicli have been obtained by our method are best arranged according to
the particular part of a nerve centre to which they appear to refer.

In considering the anatomical structure of a nerve centre as we have given it on
p. 479, it is evident that the region of greatest obscurity is that Avhich we have
termed the area or field of conjunction between the afferent and efferent sides, and

* ' Die sensorisclien Funcfcionen des Riickenmarks der Wii-bolthiere,' v. E. Pflugkr. Berlin. 18.53.
t ' Pfluuek, Archiv,' voL 30, 1883, p. 248.
X 'Pflugee, Archiv,' vol. &, 1874, p. 537.

ON THE MAMMALIAN NERVOUS SYSTEM. 485

in which there is from physiological evidence reason to believe considerable delay
occurs in the passage of the impulses through the centre.

Such histological evidence as has been referred to in Section 1 suggests that the
structure is of such niiture as not to facilitate the passage of impulses through this
region, but when we performed the following series of experiments we were surprised
to find the degree to which such hypothetical obstruction really prevailed. It
occurred to us that the obstruction might be more marked to excitatory impulses
passing in the reverse direction, i.e., from eflPerent to aflerent mechanisms. We were
not at all prepared, however, to discover, as we have done, that that obstruction was
actually an absolute block ; nevertheless, such appears to be the case.

To ascertain this we ai-ranged the experiment as follows : —

Having divided the cord at the level of the 10th dorsal vertebra (see tig. 19), we
raised its peripheral end for observation as described in previous pages and connected
it with the galvanometer by its cross section and surface. The cauda equina having
been exposed to an adequate degree, a pair of nerve roots was selected (usually the
7th lumbar), divided just above the intervertebral foramen and their central end
raised in the air by ligatures and separated. The exciting electrodes were then
applied respectively to the central ends of the posterior and anterior roots (see fig. ID,
Ex. p. and £x. a.).

The excitation in the first case invariably evoked a large deflection of the galvanometer,
but in the second case absolutely nothing even when the secondary coil was brought
to 12500 of the Kronecker scale, thus completely covering the primary, y.c., tiie zei-o
point of the centimeti'e scale. In other words, the excitatory condition, arriving
by the afferent channel or posterior root, not only passed up the few direct fibres of
the postero- external, and later of the posterior median column, but probably aroused
the afferent portion of the centre from which additional effects might ascend the cord
by the internuncial fibi-es, thus producing a very large deflection in the galvanometer.

On the other hand, the excitatory condition which arrives at the nerve centre by
passing up the anterior root or efferent channel is totally unable to reach the afferent
side of the centre, and so ascend the cord.

The importance of the negative result of this experiment caused us to vary it
as follows : —

In two Cats we divided the posterior roots of the 4th, 5th, 6tli, and 7th lumbar
nerves and the 1st sacral on the left side. We then divided the cord in the lower
dorsal region (10th dorsal), and connected the upper end of the lower fragment with
the galvanometer. On exciting the left (mixed) sciatic nerve, which was in connec-
tion with the cord by the anterior roots only, no electrical change could be evoked in
the observed region of cord however intense the stimulus used, whereas, excitation
of only one opposite posterior root gave deflections of 050", 320", and with a stimulus
of only 500, 130". This method avoided any chance escape of stinuilating current,

486

MESSRS. F. (U:)TCH AND V, IIOIISLEY

wliifli, wlioii tliis is very intense, miglii possil)ly occur in tlie case of the short
anterior root.

These experiments not only ilhistrate in a very strikinj.^ manner the extent of the
block to (ho pas.sage of impulses hnckwards throiif«h the eflierent side of a spinnl

Fiy. 19.

ifP

if

». ■

- , I.C

Cat.
Adiclt^

•

• m.D

' IX.D

— ...

xni.D

Ex.n

IV L

i S'

^ '

-'

1 il

VJI -^

IS /y,

^ilW

lip

n

Ex.P

nerve centre, they also afford iiii'oiiiiation upon the nature of the anatomical con-
nections of the anterior root fibres with the cord. The histological enumerations of
BiKGF, the developmental researches of His, and the extent of descending degenera-
tion in efferent tracts, have combined to indicate that all the fibres of the anterior
roots are in connection with nerve cells lying in the immediate neighbourhood of their

ON THE MAMMALIAN NERVOUS SYSTEM. 487

attachment to the cord, and that there are no tibres which pass directly from the
cohnnns of tlie cord (above) outwards into the anterior roots witliout interruption in a
nerve centre. If any uninterrupted fibres exist, then the excitation of the central
end of an anterior root ought to give an eft'ect in the galvanometer as arrano-ed in
this experiment, for it is well known that the excitatory state {i.e., negative variation)
travels indifferently in either direction along continuous tracts. The absence of such
effect is a very convincing proof that such through fibres do not exist. Moreover, an
additional point receives elucidation from this same experiment. We have already
drawn attention to the fact that the anatomical relations of the pyramidal tract
fibres are for the most part (and wholly as far as the lower limb is concerned) not
with the eflPerent side of the spinal centres but rather with either the field of con-
junction or the afferent aspect.

This anatomical juxtaposition of parts suggests to us that the pyramidal tract fibres
for the lower limb do not immediately end in the efferent corpuscles. Such a con-
clusion is in harmony with the experimental results now under discussion, for it is
unreasonable to suppose that the excitatory process would be so entirely blocked
when proceeding in the reverse direction along a path consisting merely of two fibres
(that of the anterior root and that of the pyramidal tract) joined by a large corpuscle
such as one of those in the ventral (anterior) horn.

But we must not anticipate the facts which bear upon the mode of termination
of the fibres of the pyramidal tract, and must return to the question of the block
offered by the construction of a centre to the passage of nerve impulses.

(2.) Resistance by the Afferent Side of a Nerve Centre to the Passage of Impulses.

After having discovered the fundamental position that it is impossible for an excita-
tory condition to pass " backwards " through a nerve centre from the efferent to the
afferent side (whence it could easily spread up the cord), it was clear to us that the
next point to be brought into relation with that just described would be the possibility
of a nerve centre discharging its energy backwards down an afferent path ; or in
other words, whether the excitatory condition, if started in the afferent side of a nerve
centre, could be transmitted backwards along the posterior roots. Upon this subject
we have accumulated a number of facts and experiments from several })oints of view,
and as these are in mutual agreement we can answer the question unhesitatingly in
the affirmative.

Many of the facts which establish the conclusion that a spinal nerve centre can and
does discharge energy down the posterior as well as the anterior roots when it is
stimulated into activity have more im]:)ortant bearings on other questions raised in the
present paper, and consequently have been referred to in Chapters VIIT. and X. We
will, therefore, simply enumerate the facts, and refer the reader to these chapters in
which tlioy are descril)ed, detailing only those whicli ha\e not been already treated.

488 MESSRS. l-\ GOTCH AND V. lloRSLKY

We reserve also to the end of lliis (.liiijjtci- (lie pertinent bearing of these observations
upon tlie important doctrine of kintesthesis (BasTIAN).

The evidence of the passage of nerve iin|)ulses, as indicated by electrical changes,
from the aroused centres in the C()rd down (lie afferent (sensorv) fibres of the posterior
roots may be grouped as follows : —

(1.) Electrical changes in the posterior roots and mixed nerve with aJI anterior roots
cat when the spinal centres discharge under the influence of strychnia.

(2.) Electrical changes in the above structures when the spinal centres are discharged
i-eflexly.

(3.) Electrical changes in the above structures when the spinal centres are aroused
by electrical excitation of the cohunns of the cord.

(1.) The Electrical Changes in the Posterior Roots when the Spinal Centres are Excited
h}/ Strychnia.

The most remarkal>Ie illustrations of the fact that impulses emerge from a dis-
charging spinal centre on its afferent as well as its efferent side, is that fiu-nished by
the following experiments.

In a Cat (211) the coi'd was divided as usual in the lower dorsal region (llth
dorsal) and both sciatic nerves exposed and divided. The central end of each nerve
was connected when desired with the galvanometer. The cauda equina was then
exposed, and the posterior roots of the 5th, 6th and 7th lumbar nerves, and those of
the 1st and 2nd sacral, were divided on the right side, whilst the anterior roots of
the same nerves were divided on the left side.

The right nerve was thus in connection with the cord by its anterior roots only,
the left by its posterior. Thirty minims of a 1 jaer cent, solution of acetate of
strychnia was now injected intraperitoneally. When the tetanic strychnia spasms
commenced, excitatory electrical effects were observed in both nerves.

The amount of the deflections varied with the intensity and direction of the spasms,
but were always larger in the right nerve (posterior roots cut) than in the left, averaging
180 in the right, and 30 in the left. The effect, though smaller in the posterior
root, was absolutely definite in character, each spasm being accompanied by an effect.

In another animal. Cat (341), we divided the cord at the 12tli dorsal vertebra and
exposed the cauda equina. We then selected the 7th left lumbar posterior root,
divided it near the ganglion and connected in accordance with our method the
central end with the galvanometer.

Thirty minims of a 1 per cent, solution of acetate of strychnia were then injected
into the peritoneal cavity. Each strychnia spasm caused excitatory electrical effects in
the posterior root, v/hich gradually increased in amount in proportion ns the discharges
became more violent and pi'olonged. The galvanometer deflections, which were very
definite in character and similar in direction to those obtained by the passage of nerve
impulses, amounted to 40 at first, then 70, and finally 120.

ON THE MAMMALIAN NERVOUS SYSTEM. 489

The employment of clamps, &c., obviated any errors due to the movement of the
animal ; the experiment mav, however, be objected to on the ground that the
strychnia discharges are abnormally intense, and must thus break through resistances
which ordinary central discharges would be unable to. That the centres can dis-
charge down posterior roots without being under the influence of strychnia is, how-
ever, shown by the frequent occurrence of similar though smaller galvanometric
eti:ects in the posterior root of an unstrychnised animal, when the cord is discharging
in consequence of previous electrical excitation and shallow narcosis.

(2.) The Electrical Changes in the Posterior Roots v'hen the Spinal Centres are
Discharged ReJIexJy.

As in the preceding experiments, so here, the evidence is that derived from the
electrical changes observed, (a) when the posterior root was directly connected with
the galvanometer, and (h) when the (mixed) sciatic nerve was connected (after
division of all anterior roots) with the galvanometer and the central end of one
postei*ior root exoited.

