Studies of cerebral function in learning

CO

in

O)
O)
CO

o

5

co

Lashley, Karl Spencer

Studies of cerebral
function in learning

Archives of Neurology and Psychiatry

Vol. 12 SEPTEMBER. 1924 No. 3

STUDIES OF CEREBRAL FUNCTION IN LEARNING

I

V. THE RETENTION OF MOTOR HABITS AFTER DESTRUCTION OF
THE SO-CALLED MOTOR AREAS IN PRIMATES *

K. S. LASHLEY. Ph.D.

MINNEAPOLIS

Since the area was first described by Fritsch and Hitzig,1 the
function of the electrostimulable cortex of the cerebrum has been the
subject of almost continuous controversy. The experiments were
immediately called in question through criticisms of the technic by
Dupuy,2 Sanderson,' Carville and Duret,4 and others, or by abstruse
metaphysical deductions such as were advanced by Hermann 5 who
objected to the motor area as violating the "unity of mind." The work
of Ferrier,6 Carville and Duret and Hitzig soon established the fact
of the electrical excitability of limited areas of the cortex, but immedi-
ately a new question arose. Fritsch and Hitzig had considered the
excitable zone as motor, if we may translate the expression, "entry of
single psychic functions into material" by such a term. In this they
were followed by Carville and Duret, who described the motor dis-
turbances following lesions in the area as "paralysie de la motricite
volontaire corticale." Ferrier also considered the area as motor. But
most of the early work had been done with dogs, and the sensory
disturbances which appear in this animal after ablation of the stim-
ulable area were emphasized by Schifr,7 Nothnagel,8 Munk 9 and

* From the Department of Psychology of the University of Minnesota.

1. Fritsch, G., and Hitzig, E. : Ueber die elektrische Erregbarkeit des
Grosshirns, Arch. f. Anat. u. Physiol. 1870, pp. 300-332.

2 Dupuy, E. : Experiment sur les fonctions moteur du cervean, Compt.
rend. soc. de biol., 1888, pp. 1025-1027.

3. Sanderson, J. B.: Note on the Excitation of the Surface of the Cerebral
Hemispheres by Induced Currents, Proc. Roy. Soc. 22:368-370, 1874.

4. Carville, C, and Duret, H. : Sur le fonctions des hemispheres cerebraux,
Arch. d. Physiol. 7:352-490, 1875.

5. Hermann, L. : Ueber elektrische Reizversuche an der Grosshirnrinde,
Arch. f. d. ges. Physiol. 10:77-88, 1875.

6. Ferrier. D. : The Functions of the Brain, London, 1876.

7. Schiff, M. : Untersuchungen iiber die motorischen Fnnctionen des Gross-
hirns, Arch. f. exper. Path. u. Pharmocol. 3:171-179, 1875.

8. Nothnagel, H. : Experimentelle Untersuchungen iiber die Fnnctionen des
Gehirns. Arch. f. path. Anat. u. Physiol. 57:184-227, 1873.

9. Munk, H.: Ueber die Funktionen der Grosshirnrinde, Berlin, 1890.

3 I

ARCHIVES OF SEMIOLOGY AND PSYCHIATRY

o
>-

llitzig.1" who interpreted the disorders of movement variously as due
to the loss of muscular and cutaneous sensitivity or to the loss of
kinesthetic images of the movements to be performed.

Following the suggestion of Tamburini, Luciani and Seppili 1X
advanced the view that the motor areas contained somesthetic projec-
tion fibers as well as motor elements, and developed a theory of the
sensorimotor function of the sigmoid gyrus and rolandic areas. In
this they were followed by Horsley,12 Mott,13 Dana.14 Rothmann 15 and
many others.

De Barenne,10 in particular, has demonstrated the existence of
marked sensory disturbances in the cat after application of strychnin
to an area which widely overlaps the stimulable area, and the general
correctness of Luciani's view for the stimulable areas of lower mam-
mals does not seem open to question, although there may be some doubt
as to whether the motor functions of the cortex in these forms are
comparable with those of primates.