(a.) In this experiment the central end of one divided posterior root was connected
■with the galvanometer and the sciatic nerve of the same side stimulated, the result
being to produce in the galvanometer a deflection of 20 scale whenever the muscles
were thrown into a fair reflex spasm (coil 8000, 341). We shall see presently that this
is the amount frequently obtained from the mixed nerve as the result of a reflex dis-
charge. It is most interesting, therefore, to see that the centre reflexly discharges
backwards down the posterior root. The bearing of this on the discussion as to
which part of a centre is probably the source of the kinetic nerve will be seen
further on when this question is raised.

(h.) We varied the experiment in another animal. Cat (268), by dividing all the
anterior roots supplying the left sciatic nerve and connecting the cut central end of
this latter with the galvanometer. We then divided one posterior root, the 6th left
lumbar, and excited its central end.

By this arrangement we obtained excitation of the spinal centres in the lumbar
enlargement, and at the same time left only the posterior roots or afferent channels
for any downward discharge such as would evoke electrical changes in the sciatic
nerve.

One observation gave an effect of 10, another of 19 in the galvanometer, the reflex
being weak in the first and fairly strong in the second.

(3.) Existence of Electrical Changes in the Posterior Boots tvhen the Spinal Centres are
aroused hij Electrical Excitation of the Columns of the Cord.

The details of these experiments, in which a posterior root was connected directly with
the galvanometer and the cut surface of the spinal cord in the dorsal region excited,

MDCCCXOI. — B. 3 R

400 MESSRS. P. OOTCH AND V. HORSLEY

are given in Chapter X. Reference to this Chapter will show that the effects in the
posterior roots may be due to (a) conduction of nerve impulses from the excited area
down directl}' continuous fibres, (h) conduction of iipive impulses down fibres which
come into relation with cells, and only indirectly with the iibres in the roots. To
these two must now be added (c) the discharge of impulses from the centi-es in the
cord aroused by the excitation.

It is evident that the discrimination of these different factors can only be achieved
by careful comparative experiments made upon the cord, \\hen its direct connections
with the roots having been interrupted, the centres are placed under different
conditions as regards excitability. These experiments have not as yet been suffi-
ciently satisfactory to enable us to form any decisive opinion on this point.

There is, however, one very important fact already referred to at the close of the
preceding Chapter X., to which we must now draw attention.

It will be remembered that, on comparing the results of the experiments detailed
in Chapters IX. and X. respectively, the inference was suggested that the indirect
cliannels by which the cord communicates with the posterior roots, are of such a
character that they offer more resistance to the passages of impulses from the cord to
the nerve than to the passage of similarly evoked impulses from the nerve to the
cord. There is thus evidence that, as regards these indirect channels, the same
phenomenon of unequal conduction is present which, in its most marked fn-m,
exists in the case of the efferent region of the spinal centre.

It must, however, be borne in mind that these indirect channels do not necessarily
form the afferent side of the centres, but are only in relation with them, since it is
quite possible that in many instances these fibres pass through cells without
entering that field of conjunction which forms the terra incognita of the centre.
Evidence, therefore, of the nature of the discharge from this term incognita backwards
down the posterior roots, as well as forwards through the anterior cornual cells, could
only be obtained through electrical excitation of the cord, if such excitation was
limited to the pyramidal tract. The simplest way of performing this experiment is
that of exciting the commencement of this ti'act in the cortex cerebri, and of
observing the changes in the posterior roots. We have endeavoured in two animals
to perform this experiment, but without obtaining as yet any satisfactory evidence,
nor is this to be wondered at, when it is borne in mind how comparatively small the
effects a.re in the mixed nerve imder these conditions.

Sununing up the evidence obtained from all methods of experimentation, we are led
to conclude that when a spinal nerve centre is thrown into activity, a portion of its
energy flows as a discharge backwards down the posterior roots as well as forwards
down the efferent fibres of the anterior roots, and upwards and outwards along
internuncial fibres to the next centres.

With this somewhat enlarged view of what occurs when a spinal nerve centre

ON THE MAilMALIAN NERVOUS SYSTEM. 491

discharges, we will proceed to show what fresh light the galvanometric method is able
to throw on the character of the issuing impulses.

Section 3.— Character of Nerve Impulses Discharged by the

Spinal Centres.

It seemed to us that an estimate of the amount of the impulses which pass along the
nerves as the discharge of a reflex centre, might be obtained by comparing the amount
of the electrical change in a peripheral nerve when (a) the centre is reflexly aroused,
and (h) the nerve itself is subjected to a stimulus of similar duration and strength.
Such an estimate can only give approximate results, but the diiference revealed
in the two cases is so striking, and the desired comparison of so much importance
in attempting to analyse the components of a reflex, that we have devoted some
time to securing it. We have done this for both a "simple" reflex, i.e., that obtained
from the same side as that of stimulation, and we have also measured the effect when
it is the result of a " crossed " reflex.

( 1 . ) The Excitatory Electromotive Change Produced in a Mixed Nerve ivhen a

Spinal Centre is Excited Rejlexhj.

{a.) Simple Reflex. — To obtain the electrical effect produced by a simple reflex
discharge into the peripheral nerves, we divided the spinal cord in the dorsal region,
then prepared and divided the sciatic nerve and connected its central end with the
galvanometer. The cauda equina was exposed to a limited extent, and a posterior
root of the same side selected for excitation ; this was divided near the ganglion and
its central end stimulated. (See fig. 20.)

The excitation was performed for a known time, as in all our experiments, and the
galvanometric effect obtained in the nerve noted. We then applied the same, or a
weaker stimulus of the same duration to the observed nerve. The results obtained
form a remarkable contrast. The strength of the stimulus is so greatly altered
in its effect by the condition of the animal, rest, depth of narcosis, &c., that what
is minimal in one case, is quite adequate, or maximal in another.

The readings of the deflections indicating the reflex electrical effects in the nerve
were : —

Highest readings .... 72 (iU 55 (three highest taken).

Average 26

Lowest readings .... 7 3 nil (three lowest taken).

Out of the whole number of observations, 9 per cent, gave no result. Finally, the
stimulus employed on the average amounted to 2500 of the inductorium scale.

Taking the average result, namely, 26 of the galvanometer scale, it is at once

3 R 2

492

MESSRS. F. GOTCll AM) \. ILUHSLEY

obvious A\hat an extraordinarily small reading this is when compared to that which is
observed in a peripheral nerve, or even the j)t;ripheral portion of" a divided posterior
root, which, with a stimulus of one-fifth the intensity (500), and similar duration,
averages 227 scale, and once reached a maximum of 445 scale.

Fitr. -20.

This definite result is very significant, but before enlarging upon it we will describe
the values obtained for the " crossed " reflex.

(b.) Crossed Reflex. — To obtain the value of a crossed reflex discharge in the
peripheral nerves, we divided, as before, the cord in the dorsal region, and then
prepared one sciatic nerve for connection with the galvanometer, and next, either

ON THE MAMMALIAN NERVOUS SYSTEM.

493

the opposite sciatic nerve, or corresponding posterior root of the opposite side, for
excitation. We made this experiment in nine Cats, two Monkeys, and one Rabbit.

Fig-. 21.

The etJect in the observed sciatic nerve was, consequently, tliat due to the discharge
of the nerve centres on the side opposite to that of stimulation, or on both sides
of the cord.

Owing to the higlier excitability of the posterior roots, as compared with that of
the mixed sciatic nerve, it is not surprising that we found considerable difference
in the amount of effect obtained by crossed reflex action, according as it was

494 MESSRS. F. UOTCH AND V. HOKSLEY

elicited by the one or the other of these two ways. We, therefore, separate the effects
into two divisions indicating this difterence.

Posterior Root.— The general results in this series of experiments were as follows : —

Effects in nerve Ity excitation of opposite root.

Highest readings 45 42 42 (three highest' taken).

Average 25

Lowest readings 8 (5 nil (three lowest taken).

These figures are, on the whole, lower than those of the simple reflex. Other
considerations emphasise this contrast. These are (l) the proportionate number of
cases in which no eftect occurred, this being, with the simple reflex, 9 per cent,
of the total number of observations, but with the crossed reflex three times as much,
viz., 27 i)er cent. ; (2) the average intensity of the stimulus required to evoke the
effect, this being for a simple reflex, an intensity indicated by the secondary coil at
2500, for the crossed reflex the higher intensity indicated by the coil at 3500. Thus,
as compared with the simple reflex, the crossed reflex is feebler, less often obtained,
and when obtainable requiring a more pronounced stimulus applied to the afferent
flbres in a posterior root.

Sciatic Nerve. — In addition to the above we have observed crossed reflex effects
which are not comparable with the simple uncrossed effects, since they were evoked
by excitation, not of a root, but of the opposite sciatic nerve.

These effects are smaller and more difficult to elicit than those just indicated.

The eftects were evoked and observed in precisely the same manner as in the
previous cases, and the amounts of the deflections in the central end of the sciatic
nerve were as follows : —

Highest readings 20 15 14 (three highest taken).

Average 10

Lowest readings 6 5

The intensity of stimulus required to evoke these ett'ects was very much higher
than in either of the previous instances, the average being represented by a position
of the secondary coil of 8300, whilst it often happened that a position of 12,000 was
necessary. It may be remarked that the use of the Helmholtz side-wire in the induc-
torium minimised any error due to electrotonic escape, such as produces the well-known
phenomenon of the so-called " paradoxical contraction."

We are now in a position to consider the signi6cance of these results.

The remarkable difterence in quantity between the electrical eftect in the nerve
obtained by exciting the fibres themselves and that which is the result of a reflex
discharge from the cord implies that, however intense the excitation of the afferent

ON THE MAMMALIAN NERVOUS SYSTEM.

495

side of a centre, the amount of" the flow and the intensity of the processes which
emerge from the efferent side is small. A compai-atively intense excitation is necessary
for the etherised centre to discharge at all, and the emerging impulses are so altered
in their time relations and intensity, or limited to so few channels, that but little
evidence of their preseT\ce is indicated by the galvanometric method.

To facilitate further explanations we would call attention to the following diagram,
fig. 22, in which are represented the hypothetical elements of a nerve centre.

Fig. 22.

Scctio/ial Surface
of Cord ^

Efferrnt,
Side.

Di

Afferent
Side.

C Sp.g

Field

■ of ■
Conjunctzon ■

In fig. 22 are shown as simply as possible the three constituent parts of a simple
nerve centre, namely, the aft'erent side, field of conjunction, and efferent side. In
addition, the diagram represents the known part of the afferent channel which forms
the direct ascending fibres in the spinal cord (Di), the internuncial fibres (I), which
also ascend, but which are not yet absolutely known, and the fibres of the pyramidal
tract (P).

If a posterior root be excited, at the point marked Ex. in fig. 22, we can obtain a
record of the electrical change evoked by such excitation, in the following parts,
namely, the portion of the cord (A) above the excitation, in the outgoing channel (B)
(em.), and in the channel of excitation (C). On taking the average of all such records
we find that they are as follows : —

Average at A 198.

B 26.

C . . . . 250.

•iOG MESSRS. F. GOTCH AND V. HORSLEY

Altliongli the cross sectional area of the root is much smaller than that of the cord,
yet tlie effect in the tihres of the root (C) is greater thaii that in the fibres of the cord
(A), suggesting that there is less nerve energy after propagation thi-ough the afl'erent
side of the centre. It seemed at first sight not improbable that the effect at A pro-
duced by the passage of nerve impulses up the direct channels which join A with C,
plus additional nerve impulses evoked by a discharge of central activities, would be
very large. It would seem, however, if the electrical indications are to be relied upon,
and we see no reason to discredit them, that the impulses which enter the central
structures from Ex. suffer a diminution which more than counterbalances any such
addition. We are, however, well aware that the reduction may be due to a spresid
of nerve impulses in their passage through the afferent side of a centre ; and since a
similar reduction occurs both in the case of impulses descending the pyramidal tracts,
and emerging in the roots, as well as in those which as described are reflex in character,
we incline to the opinion that ascending nerve impulses do suffer a diminution in
intensity in their passage through the afferent side of a centre.

With reference to the impulses descending the pyraniidal tracts, whether originating
in the cortex cerebri or the fibres of the corona radiata, the facts given in Chapters V.
and VI. suggest that a very considerable reduction occurs before they emerge in the
anterior roots.

In this connection we would here draw attention to some experiments which we
have made to record the muscular contraction evoked by excitation of the spinal
centres, and particularly to mea.sure the time between excitation and the commence-
ment of the muscular contraction. The period of delay or latency thus estimated
was ascertained by the use of Tigerstedt's signal method, as described in Chapter III.,
Section 3, and in Chapter VIII., Section 1. In the latter place we have indicated
at length the method we employed for the measurement of the rate of transmission
of the excitatory effect in the cord fibres, which we found to be about 39"5 metres
per second. We have now to point out the reasons which led vis to measvu'e the
loss of time in a Mammalian spinal nerve centre. It appeared to us that this
loss of time which the excitatory condition suffers in going through a spinal nerve
centre would not be the same if the path of its entry into the centre is different.
There are two modes of entry with which we are famihar. These are (1) the
termination of the pyramidal fibres in the nerve centre, (2) the termination of the
afferent channels in the posterior root in the nerve centre.

Now, as stated on p. 48], the latter has frequently been estimated, but Exner has
attempted the measurement of the former, and not with a very definite result.
These periods of time have not to our knowledge been measured in the Cat, and, as
we felt the importance of forming some idea of their value, we performed a few
experiments as follows.

The problem in each case is shown in fig. 23. The nerve centre is at EA, E being
its efferent, A its afferent side, and / the field of conjunction.

ON Till-: J\lAM^b\LIAN NERVOUS SYSTEM.

497

JEae

The ordinary reflex is produced by the stimulating electrodes being applied to an
afferent nerve, the posterior root at B, and the effect recorded by contraction of the
muscle M. If the distances BA and EM are known, and the latency delay in the
muscle, then it follows that by subtracting these from the total time expended in the
execution of the phenomenon we obtain a residue which is the actual delay of trans-
mission of the excitatory effect within the centre EA itself.

This we have done as follows : —

Simple Reflex Delay, two Experiments (Cat).

Second.

1. Total expenditure of time between excitation at B and

contraction at M "022

2. Time expended in transmission from B to A and E to M at

38 metres per second '006

3. Time expended in latency of muscle at M "01

The sum of 2 and 3 gives all the time expended in extrinsic duties ; the subtraction,
therefore, of these from 1, gives the delay or loss of time in the centre itself; this
amounts to '006 (in one case '004).

We now turn to what is more interesting, viz., the loss of time when the impulse

MDCCCXCI. — B. 3 S

408 MESSRS. F. flOTCII AND V. ITORSr.KY

coming to the centre ajiproaclies it by way of the pyramidal tract wliicli runs in the
lateral coltinui, and is marked as a thick interrupted line in the diagram and l)y the
letters py.

To obtain the time loss we liave as before to take the total time expended from the
moment of excitation at L to that of the contraction of the muscle M, and then to
subtract from it the time occupied in simple transmission along the line L/M, and,
finally, to subtract the latency time of the muscle. These times are as follows : — •

Dklav in the Spinal Centre during Excitation of the Lateral Column.

Average of four

experiments (Cats).

Second.

1. Total expenditure of time from L to M '0176

2. Time expended in transmission from L to f, and

E to M -OOG

3. Time expended, in latency of muscle '01

On subtracting 'J + 3 from 1, we get as a final result '001 G. This resultant loss of
time is thus several thousandths of a second shorter than the delay taken by trans-
mission of the reflex effect. This is naturally to be expected as a consequence of the
lesser amount of centre to be traversed, but possibly the smaller result is also due
to the channel of the pyramidal path in the nerve centre being " polarised " for
descending and efferent impulses.

It appears, therefore, that when the impulses have to pass through the whole
central structure (as in reflex) they suffer both a delay in time and a very marked
reduction in quantity, and that when they pass through a portion which excludes the
afferent commencement of the centre (cortical discharges), although the reduction is
very considerable, the delay is much less.

There is one point to which it is necessary to draw attention before leaving this
part of the subject, namely, the well-known susceptibility of central mechanisms to
changes from fatigue, these altering the amount of nerve energy discharged. It will
be sufiicient here to indicate by two examples the influence which previous activity has
upon the amount of nerve energy reflexly discharged from a spinal centre, the latter
being indicated by the electrical effect in the issuing nerves.

ON THE MAMMALIAN NERVOUS SYSTEM.
Effect of an Interval of Eest on the Simple Reflex Discharge.

499

ExcitaHon of tlip central end of the left 7th lumbar posterior

Galvanometric deflections.

Before rest.

After rest of
1-5 iniuntes.

16

20

42
39

Excitation of 1st sacral as above

Effect of Want of Rest on the Simple Reflex Discharge.

f 18 Coil 2000
Cat (263) Excitations following at two minutes' interval of right | 15 „ 4000
7th lumbar posterior root gave in the right sciatic <( After 15 min. rest
nerve | 12 „ 2000

L 17 „ 4000

f 26 Coil 2000
Cat (378) Excitations following each other at 1 minute's interval < oi " goon

L 8 ", 8000

Character- of the Electromotive Change which is Produced by the Discharge of a
Nerve Centre as Contrasted with that Artificially Induced in a Nerve Channel, i.e.,
Fibres.

The galvanometric method appears to us to suggest differences between the passage
of the nerve excitatory condition (so called " nerve impulses ") along fibres according
to whether it is a centre which is the source, or an artificial stimulus in the actual
course of the fibres. Thus, if a peripheral nerve be excited directly by any form of
excitation, i.e., electrical or mechanical, the effect in the nerve {i.e., the excitatory
condition) shows itself in the galvanometer as a movement which begins sharply and
proceeds moderately rapidly, but steadily, so long as the excitation lasts, stopping
when the stimulus is left off"; upon which return of the needle to zei-o begins.

It is quite otherwise when the effect in the nerve is due to the passage of the
excitatory condition (the "discharge") from a nerve centre. When this latter is
excited and the electrical change in the efferent path observed, it shows itself as a
slowly developed deflection which gains speed as it moves, does not stop immediately
when the stimulation is left ofi", and only slowly returns towards zero, which it
frequently does not completely reach. This peculiar character of the deflection move-
ment is special and easily recognised. If the activity of the centre be impaired,
eg., by cooling, drying, &c., the movement is still more sluggish, and is perhaps of
more interest ; it is late in development, i.e., does not commence for one or two seconds
after the excitation has begun.

3 s 2

500

MESSRS. F. (lOTCH AND V. IIORSLKY

A similar increase of the normal latency in the discharge of the cord, we have also
observed by the graphic method, and delay of the kind is frequently noticed in the
cortex cerebri (Francois Fuanck) when its activity is depressed l)y any of the
asrencies referred to on ik ■18.'^.

Section 4.— Tiik Lvfluence Exekted by its Central Attachments upon the
Electrical Changes evoked in a directly Excited Nerve.

We found, in the course of our experiments upon the nerve roots, that the amount
of the electrical change in the nerve evoked by direct excitation of its fibres in any
root, varied with the condition of the nerve centres, and was always less when the root
was cut away from its central attachments.

We, therefore, designed a series of experiments to test the extent of tliis influence
in the case of the anterior and posterior roots respectively.

This series of experiments consisted in the exposui'e of the cauda equina, after
division of the spinal cord in the dorsal region, connecting the central end of the cut
sciatic nerve with the galvanometer, and then raising the posterior roots seriatim, and
exciting each with the platinum electrodes vinder all the precautions stated in
Chapter lit.. Section 4. After measurement of the effect observed in the nerve
to follow excitation of any posterior root in continuity with the spinal cord, the
root was divided about its middle and the peripheral end gently ligatured, raised in the
air and excited. The resulting electrical change or effect in the nerve was noted
and compared with that obtained from the root when first excited in continuity with
the cord.

In every case a diminution was observed in the effect as a direct consequence
of severing the peripheral portion from the central apparatus. We feel justified
in attributing this change to the cause assigned since the immediate local effect of the
section on the peripheral fibres would be to raise rather than depress their excitability.
The result, if marked, although conceivably due to the depressing effects of injury,
is possibly associated with a loss of a beneficial (nutritive 1) influence derivable from
the spinal centres. The deflections observed were as follows : —

Averages.

Posterior root excited.

Effect in nerve.

Root excited in
contiuuity
with cord.

Root cut and

peripheral

end excited.