Recent more critical work with primates, however, shows that in
them a further specialization has occurred with the development of the
fissure of Rolando as a line of demarcation between centripetal and
centrifugal projection areas.17 The work of Schafer,is Mills, 1(l Griin-

10. Hitzig, E. : Physiologische unci klinische Untersuchungen iiber das
Gehirn, Berlin, 1904.

11. Luciani, L., and Seppili, G. : Die Funktion-Localization auf der Gross-
hirnrinde, Deutsche Ausgabe, Leipzig, 1886.

12. Horsley, V. : On the Analysis of Voluntary Movement, 19th Century
29:857-870, 1891.

13. Mott, F. W. : The Sensory-Motor Functions of the Central Convolutions
of the Cerebral Cortex, J. Physiol. 15:464-487, 1893-1894.

14. Dana, C. L. : A Study of the Functions of the Cortex of the Motor Area
of the Brain, J. Nerv. & Ment. Dis. 21:761-785, 1894.

15. Rothmann, M. : Leber die elektrische Erregbarkeit der Zentral-
windungen, Monatschr. f. Psychiat. u. Neurol. 32:489-502, 1912.

16. Barenne, J. G. D. de: Sensory Localization in the Cerebral Cortex, Quart.
J. Exper. Physiol. 9:355-390, 1916.

17. The lack of any persistent paralysis in rodents, carnivora and ungulates
after destruction of the stimulable areas and the occurrence of partial paralyses
after destruction of parts of the corpus striatum in lower forms suggest that
the differentiation of function of precentral and postcentral gyri in primates
may be due rather to the acquisition of primative striate functions by the
cerebral cortex than to a division of cortical functions which overlap in lower
forms.

18. Schafer, E. A.: On the Alleged Sensory Functions of the Motor Cortex
Cerebri, J. Physiol. 23:310-314, 1898.

19. Mills, C. K. : The Separate Localization in the Cortex and Sub-Cortex of
the Cerebrum of the Representation of Movements and of Muscular and Cutane-
ous Sensibility, J. Nerv. & Ment. Dis. 38:595-619, 1901.

LASHLEY— CEREBRAL FUNCTION 251

baum and Sherrington,2" C. and O. Vogt,21 Lewandowsky and Sim-
mons,22 Hoppe,23 Cushing,24 Franz,23 and Leyton and Sherrington,26
all points to the conclusion that primary excitability is limited to the
precentral gyrus and that sensory disturbances rarely result from lesions
confined to this area.

These studies have defined the centrifugal function of the cells of the
precentral gyrus and have confirmed the view that the area is motor
rather than somesthetic. but in spite of the fairly general recent agree-
ment as to the location of "motor" area, there is a wide diversity of
opinion concerning the significance of the "motor" function. The area
is somehow concerned with the production of movement, but how it
acts and what kinds of movement it controls, are still debated points.
Ferrier 6 emphasized the "automatic" character of movements after
destruction of the motor areas and held that "All reactions not so
(automatically) organized, and still dependent upon conscious discrimi-
nation and exercise of attentive volition are effectually and permanently
annihilated." Munk 27 classified movements according to their somatic
distribution and maintained that small unilateral adaptive movements
( Einzelbewegungen ) are permanently lost after destruction of the motor
areas. His definition of these movements is by no means clear, but
his accounts seem to imply that the movements had been learned, and
he denies that a dog can learn to give his paw after destruction of both
motor areas. Wagner 28 maintained that the chief function of the motor
areas is in learning, and that animals deprived of them are incapable of
forming any new habits. Bechterew 29 also refers to experiments lead-

20. Grunbaum, A., and Sherrington, C. S. : Observations on the Physiology
of the Cerebral Cortex of the Anthropoid Apes, Proc. Roy. Soc. 72 B: 152-155,
1903.

21. Vogt, C, and Vogt, O. : Zur Kenntnis der elektrisch erregbaren Hirn-
rindengebiete bei den Saugetieren, J. f. Psychol, u. Neurol. 8:277-456, 1907.

22. Lewandowsky, M., and Simmons, A. : Zur Physiologie der vordern und
hintern Zentralwindung, Arch. f. d. ges. Physiol. 129:240-254, 1909.

23. Hoppe, H. H. : A Critical Study of the Sensory Functions of the Motor
Zone (Pre-Rolandic Area) : More Especially Stereognosis, J. Nerv. & Ment.
Dis. 36:513-527, 1909.

24. Cushing, H. : A Note Upon the Faradic Stimulation of the Postcentral
Gyrus in Conscious Patients, Brain 32:44-54, 1909.