6th lumbar

7th lumbar

1st sacral . . . .

267
259
212

227

172

75

ON THE MAMMALEAN NERVOUS SYSTEM.

501

Averaging all the observations together, we find that the excitatory effect observed
in the nerve, when the uncut root was excited, amounted to 237 scale, whereas that
evoked by stimulating the peripheral end was 152 scale. The reduction amounts
roughly to about 30 per cent, of the original value, and is especially marked only in the

Fiff. 24.

case of the posterior roots. This fact led us, at first, to believe that when a posterior
root was excited in its continuity, the spinal centre being of necessity aroused, there
must be a flow of energy from the latter down the root, and that this would provide
the greater effect noted. This view, however, was materially shaken by the small
amount of the reflex discharge obtainable from a nerve centre. We, therefore, regard

502

]^1KRSRS. F. GOTCil AND V. IIORSLEY

the plienonieiioii from the more general point of view expressed above until i'nrther
research shall make the question clearer.

MeanwhiU' in this ])laee we wish to add a few remarks on a curious result which we
occasionally obtained (and the meaning of which is very doubtful) by varying the
above-mentioued experimental procedure.

As before, we connected the sciatic nerve with the galvanometer, and exposed the
Cauda equina after section of the cord in the dorsal region. Each posterior root was
now excited in continuity with the cord, seriatim. Tlie anterior roots were then cut,
and the excitation repeated.

In this preparation, therefore, we had the arrangement indicated in the diagram,
(Fig. 24.) The first experiment of the kind was simple and uncomplicated, and may
therefore serve as an examjile.

Cat (177). Sciatic Nerve and Roots of the Right Side prejmred after Section of

the Cord in the Dorsal Region.

Time.

Excitation.

Duration of

.same.

Posterior root excited
in continuity.

Detlection,

galv. scale.

Reflex action as
observed in muscles.

12.2(3

4000

5 seconds

7th lumbar ....

70

Good reflex action

]2.3i

J»

51

1st sacral ....

46

12.39

1)

11

6th lumbar ....

62

12.45

Anterior roots of the above

numbers divided.

12.50

4000

5 .seconds

7th lumbar ....

380

Very vigorous reflex,
much more than before

12.53

»>

)i

1st sacral ....

240

12.65

))

))

6th lumbar ....

280

i» )) ))

It will thus be seen that section of the anterior roots seemed to enormously
increase the efi'ect in the nerve. We were quite at a loss to understand this result
and therefore repeated the experiments. We then found that this striking difference
was not always obtainable. In one experiment, for example, in which the section of
each anterior root alternated with excitations of the corresponding posterior root, the
rise of effect was only noted with one root — viz., the 7th lumbar — in which the effect
mounted from 130 to 160, 147, 250, whereas the other roots showed no rise. In this
experiment we noted that there was coohng of the preparation, especially of the lower
roots. In a succeeding experiment (Cat, 182) we found that wai-ming the cord had a
very noteworthy effect on the result, whereas warming the root had not. We then
made a fourth experiment (Cat, 268), in which the arrangement of the first was more
closely adhered to

ON THE MAMMALIAN NERVOUS SYSTEM.

503

Cat (268). Influence of Section of Anteriox' Roots on the Excltatoiy Effect in the
Nerve, when one Posterior Hoot is excited in continuity.

Time.

12.3
12.6
12.10

12.20
12.25

12.38

12.46

Excitation.

1000
2000

Duration.

3'5 seconds

Root excited.

7th lumbar posterior, in
continuity

Deflection.

230 Average of two

observations
275

Section of the anterior roots of the 5th, 6th, and 7tli lumbar, and 1st and
2nd sacral.

1000
2000

2000
2000

245 Aver.ige of two

observations
320

Section of the 6th lumbar and 1st sacral posterior roots.

292
310

On summing up these experiments it is evident that the division of the anterior
roots does raise the value of the effect in the sciatic nerve evoked by excitation of a
posterior root in continuity. In the last three experiments that rise is not so greatly
marked as to be more than the supplementary discharge of the spinal centres might
account for, but this, of course, does not explain the high readings in the first case.

Section 5. — The Speead of Reflex Discharges up and down the Cord

(Internuncial Fibres).

We have frequently spoken of the existence of internuncial fibres connecting the
various centres. It must, however, be confessed that the evidence of their existence is
founded almost entirely upon our knowledge of the laws of reflex spread as stated by
Pfluger. (See Section 1, p. 484.)

We endeavoured to ascertain what light the present galvanometric method would
throw upon this spread, and thus upon the localisation and characteristics of these
paths. To elucidate this we made a series of experiments in which the cord was
excited^ and the electrical changes noted at the same time that the muscular
movements were either recorded or carefully observed.

(1.) Ascending Discharges (Reflex).

In the case of reflex discharges it is possible, after division of the cord, to observe
the electrical changes in the upper end of the lower fragment of cord, and at the same
time to note the contractions of a muscle of the lower (hind) liml).

504

MESSES. l'\ (;UTCJI AND V. IIUUSLEY

Since it was desirable to ascertain tlie relationship between the pei'iodicity of the
electrical changes in the cord and that of the muscular twitches, the capillary electro-
meter was used.

The spinal cord of the Cat being divided in the mid-dorsal region, tlic upper end
of the dorso-lmnbar portion was prepared for connection with the electrometer, wliilst
the lower tendon of the rectus femoris muscle was cut and attached to one of Pick's
spring recording levers (isometric method). The cord was then exposed over the
dorso-lumbar junction and the lateral column excited with stimuli of vai-ying degrees
of intensity. It is obvious that electrical changes thus produced in the cord are due
((f) to the excitation of continuous fibres joining the excited portion of the lateral
column with the observed region, (/>) to the excitation of indirect fibres in the
lateral column, internuncial or otherwise, (c) to the discharge of nerve impulses
up the cord from the aroused centres.

The main interest of the experiment lies in the fact that with a weak stimulus no
muscular eftects were observed, although distinct electrical effects were produced ;
whilst with a strong stimulus the centres in the cord were so aroused that they
continued to discharge after the stimulus had ceased, these discharges being evidenced
both by muscular contractions and by effects in the electrometer.

The following Table indicates this result : —

ExPEElMENT. Cat (104). — Afferent Effect in Cord compared with Contraction of
the Muscle, simultaneously evoked by Intermediate Excitation of the Cord.

Time.

Excitation.

Duration
of same.

Effect in cord ; electro-
meter.

Effect in muscle ; spring
myograph.

4.50
4.54
4.55

4.57

1500
3000
4000

5000

5 sees.
J)

Persistent 2 divs. . . Nil

6 divs. . . Twitch
,, 7 divs. . . Initial effect and continued

tetanus
„ 11 divs., fol- Initial characteristic (cord) '
lowed by intermittent contraction followed by |
after-effects well-marked after-effect

1

It is thus seen that Avhen a lumbar spinal centre is aroused to discharge, a flow of
nerve energy takes place up the cord as well as out by the efferent roots, and, in
all probability, impulses thus pass upwards to various groups of centres both in the
upper regions of the cord and in the encephalon.

As to the spinal channels by which these impulses travel, — whether internuncial,
direct efferent fibres (pyramidal tract) or direct afferent fibres (posterior column) — the
experiment gives no information, and we have not had an opportunity of repeating it
on an animal in which, by previous operations, the two latter have been more or less
eliminated. Some evidence is afforded by the following experiment, in which the left

ON THE MAMMALIAN NERVOUS SYSTEM.

505

posterior column having been divided at the 9th dorsal vei'tebra, the left 1st lumbar
posterior root was excited, and the electrical changes in the cord noted galvano-
metrically at the same time that the muscular movements were observed.

It will be seen that the electrical change in the dorsal region of the cord was
indicated by an effect of 55 in the galvanometer, when the lateral column on the side
of the excitation was uninjured ; whereas hemisection between the root and the
observed region cut down the galvanometric effect to only 7. Before the interruption
by the hemisection, a reflex contraction was observed during the excitation in all the
muscles supplied by the various nerve segments between the excited root and the
observed area of cord, whilst, after the section, such reflex contraction was limited to
the piece of cord below the hemisection. The particulars of this second experiment,
which appear to suggest that the internuncial fibres are localised in the lateral
column, are as follows : — ■

Cat 376. Cord divided, and peripheral end connected to the galvanometer at the
intervertebral disc between the 8th and 9th dorsal vertebrte, section of the left
posterior column opposite 13th dorsal vertebra.

Time.

Excitation.

Duration of
excitation.

Part excited.

Effect in
galvanometer.

Effect in neighbouring
muscles.

6.3

8000

(Strong

stimulus)

5 sees.

Left 1st lumbar
posterior root

55

Powerful reflex in
trunk muscles

Left hemisection of the cord was then performed at disc between 13th dorsal and
1st lumbar vertebrae.

Time.

Excitation.

Duration of „ , ., ,
excitation. ^^^'^ '^^'''*«"^-

Effect in
galvanometei".

Effect in neighbouring
muscles.

6.12

8000

5 sets.

Left 1st lumbar
posterior root

7

Reflex limited to
muscles below the
point of hemisec-
tion

(2.) Descending Discharges.

While PFLtJGER's law concerning the upward internuncial discharge of spinal nerve
centres undoubtedly prevails, nevertheless the anatomical investigations of Lockhart
Clarke, the physiological observations of Brown-Sequard, and the anatomical
researches of GoLC4i, recently confirmed by Kamon y Cajal, show that there must

MDCCCXCI. — B. 3 T

506

MESS US. F. GOTCII AND V. IIORSLEY

be descending channels from at least the afferent sides of the spinal nerve centres,
even if there are not direct descending- channels. In fact, Golgi's work goes to show
that there are such direct paths, and this is extended by Kolliker. We commenced
the investigation of this point by the galvanometric method with an attempt to

Fig. 25.

ascertain whether any electrical change can be detected in the cord below the entry
of the stimulated afferent nerve, and have obtained positive results. Our plsm was
to divide the cord in two places, as in Chapter VIIL, and to connect the lower
lumbar end of the tract included between the sections with the galvanometer. (See
ficr. 25.)

OS THE MAMMALIAN NERVOUS SYSTEM.

507

The trunk of the last dorsal nerve was then excited.

Cat 378. Cord divided between 8th and 9th dorsal vertebrae, and at disc between
3rd and 4th lumbar vertebrae.