25. Franz, S. I. : Variations in Distribution of the Motor Centers, Psychol.
Monogr. 19:80-162, 1915.

26. Leyton, A. S. F., and Sherrington, C. S. : Observations on the Excitable
Cortex of the Chimpanzee, Orang-Outan and Gorilla, Quart. J. Exper. Physiol.
11:135-222, 1917.

27. Munk, H. : Ueber die Fuhlspharen der Grosshirnrinde, Sitzungsber. d.
Berlin Akad. Wiss., 1892, 679-723; 1893, 759-781; 1894, 823-833; 1896, 1131-1159.

28. Wagner, V: Discussion in Neurol. Zentralbl. 24:1022, 1905.

29. Bechterew, W. v.: Die Funktionen der Nervencentra, Jena 3. 1911.

2?2 ARCHIVES OF NEUROLOGY AND PSYCHIATRY

ing to the same conclusion. He states that after removal of the motor
centers associative motor reflexes are lost. "With unilateral destruction
of the motor region in dogs the associative motor reflex may he elicited
in the homolateral fore-leg, but in the contralateral fore-leg the reflexes
learned earlier are forever lost and can not be reestablished even after
a number of conditioning associations (p. 1551ff.)." Gierlich 30 also
supports this view of the exclusive motor function of the stimulable
areas.

In opposition to these results, several writers have reported the
acquisition of habits after the destruction of the motor areas or of the
pyramidal tracts. Starlinger 31 trained a dog to give his paw after
total destruction of both pyramidal tracts. Rothmann 32 observed learn-
ing in a rhesus monkey in which one precentral gyrus had been extir-
pated and the pyramidal tract of the other had been sectioned in the
cervical region. Franz and Lashley 33 and Lashley 34 found learning
ability in the rat unaltered by total destruction of the stimulable cortex.
This re>ult has been confirmed by Jellinek and Koppanyi.33

In the contradiction of evidence here, we must favor the positive
results. Failure to learn may be due to any one of a number of factors
in addition to specific destruction of tissue, and a single positive case
with certain destruction of the motor area is sufficient to discredit any
number of negative findings such as are cited by Bechterew and Munk.
It seems quite certain that the formation of conditioned motor reflexes
is possible in the absence of the electrostimulable cortex, but this fact
fails to reveal the normal function of the area in the performance of
complex activities. Both Rothmann and Brown38 seem to believe that
in the intact animal the motor areas form the chief centrifugal path for
complex adaptive reactions and that when learning occurs in their
absence it is to be considered as due to vicarious function of other

30. Gierlich. X.: Ueber Symptomatologie, YYesen, unci Therapie der hemi-
plegischen Lahmung, Wiesbaden, 1913.

31. Starlinger, J.: Die durchschneidung beider Pyramiden beim Hunde,
Neurol. Zentralbl. 14:390-394. 1895.

32. Rothmann, M. : Ueber die physiologische Wertung der cortico-spinalen
(Pyramiden) Balm, Arch. f. Anat. u. Physiol. (Physiol. Abt.) pp. 217-275, 1907.

33. Franz, S. I., and Lashley, K. S. : The Retention of Habits by the Rat
After Destruction of the Frontal Portion of the Cerebrum, Psychobiol 1:3-18
1917.

34. Lashley, K. S. : Studies of Cerebral Function in Learning, Psychobiol.
2:55-135, 1920.

35. Jellinek. A., and Koppanyi, T. : Lernfahigkeit gehirnverletzter Ratten.
Anzeiger d. Akad. d. YViss., Wien, 1923, No. 17.

36. Brown, T. (i.: Studies. XXVII. (>. The Motor Activation of Parts of
the Cerebral Cortex Other Than Those Included in the So-Called "Motor"
Areas in Monkeys, Quart. J. F.xper. Physiol. 10:103-143, 1916.

LASHLEY— CEREBRAL FUNCTION 253

parts rather than as an expression of their normal function. On the
basis of Brown's work and of clinical evidence, however, Monakow 36a
is inclined to minimize the importance of the pyramidal areas for
••voluntary" movement. He suggests that "We must give up or essen-
tiallv modify the view that the precentral convolution alone conducts
impulses to voluntary movement. It is probable that the pyramidal
areas serve less for the execution of voluntary movements than for the
inhibition of the kinetic functions of spinal coordination. Their func-
tion seems to be in the class of reflex activity." Lashley :JT has
reported the survival of visuomotor habits after the complete destruction
of the stimulable area and serious injury to the caudate and lenticular
nuclei 3S and has suggested that the primary function of the stimulable
area is the reflex regulation of postural and kinetic mechanisms. Similar
results have since been obtained for the stimulable cortex with problem-
box and maze habits.

Thus we find in the literature claims that the electrostimulable cortex
is motor, that it is sensory, that it is sensorimotor, that its motor func-
tion is exercised through the storing of images of movement, that it
is the final common path for all voluntary movements, and that it is a
reflex center not primarily concerned with voluntary activity.

Much of the literature on the function of the electrostimulable
cortex, as on cerebral localization in general, presents an inextricable
tangle of physiologic fact and psychologic speculation. The long con-
troversy between Goltz, Munk, and Hitzig was largely due to their
inability to grasp each other's psychologic theories, and recent progress
in psychology tends to invalidate much of the cerebral localization which
was based on older conceptions of mental faculties.

Images have fallen into disrepute, and even the psychologists who
still deal with them deny that kinesthetic imagery has any demonstrable
relation to the initiation of movement (Thorndike 39). The conception
of volitional activity is too vague to have any scientific value. At best
it represents an indefinite distinction between more or less complexly
conditioned activities, and the conception of conditioned reflexes leaves
the distinction without significance. The "will" has been largely dis-
carded in psychology, although it is still in good standing in neurologic
discussions. Such a statement as that the stimulable cortex is not

36a. Monakow : See footnote 57.

37. Lashley, K. S. : Studies of Cerebral Function in Learning. The Motor
Areas, Brain 44:255-286, 1921.

38. When these experiments were reported, I believed that the lesions were
largely confined to the caudate nuclei. Dr. J. B. Jonhston has since called my
attention to the fact that the caudate nucleus in the rat comprises only a
narrow median band in the corpus striatum and that the lesions reported actually
included a considerable portion of the lenticular nucleus.

39. Thorndike. E. L. : The Mental Antecedents of Voluntary Movement, J
Philos., Psychol, and Sc. Meth. 4:40-42, 1907.

254 ARCHIVES OF NEUROLOGY AXD PSYCHIATRY

motor, but "psychomotor," means no more than that the pyramidal
cells excite patterns of spinal motor cells rather than individual cells.
It contributes nothing to our understanding of cerebral function. In
the present state of psychologic science, we can not do better than fol-
low the dictum of Bubnoff and Heidenhain,40 "Es will uns uberhaupt
scheinen als miiste die Untersuchung der physiologischen Processe in
dem Gehirn von den jene Vorgange begleitenden Bewusstseinsvorgangen
moglichst absehen, wenn es sich urn eine Deutung physischen Gesche-
hens handelt," and rigidly exclude from neurologic discussion every
subjective concept which cannot be translated into objective terms.

There is no evidence for the localization of any "mental function"
in any part of the cerebrum. All that can be concluded from the
existing evidence is that the conducting pathways concerned in par-
ticular kinds of behavior lead from receptor to effector through certain
cerebral areas. Cerebral motor localization is a problem of the origin .
and function of the centrifugal neural impulses of the cortex. Their
"volitional" or "automatic"' character can be defined only in terms of
their complexity of organization and their relative importance in the
total motor integration or kinetic melody, and until so defined the terms
are meaningless. The "reflex" conception of cerebral function, although
still a theory and notably inadequate to account for all the phenomena
of cerebral function because of oversimplification in its formulations,41
is too well supported by evidence on nerve conduction and analogy with
spinal functions to be disregarded in favor of any speculations con-
cerning the localization of "psychic" functions.

STATEMENT OF PROBLEM

Stated objectively, three mutually incompatible theories concerning
the function of the precentral gyrus are to be found in the current
literature. They are: 1. This area is the only centrifugal outlet from
the cerebral cortex for complexly integrated movements or for move-
ments acquired as a result of training ( voluntary movements or con-
ditioned reflexes). 2. In the intact animal, the Betz cells are the
principle centrifugal paths, but some neural impulses of like function
may descend by extrapyramidal tracts, and these tracts may assume
vicariously all the functions of the motor area. 3. The motor area
is a part of the mechanism functioning in the regulation of tonus and
posture and is not directly concerned in conditioned reflex activity.