Excitation of left 13th dorsal nerve, effect In lower (lumbar) ]>ortion of cord.

Time of
observation.

Intensity of
stimulus.

Duration of

stimulus.

Effect in
galvanometer.

Effect in

muscles.

11.83
11.8(5

2000
8000

seconds.
5
5

26
31

Fair reflex
Fair reflex

In order to see whether this effect was transmitted by the fibres in the posterior
or lateral column, the left posterior column was now divided opposite the centre of
the body of the 1st lumbar vertebra. The excitation of the root now gave the

following effect : —

Time of
observation.

Intensity of

stimulus.

Duration of

stimulus.

Effect in
galvanometir.

Effect in
muscles.

11.58

12.4
12.15

8000

Left lateral col
8000

seconds.
5

umn divided (he
5

Left 13th -dorsal 8
nerve, central cud

nee complete left hemisection).

Left 13th dorsal G
nerve, central end

Good reflex

We here see most inimistakably that the descending effect is very different in
character and localisation from the ascending effect previously described, for (l) it is
much smaller, (2) it is mainly dependent upon the continuity of the posterior column,
wdiereas the ascending indirect effect is dependent upon the integrity of the lateral
column.

These results, which were quite unexpected by us, are in very remarkable
agreement with the anatomical researches of Golgi, &c.

Section 6. — On the Electrical Changes in the Cortical Nerve Centres.

In all the preceding investigations we have studied the characteristics of a nerve
centre by examining the electrical changes, and thus the excitatory processes present
in the nerve fibres connected with such a centre. We chose the spinal cord since it
contains the simplest form of a complete centre that the Mammalian nervous system
presents.

3 T 2

508

l^IKSSKS. V. r;OTC]l and v. HOllSriKY

]n ;in early stage of oui' work the possibility occurred to us of investigating the
centres in the cerebral cortex, by the use of the electrical method, this plan having
been previously employed by Caton (1875). Wo selected the occipital lobes (see
fig. 26), and endeavoured by stimulation of the retina to obtain evidences of definite

electrical changes in the cortex of the moderately etherised Cat, consequent upon the
arrival therein of nerve impulses. The difficulties of effectual isolation and the
uncertainties of the area of connection, comprising as it does a mass of subjacent
nerve fibres, seemed to us to account for the indefinite and capricious character of
the electrical indications as displayed by the galvanometer.

ON TllIO MAMMALIAN NERVOUS SYSTEM. 509

The experiments of Beck, of Fleischl von Marxow, and of Danilewsky in
this connection have been referred to in Chapter 11., which deals with the history of
the electrical method. With reference to these we would here only point out that
the experiments of Setschenow, Weriho, and ourselves, to which in his Polish paper
Beck alludes, seem to us to indicate that the observed electrical changes may have
their seat in nerve libres proceeding from or to centres, rather than in any form of
central, i.e., corpuscular, nerve mechanism.

Any electrical changes in the contacts of the electrodes with the cortex might,
therefore, be as much caused by the impulses travelling in the subjacent fibres as by
chano-es in the cells or other surface termini.

The above criticism obviously does not affect the evidence which the presence of
definite electrical changes offers as to localisation, but seems to us to discount any
advantage which the method might offer for the investigation of the functions of a
nerve centre by the study of the actual changes in its cells.

Summary.

It is obviously impossible to formulate with sufficient accuracy many general
conclusions as to the mode in which the functional activity of the spinal nerve centres
is generated and discharged, but we cannot refrain from pointing out that the
electrical method supplies many ways of investigating this very important and
difficult subject, and further that it throws most unexpected light on the obscure
questions relating to the working of the several parts of nerve centres.

Most prominently stand out two principal facts : —

(1.) The kinetogenetic portion'" of a spinal nerve centre is probably the aff"ei-ent
side. (See Section 2.)

(2.) The part in which the delay and diminution of impulses passing through is
effected is the efferent side of such a centre with the field of conjunction.
(See Section 2.)

The consideration of these facts, unexpected as they were, shows clearly that the
interpretation by BASTiANt of sensori-motor nerve phenomena is j)robably the most
correct yet advanced. The basis of that interpretation, namely, the doctrine of
kinpesthesis as formulated by him, is in complete harmony with oui- experimental
results.

Bastian has for many years contended that the ordinary division of nerve centres
or parts of the same into sensory and motor respectively is misleading, and that not
only has the statement that the source of a nerve centre's discharge is the motor part

* By tlieterm "kinetogenetic portion" we have styled that part of a centre in which the potential
energy is converted into kinetic.

t " The Bruin as an Organ of Ifind." " Paralysis, Cerebral, Bulbar, and Spinal."

510 MESSRS. F. GOTCH AND V. HORSLEY

of it no basis of actu.al fact, hut tliat it is more pliilosopliical to associate such source
witli the sensory jmrt.

Among the considerations which favour this view are the following : — The
undoubted physiological resemblance which exists between the succession of events in
a voluntary or reflex movement, respectively, and the analysis of such psychological
events as are known to be concomitants of the former. The psychological phenomena
of reflex action show that the afferent impulse in all cases precedes the efferent dis-
charge ; psychological analysis also shows that muscular sense impressions similarly
precede those discharges which ev^oke " voluntary " coordinated movements. Hence
he concluded that since the sensory excitation always precedes the eflerent output, the
former must have primary importance, in other words, that the development of kinetic
energy must take place in the afterent side of the centre. To express this view
briefly he coined the term kincesthesis.

As the electrical method aflbrds the first opportunity of an experimental contribu-
tion to this subject, it is interesting to find how strongly its application bears out
Bastian's position. In this chapter it appears probable that the kinetogenetic
portion of the centre is the affei'ent side of it, and the more especially when it is
seen how readily the centre discharges into the afferent nerve channels (i.e., actually
" backwards," as compared to the course of ordinary afferent impression).

The method has further enabled us to ascertain what connections and facilities for
conduction the efferent, or so-called motor side of a nerve centre possesses, and
instead of finding, as might have been expected from the ordinarily expressed beliefs on
this subject, that we had to deal with a source of energy that was readily ai'oused,
and freely connected with its neighbours, we found, to our surprise, that it afforded
nothing of the sort, and that its power of conducting impulses centripetally was
apparently nil.

Curiously enough, this last point was foreseen also by James, who, in his celebrated
work on the 'Feeling of Eftbrt,' 1880, while endorsing the views of Bastian, says
that the "electrodes of the physiologist," if applied to the central end of the anterior
root, would not arouse any "sentient," i.e., afterent impulse in the cord. We are
happy to find that our experimental results, unusual though they were, are, never-
theless, in close agreement with the deductions of the logical method of these
distinguished writers.

ON THE MAMMALIAN NERVOUS SYSTEM. 511

CHAPTER XII.— ON THE ELECTRICAL CHANGES EVOKED IN THE SPINAL CORD
AND NERVES BY THE ACTION OF ABSINTHE AND STRYCHNIA.

Section 1. — Experiments (control) including the Use of Absinthe.

Very early in the present investigation on the effects produced by electrical excita-
tion of central apparatuses, the advisability of obtaining a control of our results
presented itself The use of strychnia by DU Bois-Keymond as a control demon-
stration of the excitatory electrical state in nerves, formed, as it always must, the
basis of such a method.

We, however, first selected absinthe as particularly exciting nerve centres, and,
according to one of us* (V.H.), specially the cortex. Absinthe has been known since
the experiments on animals by MABciit to cause, in small doses, effects only explicable
as poisoning of the highest cerebral centres, thus producing mental changes, delirium,
hebetude, &c., while in larger doses it evokes clonic epileptiform convulsions with
stertorous respiration, &c.

MagnanJ has, of all writers, contributed the most to our knowledge of the action
of this substance.

By injecting small quantities of the essence of absinthe either into the stomach or
into a vein, he produced a very striking series of phenomena, which he showed were
identical with idiopathic epilepsy.

Following up its action more closely he endeavoured to ascertain the share taken in
the production of the fits by the brain and spinal cord respectively. This he investi-
gated by first dividing the cord at the atlanto-occipital articulation, and then injecting
absinthe, artificial respiration being kept up.

Unfortunately, he only desciibes one experiment (No. 6) in which the cord was
completely divided. He concludes, however, from his other experiments that the
drug excites simultaneously the lower (spinal) and the upper (cerebral) centres. It
has, however, been shown by one of us§ that complete section of the cord at the
8th dorsal vertebi'a prevents the appearance of the characteristic convulsions in the
muscles of the parts below the level of the section ; and, further, that the corresponding
parts or limbs, if the excitable cortex of one side be removed, will not take part in
the first general epileptiform convulsions which follow the injection of the drug,
whilst, even when in the subsequent fits such parts are affected, the muscles only pass
into a slight degree of tonus.

Until a more extended series of I'eseai'ches should negative these points, it is

* ' Brown Lectures.'

t ' Comptes Rendus de I'Academie dcs Sciences,' vol. 58, 1864, p. G28 ; also Ajioky.

X ' Reclierches sur les Centres Nerveux,' Paris, Manon, 1876.

§ 'Reports of the Brown Institution.'

512 MESSRS. F. COTCII AND V. HORSLEY

reasonable to accept the conclusion tluit ;il)sinthe excites par excellence the complex
cortical centres.

It will l)e well, perhaps, to mention that by the method of simple observation the
following phenomena are elicited by the excitatory action of absinthe.

After injection of two drops of essence of absinthe* into the jugular vein of an
animal, narcotised with ether to the degree of unconsciousness, there occur, after a
sufficient interval has elapsed (30 seconds or so) to permit of the translation of the
poison through the heart and lungs to the arterial system, and so to the brain, the
following events : —

The small facial muscles begin to twitch in single clonic spasms, next passing into
a state of tremulous tonic spasm. This order of convulsion passes rapidly down the
body, until the tonic spasm in the limbs is extremely marked. After this has
obtained for a period varying with the dose injected, the tonic spasm gives way to a
long series of clonic twitches. Accompanying these motor " discharges," there is
profuse salivation, and sometimes escape of urine, while in cases in which narcosis
has not been preliminarily employed, unconsciousness and coma are early symptoms.

It is thus obvious that absinthe affords a very simple and efficient means of
chemical excitation for the purpose of testing the validity of the results of electrical
stimulation of the motor cortex.

We have employed it under the following experimental conditions : —

(1.) Connecting the nerve to the galvanometer or electrometer.

(2.) Connecting the spinal cord to the galvanometer or electrometer, at the same
time observing the contractions of the muscles in different parts.

The method employed was, so far as the electrical connections of the observed parts
were concerned, precisely that described in the preceding pages. Special precautions
had to be taken against any agitation of the preparation. The narcosis with ether
being temporarily maintained, the external jugular vein of either side was exposed,
penetrated by the needle of a hypodermic syringe, and two minims of the essence of
absinthe injected, the narcosis of this drug taking the place of the etherisation,
which was forthwith terminated. It was invariably observed that the galvanometer
showed evidences of excitatory electrical changes in either nerve or cord, before the
muscular contractions in the immediate neighbourhood of the observed level became
accentuated.

In respect of what has been said above, we may here add that we never saw con-
tractions of muscles innervated from points below the level of section in the spinal
cord when that was divided.

* ObtainaUc of Messrs. Hopkin and Williams.

ON THE MAMMALIAN NERVOUS SYSTEM. 513

(1.) Excitation of the Intact Central Nervous System hij Ahsinthe. — Electrical Effects

in the Sciatic Nerve.

The results obtained by this arrangement were as follows : —

(a.) Observations loith the Galvanometer. — The experiments were made on four Cats,
one to fifteen drops of absinthe being injected. The preliminary effects noted in the
galvanometer while as yet the spasms were limited to the upper anterior muscles of
the body amounted to small deflections, of which tlie following are examples : 4, 5, 5,
8, 10, 10, 13, 15.

When, however, the toxic effect of the drug became more marked, and the
discharges summated produced a fully developed fit, as above described, then the
deflection of the galvanometer was notably increased, and ran up to maxima of
135, 150, 150, 185, 210, 272, 275, these averaging nearly 200. In smaller attacks,
but still generalised, the deflection varied from 45 to 85. With regard to the
magnitude of these deflections, it is to be noted that their large amounts are
doubtless due to the completely bilateral and prolonged character of the cortical
discharges and the continued effect on the galvanometer.

{b.) Observations ivith the Electrometer. — When during any one of the above
observations the electrodes in contact with the central end of the nerve were switched
over so as to bring it into connection with the electrometer, the eff'ect was visible, but
varied notably in amount from a small movement of the mercurial meniscus to 1
division.

It was thus abundantly clear that the employment of absinthe gave maximal
electromotive effects in the sciatic nerve througli the overwhelming discharge of the
highest cortical centres, and the figures obtained from the galvanometric deflections
give a fair notion of the comparative magnitude of these discharges, the more
especially as the degree to which the epileptic discharge developed could be plainly
seen in the amount and duration of the spasm iii the neighbouring muscles. This
being so, it is, of course, hardly necessary to add that the cessation of the convulsion
was synchi'onous with a halt in the swing of the galvanometer needle.

(2.) Excitation of the Nervous System by Absinthe. — Electrical Clianges in the

Sjnnal Cord.

As might have been anticipated, the electrical effects observed in the spinal cord
when the central mechanisms were excited by a diffuse stimulus, such as that of
absinthe in the cir-culation, were larger than that witnessed in the peripheral nerve.

We have made four experiments in Cats, and one in the Monkey {Macacus rhesus),
and have usually measured the effect with the electrometer. With this instrument
the excursions of the mercury were very marked, varying from slight movement in
the case of initial spasms to excursions of even five divisions in severe fits.

MDCCCXCI.-B, 3 u

514 MESSRS. P. GOTCH AND V. HORSLEY

la the galvanometer, similarly, the lowest recorded deflection was 110, the hio-hest
being over 300 and obviously conditioned by the duration of the convulsions.

No comparison can be drawn between the magnitude of the deflections and those
obtained from the nerve, for the above and other reasons. In general they confirm the
validity of the views advanced, respecting the results obtained by electrical excita-
tion of the central nervous system as detailed in the foregoing chapters.

An experiment in this connection may here be mentioned as of special interest
regarding the action of absinthe. The toxic action of this drug is unquestionably
excitatory as far as the highest nerve centres are concerned. FRANgois France*
found that it produced inexcitability of the cortex. One of us has pi'eviously
suggested that this is a question of dosage, in common with other narcotic agencies.
In the present series, while observing in one case the deflection produced by absinthe, we
superadded electrical excitation of the previously exposed cortex. The result (Cat 99)
was to notably increase the effect. Thus, while the chemical stimulus was actually
evoking a powerful discharge, the additional electrical excitation caused the centres
to produce still more kinetic energy. In this stage, therefore, the absinthe had not
exhausted the cortex. The additional eftect was visible, both in the electrometer and
galvanometer, when either was switched on to the cord.

Section 2. — Experiments (control) including the use of Strychnia.

The method of employing strychnia differed from that detailed in the case of
absinthe in one important jjarticular, namely, that the spinal cord was severed from
its connection with the encephalon. The spinal cord in two Cats and one Monkey,
and the sciatic nerves and the posterior root in two Cats, were connected with the
galvanometer for observation of the excitatory electrical changes. The solution of
strychnia employed was a 1 per cent, solution of the acetate. It was injected into
the peritoneal cavity, and the toxic symptoms observed. The tetanic spasms,
like those obtained by absinthe, were of very varying force and duration, but gave far
higher readings than absinthe. We will return to this point later, and meanwhile
briefly state the results obtained.

(1.) Electrical Effects in the Sciatic Nerves. — The observation of the effects in the
Mammalian nerve during the discharges due to strychnia has furnished valuable
control results. We have, however, used the method to obtain by its means the
evidence referred to in the preceding chapter on the discharge of a nerve centre
backwards down the posterior root as well as the discharge down the anterior roots.

(2.) Electrical Effects in Posterior Root. — Similar changes to those in the nei've have
been obtained in the case of the posterior root. These have been already referred to
in Chapter XL, Section 3.

(8) Electrical Effects in the Spinal Cord. — When the spinal cord of an animal was

* Loc. cit.

ON THE MAMMALIAN NERVOUS SYSTEM.

515

severed, and the lower fragment connected in the manner practised in all the
preceding experiments, and strychnia injected, very large electrical changes were
observed. These were pi-oduced in the upper, as well as the lower, end of the dorso-
lumbar portion of the cord. The maximal effects were over 500 scale of galvanometer,
and the average in one experiment of all the readings was as high as 300.

To sum up these experiments with a chemical stimulus, from the point of view from
which they were designed, it is clear that the excitatory electromotive changes {i.e.,
diminution in the resting difference) observed to occur in the central nervous system
when that is excited electrically or mechanically, are true indications of physiological
phenomena accompanying functional activity, and, further, that tlieir amounts vary
directly in proportion with the intensity and duration of such activity.

CHAPTER XIII.— SUMMARY AND CONCLUSIONS.

The consideration of the results given in the foregoing chapters shows, we venture to
think, that the electrical method of investigating the localisation of nerve impulses in
the Mammalian nervous system is one which has furnished several new aspects of nerve
function, and we believe that if further pursued it will prove one of the most valuable
means of differentiating the structure of the nervous system, and gauging the nature
of the functional activity of the nerve centres.

In view of the extended scope of our present research, we feel unable to give a suffi-
ciently brief summary of the results. We therefore propose to enumerate, by way of
conclusion, some of the general principles which we think we are justified in deducing
from our work.

The following remarks, therefore, cannot in any sense be regarded as embodying
the whole of our investigations, and we must consequently refer our readers to the
individual chapters, and especially to the remarks at the end of each, for information
as to points upon which we do not here touch.

(1.) Resting Electrical Difference.

The resting electrical difference between the cut and uninjured longitudinal
surface respectively in the Mammalian mixed nerve, spinal nerve root, and spinal
nerve, has been found by us to liave the following value: —

Nerve

Root

Cord

Cat.

Monkey.

(69 cases) -01 Daniell
(5 cases) -025 „
(50 cases) -032 „

(12 cases) '005 Daniell
(10 cases) -022

3 u 2

51 G MESSRS. F. nOTCII AND V. TTORSLKY

We have foimil, I'lirtlioi-, that the (lifVerence is subject to variations of which
the following are the most important : —

(of.) A notable fall is observed in all three; tissues in consequence of systemic
cleatli.

(b.) The difference in the cord is inci-eased after the functional activity of the organ
has been aroused.

(('.) The difference in the cord is more pronounced when the tissue is in connection
with the encephalon. (See Chapter IV.)

(2.) Effect in the Spinal Cord on Excitation of tlie Cortex Cerebri.

In extension of our discovery that excitatory electrical effects can be observed in the
spinal cord to result from excitation of the cortex cerebri, as previously described, and by
means of which the character of the impulses derived from the cortical centres may be
studied, we have obtained effects in both the cord and the mixed nerve, following
similar excitation of the cortex, and by comparing the result of these observations
together have found that the excitatory state evoked by cortical activity undergoes
a diminution of over 80 per cent, in the passage from the cord into the sciatic nerve.

We have also applied the galvanometric method to differentiate the cortical excit-
able areas, by recording and comparing the discharges from the same in the spinal
cord, and have found a striking degree of localisation demonstrated thereby in the Cat
as well as the Monkey. (See Chapter V.)

(3.) Effect in the Sj^inal Cord on Excitation of the Corona Radiata.

By a comparison made between the amounts of the electrical effects produced
in the spinal cord and the mixed nerve respectively after excitation of the corona
radiata, we found that the cord effect is four times as great as that in the nerve, and
further, that this corona-radiata-to-cord effect is little more than half the complete
cortex-to-cord change. (See Chapter VI.)

(4.) Bilaterality .

By comparing the records of the extent of the electrical effects in each half of the
longitudinally divided cord, and in the mixed nerve of each side, we have made fi-esh
observations on the important and complex question of bilaterality of representation
in the central nervous system. We have found —

(o.) That it is possible to obtain strictly unilateral effects in both the spinal cord
and sciatic nerve with complete excitation of both cortex cerebri and corona radiata.

[b.) That the circumstances which favour the production of bilateral effects are
such as bring into play other portions of the central nervous system, e.g., the opposite
excitable cortex, cerebellum, and basal structures ; and,

ON THE MAMMA.LIAN NERVOUS SYSTEM. 517

(c.) That such bilateral effects, under these circumstances, can be evoked nioi'e
readily by excitation of the coi'ona radiata than of the cortex.