40. Bubnoff, V. and Heidenhain, R. : Ueber Erregungs- und Hemmungs-
vorgange innerhalb der motorischen Hirncentren, Arch. f. d. ges. Physiol. 26:
137-200, 1881.

41. The data on direct adaptation of unpracticed organs to the solution of
|irol)lem-boxes presented later in this paper seem wholly inexplicable in terms
oi simple conditioned reflexes.

LASHLEY— CEREBRAL FUNCTION 255

The first of these views is definitely ruled out by the evidence cited
above for habit formation after destruction of the areas. The evidence
for the second and third views, as applied to primates, is inconclusive,
although a considerable mass of evidence derived from the cerebral
paralyses seems to support the second. The validity of this evidence
will be considered after the experimental data are presented.

A simple test of the hypotheses is possible. If after total destruc-
tion of the precentral areas an animal shows undiminished ability to
carry out activities of all degrees of complexity, acquired before the
operative destruction, this will be conclusive evidence that the lesion
did not destroy any part of the conditioned reflex arcs involved in the
activities, and, as a corollary, that these arcs do not traverse the motor
areas. Loss of the conditioned reflexes following the lesion with their
later reestablishment through training will support the second view, that
the pyramidal cells of the precentral region are the principal efferent
paths involved in habitual movements.

Such a test was made with the rat and gave unmistakable evidence
against the participation of the motor areas in the activities of the
simple maze and visual discrimination box ( Franz and Lashley, ;
Lashley37). Since the publication of that report, the results have been
confirmed for a complex maze and the "double-platform box." As the
rat shows no paralysis after lesions to the stimulable areas alone, it is
not possible to generalize from it to higher forms which do develop a
paralysis. I have therefore repeated the tests in a series of experiments
with monkeys.

EXPERIMENTAL METHODS

The parlysis which follows lesions to the precentral gyrus in monkeys
necessitates a modification of the technic used with the rat. The general
procedure was as follows : The animals were trained in rather simple
manipulative acts involving a new and easily recognizable pattern and
sequence of movements. They were then kept without practice for
about two months, at the end of which time their retention of the habits
was tested. This gave a measure of the normal loss to be expected
from disuse of the habits over a period equal to that required for
recovery from cerebral paralysis. After these preliminary retention
tests, the motor areas were destroyed. The animals were then kept
without further practice in the habits until the paralysis was so far
improved that they were judged capable of making the movements
required. They were then given a final series of retention tests (post-
operative retention tests) and brought to necropsy.

Train ing Methods. — For training, the familiar problem-box method
was used. The animals were confined in a large cage. 5 by 5 feet, to

256 ARCHIVES OF NEUROLOGY AND PSYCHIATRY

the floor of which a small problem-box containing food was bolted.
Three problem-boxes were used :

1. Pull box. The animal must reach through a circular hole. 2
inches in diameter, grasp and pull forward a rod which passed trans-
versely 3 inches behind the hole. This released the lid of the box,
which was thrown open by a spring.

2. Crank box. The animal must grasp and turn a crank projecting
from the front of the box. The crank handle described a 6 inch circle.
It offered a resistance of about 200 gm. at all points of the circle. It
was set at "one o'clock'' and must be turned counter-clockwise through
270 degrees, at which point it released the lid of the box. The lid was
thrown back by a spring.

3. Hasp box. The animal must open an ordinary gate hasp, closed
with a wooden plug inserted loosely through the staple, withdrawing
the plug and lifting the hasp from the staple, over which it would fall
again if released. He must then raise the lid of the box and hold it
open while he reached in for the food.

Five trials a day were given with each box ; the time spent in each
trial was recorded, and detailed notes were made as to the use of right
or left hand and the exact method employed in opening the boxes.
Training was continued until the latches were released in a stereotyped
manner without random movements. Several interruptions of training
occurred lasting from one to several weeks, so that the learning curves
do not represent the rate for continuous training.

In addition to the problem-box habits, each animal was trained to
pick out cubes of banana from among cubes of wood of similar size
and appearance. The cubes were scattered in irregular order under a
sheet of glass supported 2 inches above the floor of the cage. They
were placed about 6 inches back from the edge so that the animals had
to reach under the glass to get the cubes, which they could see but
could not distinguish by odor.