Hence, we conclude that, as far as the cortical efferent representation of the lower
limb in the Cat and Monkey is concerned, the normal condition is that of
unilaterality. (See Chapter VII.)

(5.) Electrical Changes evoked in the Spinal Cord hy Excitation of its Columns.

Observation of the electrical changes in the dorso-lurabar spinal cord, when
evoked by direct excitation of its fibres after severance from the encephalon, has
revealed by comparison of the electrical changes produced the proportionate existence
of direct and indirect channels in the various columns of the cord.

We have thus examined the columns so far as they conduct ascending and descending
impulses respectively in the Cat and Monkey. This analysis we extended by employing
the exclusion method of intervening sections, as an addition to observations on the
intact cord. The results show that . —

(o.) In the Monkey a relatively larger number of direct fibres are contained in the
lateral column than in the posterior column, the reverse being the case in the Cat.

[h.) For both classes of impulses and of animals observed we have obtained (a) no
evidence of crossing between the lateral columns, (/S) evidence of indirect connections
between one posterior column and the lateral column of the same side, (y) evidence of
cross connections between the posterior columns.

(c.) There is no evidence of the existence in the anterior columns of the cord (Cat
and Monkey) of any continuous fibres between the mid-dorsal and lumbar regions.

(d.) The spread of impulses from path to path in the spinal cord appears, in
addition to what is stated under (6), to be conditioned as follows : — The posterior
column fibres spread into other columns more as they ascend than as they descend,
whereas the fibres of the lateral column spread in a converse manner. (See
Chapter VIII.)

(G.) Tlte Relations of the Spinal Cord to the Lumbar Nerves.

We have investigated the relation of the peripheral nerves and their roots to the
paths and to the bulbo-spinal centres in the dorso-lumbar region of the spinal cord.

The investigations consisted in observing (A) the electrical changes produced by
exciting the mixed nerves or their roots in the spinal cord when separated from the
encephalon, (B) the electrical changes produced by excitation of the divided cord in
the spinal nerves, and (C) the excitatory changes produced by stimulation of the
spinal centres.

A. The results obtained by the first method may be grouped as follows : —

(a.) By far the majority of afferent impulses ascend the cord on the same side as

518 MESSRS. P. GOTCH AND V. TTORSLEY

the entcrliiy root, Wotli by direct iiiul indirect paths, a small minority ascend by the
posterior column of the opposite side, and a mere fraction by the lateral cohniui of tlie
opposite side.

(b.) Tlie direct path of afferent iinpulses is localised in the posterior cohnnn of the
same side as that of the nerve or root excited.

(c.) The indirect paths of afferent impulses are localised in the posterior columns of
both sides, and in the lateral column of the same side as that of the excited ncrvo.

{(I.) The proportionate development of both systems of nerve paths in the two
sides of the cord may be inferred from the percentages of the total transmission of
excitatory electrical changes from the afferent nerve to the cord.

Of the electrical changes there are transmitted by : —

Same side.

Posterior column of same side as excited nerve . . 60 per cent.] ^^

X , 1 ^^ rSO per cent.

Lateral ,, ,, „ ,, . . 20 ,, J ^

Posterior column of opposite side to ,, . . 15

(See Chapter IX.)

Opposite side.
1 20 per cent.

B. The descending electrical effects, as far as the relationship of the cord and nerves
is concerned, vrere investigated in the mixed nerA^e, the spinal cord being severed from
the encephalon and excited.

The results of these experiments, obtained with both minimal and maximal
stimuli, and controlled by the exclusion method of intervening sections, were as
follows : —

(«.) On minimal excitation of the posterior column, impulses are directly trans-
mitted into the posterior roots of the same side, and so into the mixed nerve ; with
maximal excitation some impulses are similarly transmitted through indirect paths.

(6.) On maximal excitation of the posterior column, impulses are transmitted i)y
indirect paths across to the posterior roots of the opposite side, and so to the mixed
nerve of that side.

(c.) On excitation of the lateral column, impulses ai'e indirectly transmitted to the
mixed nerve of the same side as that of the excited column.

(d.) The proportionate percentage of the total transmission of excitatory electrical
states from the vai"ious excited columns of the dorsal cord into the mixed nerves is as
follows : —

per cent. per cent.

Excitation of posterior column of same side as observed nerve 73~1 ^ .,

^ S'bame side 82

,, lateral ,, ,, „ 9 J

,, posterior ,, opposite „ IS"!

,, lateral ,, „ „ o J

ON THE MAMMALIAN NERVOUS SYSTEM. 519

(r.) Olu- evidence shows that electrical states {i.e., impulses) are transmitted w itii
great facility from the excited areas in the cord down the afferent channels of the
cord, posterior roots, and nerves (see Chapter X.).

C. The researches into the relationship between the spinal cord in the doi'so-
lumbar region and the mixed nerve and spmal nerve roots, have enabled us to formu-
late some generalisations on the functional activity of the s[)inal nerve centi'cs as
follows : —

(a.) There is complete obstruction to all centripetal impulses which may reach the
cord by the central end of the anterior root.

(b.) A marked quantitative diminution as well as delay in time is suffered by impulses
which leave the spinal cord by the anterior roots, whether these have originated in
the cortex cerebri, corona radiata, or the lateral columns of the cord.

(r.) An increased resistance to descending, as compared to ascending, impulses by
certain indirect paths is offered connected with the afferent side of the spinal centres.

(d.) Whenever a spinal centre discharges, nerve impulses pass from it down the
posterior I'oots as well as the anterior.

(e.) The effect produced in a mixed nerve by the reflex discharge of a spinal centre
down the nerve fibres is notably small when compared with that evoked by their
direct excitation (see Chapter XI.).

It will be gathered from the extent and vai-iety of the above conclusions that the
employment of the method used in the above research has led us on from one investi-
gation to another. We commenced our experiments with the object of ascertaining
the character of the cortical discharge, and we employed for this purpose the capillary
electrometer.

We then made use of the galvanometer for the same purpose, and at once found
that a method was opened up for investigating, not merely the general characters of
a cortical discharge of impulses, but the comparative amounts of such impulses when
generated in different parts of tlie Mammalian nervous system. This led to its em-
ployment as a means of ascertaining the distribution of the channels in the spinal cord
along which these impulses passed, and thus to the determination of the extent to
which afferent as well as efferent nerve impulses were localised in fibres on one or the
other side of the cord. The necessity of stimulating for this purpose the various roots
of the nerves brought before us in a most striking manner the remarkable difference
between the central connections of the two kinds of roots, and thus finally ojaened up
the possibility of new investigations into the anatomical relations of a centre and the
particular physiological attributes which characterise it.

The correctness of the method is, we think, shown by the way in which the results
set forth in the foregoing chapters were progressively obtained by its use. Many of
them were unexpected by us, and needed ample verification. We had thus an
opportunity, in frequent repetition of the same experiment, of probing the extent to
which the data our method furnished could be relied upon. From this point of

520 MESSRS. F. GOTCII AND V. HORSLEY

view wc arc firmly convinced that, v^^lien proper precaiitions are used to avoid the
disturbing influence of capricious factors, the present plan is one which gives as sure
indications as any other method which has been used in the carrying out of investi-
gations into the central nervous system, whilst by its mode of application it has the
merit of ensuring definite localisation in hitherto unexplored regions, 'i'he wide extent
of the field of research which the use of the method opens up is obvious. We have
only einployed it for elucidating a few of the phenomena which are exhibited by the
functions of the brain and spinal cord ; but the functions of various ganglia, the
relations of the sympathetic system, of the more central portions of the bul bo-spinal
system, and finally of the encephalic structures may, and undoubtedly will, be satis-
fixctorily approached in the future by its means.

ON THE MAMMALIAN" NERVOUS SYSTEM. 521

Appendix A. — Topogi-aphical Relations existing between the Superficial Origins from

the Spinal Cord of the Spinal Nerves, and the Bodies of the Vertebrae in the Cat.

Ci£Rvic.\L Nerve. Corresponding Vektebr.e.

I Upper border of 1st cervical

II Upper half of 2nd cervical

III Middle of 3rd cervical*

IV Upper border of 4tli cervical*

V Intervertebral disc between 4th and 5th cei'vical

VI Lower border of 5th cervical and disc between 5th and 'ith

cervical

Vir Lower half of 6tli cervical

VIII Centre of 7th cervical

Dors.vl.

I Disc between 7th cervical and 1st dorsiil

II Lower border of 1st doisal and disc between 1st and 2nd dorsal

III Lower half of 2nd dorsal

IV Lower half of 3rd dorsal

V Lower two-thirds of 4th dorsal

VI Centre of body of 5th dorsal

VII Lower half of 6th dorsal, and disc between 6th and 7th dorsal

Vlll. . Lower half of 7th dorsal, and disc between 7th and 8th dorsal

IX Disc between 8th and 9th dorsal

X Intervertebral disc between 9th and 10th dorsal

XI Upper half of body of 11th dorsal

XII Middle of body of 12th dorsal

XIII Lower half of body of 13th dorsal

Lumbar.

I Lower half of 1st Ininbar, and disc between 1st and 2nd biTnbart

II Lower quarter of 2nd lumbar, and disc and upper quarter of

the 3rd lumbar centre opposite discf

III Disc between 3rd and -Ith lumbar

IV Lower border of 4th lumbar

V ]\Iiddle of 5th lumbar

VI Lower quarter of 5th lumbar and disc between 5th and 6tii

VII Upper fourth of 6th lumbar

Sacr.il.

1 2nd fourth of 6th lumbar

II 3rd fourth of 6th lumbar

III Disc between 6th and 7th lumbar

Coccygeal. — The cord tapers gradually to 3rd sacral vertebra.

* These nerves consequently run slightly forwards from the cord to leave the intervertebral
foramina.

t These nerves are directed slij;hlly forwards on leaving the cord.
MDCCCXCI. — B. 3 X

522

MESSRS. V. (iOTCH AND V. IIOUSLEY

Appendix B. — TaWe sliowinj^ Persistent Electrical Difference between Surface anrl
Cross Section. Distance between Contacts 1 centini. Difference expressed in
fractions of I DanielL

1. Sciatic Nerve.

( 1 ) In Connection with Brain.

(71)*

(73)

(75)

(166)

(273)

(2^8)

(290)

(291)

(292)

(301)

(801)

(296)

(296)

(298)

(298)

(299)

(299)

(303)

(305)

(305)

(382)

(3H4)

•018

•013

•018

•008

•009

•008

•009

•00-1.

•000

•004

■006

■009

•008

•01

•009

•009

•018

•009

•009

•01

•009

•0075

Monkey (47)

•007

„ (217)

•005

„ (217)

•006

., (221)

•005

„ (221)

•006

„ (262)

•003

„ (270)

•003

„ (280)

•004

„ (338)

•005

„ (338)

•00i>

„ (868)

•005

„ (368)

•007

Average 005 D;

Liiit'U.

Averasfe '0094 Daniell.

(2)

Cat

(32)
(36) ,
(36)

(157)

(lol)

(164)

(168)

(172)

(173)

(175)

(177)

(179)

(181) ,

(190)

(200)

(201)

(203)

(204)

(363)

Cord Divided

. . •OOo

. . •oo?

. . •oo?

. . -009

. . •Oil

. . •oog

. . -008

. . -012

. . •oil

. . -007

. . -008

. . -008

. . on

. . -01

. . -008

. . •on

. . •oo?

. . -008

. . ^009

ill Dorsal Region.
Cat (64) . .
„ (65) . .

(206)
(209)
(209)
(254)
(263)
(265)
(267)
(267)

(275) .

(276) .
(311)
(312)
(329) .
(829)
(331) .
(331) .
(363) .

■012

■013

■01

•012

■01

■Oil

•009

•008

•006

•007

■0085

•0125

•0125

•012

•01

•012

•008

•Oil

•007

Average ^0094 Daniell.

* As stated on p. 303, the numbers in brackets following the nientiou of aa animal refer to the page
in our note ledger in which the observation is recorded.

ON TiiK j\i a:\imaltan nervous system.

523

(3.) After Previons Operations on Cord.

Cai (227) division of left posterior roots .

. (227)

, (2'50) division of right posterior column

, (230)

, (247) „

, (2-')l) division of both posterior columns

, (259) left hemisection

, (250) „

•007 left
■009 right
■1)08 left
•005 right
■012 right
■012 right
■009 left
■008 right

II. LuMBAK Roots.
7th Left Lumbar Posterior Root.

Cat (341) 026

„ (348) -02

,. (3G2) -02

Cat (381) -018

„ (382) -OIG

Average 02 Dauiell.

Q)th lAimhar Posterior Root.
Cat (383) Oil

(Jth Lumbar Anterior Root.
Cat (383) '0045

III. Spinal Cord.

Electrical Difiereiice between Cross Section aiid Surface. Cord Divided and
Isolated at Position Indicated in Table. Contacts 1 centim. apart.

(1.) Cord in Conneetion ivith Brain

Cat

(51) 1st lumbar . . .

■033

(80) 12th dorsal . . .

■044

(97) 12th dorsal . . .

•03

(114) 2nd lumbar . .

■037

(118) 4th lumbar . . .

•04

(124) 3rd lumbar . . .

■046

(120) 2nd lumbar . .

■04

(288) IStli dorsal . . .

■033

(308) ].'t lumbar . . .

■033

(314) 1 2th d<n-snl . . .

■025

(315) 13th dorsal . . .

■032

(317) 13th dorsal . . .

■032

(319) 12th dorsal . . .

■029

(323) 13th dorsal . . .

■031

(324) 13th dorsal . . .

■033

Average '034 Daiiiill.

onkey (41) 12th dorsal . .

. -018

(48) 12th dorsal . .

. -024

(52) 12th dorsal . .

. ^029

(54) 12th dorsal . .

. '027

(234) 12th dorsal .

. 025

Average "025 Dauiell

3 X 2

524

JIRSSKS. F. GOTCIT AXR V. iri)RSLKV

(2.) Cord Severed from Brain.

Cat (37) 8 th dorsal
(G7) 9tli dorsal .
(100) 9tL dorsiil .
(104) 9th dorsal .
(121) 11th dorsal .
(140) 10th dorsal .
(145) 10th dorsal.
(148) 11th dorsal .
(151) 10th dorsal .
(153) lltli dorsal.
(192) 11th dorsal.
(194) 10th dorsal .
(196) 10th dorsal .
(243) 8th dorsal .
(255) 8th dorsal .
(289) 7t)i dorsal .
(327) 9th dorsal .
(337) 6th dorsal .
(339) 8th dorsal .
(344) 8th dorsal .
(346) 8th dorsal .
(349) 10th dorsal .
(355) 8th dorsal .
(375) 8lh dorsal .

Average "029 Daniell

•043

•02G

•03C

■03

■031

•03

•03

■025

■035

■025

■023

■018

■03

•025

•03

■014

•028

•026

•03

•04

•026

•02

•04

■039

Monkey (271 j 8lh dorsal . . . •■)17

(281) 8th dorsal . . . ^027

(321) 7th dorsal. . . •013
Averafje ^019 Daniell.

Cat

(37) 2nd lumbar . . .

•035

(124) 4tli lunil)ar. . .

•033

(170) 1st Innibar . . .

•029

(243) 2nd lumbar . .

■033

(301) 1st lumbar . . .

■037

(357) 2nd lumbar . .

•036

(366J 2nd lumbar . .

•03}

(371) 1st lumbar . . .

•03

(378) 3rd lumbar . . .

•028

Average ■033 Daniell.

Monkey (321) 2nd lumbar . . ■018
(232) 2nd lumbar . . ■OlS

(3.) Cord ill Connection loith Brain, hut divided Longitudinally into Right and

Left Jlrdves.

Cat

(129)
(130)
(133)
(135)
(309)
(318)
(324)

2nd lumbar
1st lumbar
2nd lumbar
3rd lumbar
1st lumbar
13tli dorsal
13th dorsal

Left.

•029

•01

•028

•022

•018

•018

■016

Average '02 and 02 Daniell.

Right.

018
■014
■021
■019
•027
•019
•022

ON THE MAMMALIAN NERVOUS SYSTEM.

525

Monkey (213) .
„ (215) .
„ (217) .

(235) .

(287) .

10th dorsal . . .
10th dorsal . . .
11th dorsal . . .
11th dorsal . . .
2iid lumbar . . .

Left.

Right.

■u2

■023

■013

■013

■013

■014
■012
■Oil
•Oil
•009

Average '01 (5 and

•Oil Dauiell.

(4.) Cord Severed from Brain and Divided LmgitudinaUi/.

Cat (137) . . .
„ (143) . . .

8th dorsal . . .
',1th dorsal . . .

Left.

Right.

•007
■021

■016
■013

(5.) After Section of Columns 1-3 Months Previous to Observation. Cord Severed

from Brain.

Cat (228) division of posterior roots on one side, 11th dorsal '012

(231) division of rig^ht posterior column at 7th dorsal, 11th dorsal . . ^025

(252) division of both posterior columns at 10th dorsal, 6th dorsal . . ^022

(260) left hemiscction at 9th dorsal, 4th dorsal -017

(283) left liemisection at 12th doisal, Cth dorsal •OSS

(228) hemiscction on right side at 9th dorsal, 1st lumbar •OlS

Average ^022 Daniel).

(6.) General Average.

Average, all cases, cord, Cat

,, ,, Monkej-

•032 Daniell.

•022 „

526 ON Till'; MAXIMA!, I AN NERVOUS SYSTEM.

Description ok Pf.ates 2i)-:i5

Plate 29. Arrangemeut of an experiment on a Cat for the investigation of electrical
changes in the spinal corrl. consequent upon excitation of the peripheral nerves.
See Chapter TIL
Plate 30. Photograph of the encephalon and spinal cord of an adult Cat, two-thirds
natural size, Khowing correctly the propoi'tionate size of the various i-egions of
the neural axis.

The spinal nerves in this and the succeeding Plates are indicated by Roman

figures and cajiital letters.
a, h, c, d are points on the cortex cerebri referred to on page 329.
Plate 31. Photograph of the neural axis in a Cat; the same preparation as in

Plate 30, as arranged for the experiments on bilaterality, kc.
Plate 32. Photograph of the encephalon and spinal cord of an adult Rhesus Monkey,
to show, in comparison with Plate 30. the proportionate size and details of the
neural axis in the Monkey as contrasted with the Cat.
Plate 33. Photograph of the neural axis in the Monkey ; the same preparation as in

Plate 32, as arranged for the experimeiits on bilaterality, &c.
Plate 34. Photograph of a recent dissection of the lumbar plexus in an adult
Monkev, showius: the arrano-ement of the roots enterino- the plexus.
Ax. = Anterior crural nerve.
Oh. = Obturator nerve.
E.r. = External popliteal.
r.P. = Internal popliteal.
lis. = Branches to hamstrlnsj muscles.
An. ^= Branches to anal and pudic' plexuses.
"•■ = Filum terminale.
Plate 35. Photograph of a recent dissection of the lumbar plexus in an adult Cat.
Lettering as in Plate 34.

S.-'i. = Small sciatic.

CoTcIi A' Horslcy.

Phil. Trans. 1891. ^.Flate Z2

LJB. Fleming HahwcII

MACACUS RHESUS. ADULT 9

Go tc J I f f- Ho rs 1 1' 1/

Fhil. Trans. 1891. B.PUtt^ 33

P)ioto -fr-nv, L.B -Fleming Hanwell

MACACTTS RHESUQ. ADULT 9

Gotch ^Horslei^.

Fhil. Trans 1801. B.Plate o-/

froTrh (^- florslei/-.

Phil.Tnms 1891.B.P/«fe 35

y

J

Gotch &' Horsle I/.

F?iil. 7'nnis 1891 . ^.PlaU 29

I. B-FlBmlng Hanwell

Go ir I I A'- He i-H /«" 1/

Pill/ 'fnnis 1801. \i. Plate 50

CAT. ADULT 9

L.B Fleming HanweU

Gotch &■ Horslei/.

Phi/. Trans. 1891. BJ>/<ifi- 5/

WP3fcl

G71

Goteh s<,

... Croonian lecture : on the mam-
malian nervous system.

i::>r

' ' ' ^^:

.>K

^^.>i

fc 1?

.-«rn*! ^*V

*> -4

V. ^,

^^VlJ.^^

-/&■■*

^^