Operative Technic. — Destruction of the motor areas was made under
ether anesthesia, with aseptic precautions. The region of the precentral
gyrus was exposed by trephining and identified by electrical stimulation.
The opening was enlarged by bone-forceps until the precentral gyrus
and surrounding areas were exposed. Arm, leg and face areas were
verified by stimulation, and the limits of the excitable area determined.
The entire area was then undercut by thermocautery to a depth of
about <> mm. To avoid injury to the longitudinal sinus, a median bridge
oi bone, 1 cm. in width, was left intact. In one specimen the leg area of
this region was undercut by passing the cautery diagonally mediad and
downward until the resistance of the falx was felt, then cutting longi-
tudinally across the gyrus. The dural flaps were then replaced and
the wound closed. In the cases reported below the wounds healed with-
out infection.

LASHLEY— CEREBRAL FUNCTION 257

Retention Tests. — The animals all showed marked paralysis after
operation. This cleared up gradually, and the retention tests were
given about two months after the operation. (Numbers 1 and 3, v.i.,
recovered somewhat more quickly than is usually the case when the
lesion is restricted to one hemisphere.) In the postoperative retention
tests, the animals were placed singly in the large cage with each of the
latch boxes in turn. The time required to open the boxes was noted,
and the methods were recorded in detail for comparison with methods
employed in learning and in the preliminary retention tests.

Verification of Lesions. — When the tests were completed, the opera-
tive fields were again exposed and explored by electrical stimulation.
Excitable points found were mapped. The brains were then removed,
fixed in 10 per cent, formaldehyd. and sketches made under a camera
lucida. Serial sections of the region of the lesions were then prepared.
Camera drawings of these were made and the lesions reconstructed
from them.

PROTOCOLS

Number 1. — This was a small male cebus, trained on the crank box, pull box
and hasp box. The skull was trephined and opened on both sides in front of
the precentral gyrus. The openings were extended backward to the fissure of
Rolando. The leg, arm and face areas were located by electrical stimulation
and destroyed by cautery. Cauterization extended beyond the stimulable area
except in the median line. The wound was covered with mica and closed.

On the following day, there was a partial paralysis of both sides with great
spasticity. Coordinated walking movements were possible, but there was great
weakness of the legs. The arms were extended toward food. He could not
grasp with his left hand. Partial grasping with the right hand appeared, but
there was inability to raise food to the mouth. He recognized a banana, and
made efforts to grasp it. The arms were usually hyperextended.

Four days later, he moved clumsily, his arms and legs spread out frequently,
letting him fall prone. He grasped with his right hand, but was unable to hold
food or lift it to his mouth. He ate by thrusting his mouth against the bread.
He was well oriented in the room. Two weeks before operation he had learned
to slip out of the crack as the door was opened and to run into an adjoining
room. He did this twice on the fourth day. He had a tendency to stay near a
cage containing other monkeys and to hide under it when pursued. There was
marked tremor after slight effort.

Ten days after operation, he stood and walked without falling, fumbled in
grasping, but was able to hold food in the right hand or to lift it to his mouth.

Thirty-five days after operation, he climbed and ran accurately, picked up
small pieces of banana with the right hand without noticeable clumsiness, and
made quick movements in efforts to catch flies. Retention of the problem-box
habits was tested at this time.

The average time per trial in each day's practice (five trials daily) is given
in Table 1 for each of the problem-boxes. This is followed by the average time
per trial on each day of the preliminary retention tests, and similarly for the
postoperative retention tests.

258

ARCHIVES OF XEUROLOGY AND PSYCHIATRY

Table 1.— Average Time in Seconds Per Trial Consumed in Opening Each

Problem-Box in Each Day's Practice of Training, Preliminary Retention

Tests and Postoperative Retention Tests

Training Tests

Prelimi

nary Retention

. A

Tests

l'n II Box

< rank Box

Hasp Box

Pull Box

Crank Box

Hasp Box

•206.2

135.2

1.6

3.2

44.4

."..4

390.0

34.4

3.0

1.6

81.8

1.8

47.2

10.0

2.4

11.0

14.0

•:.o

12.8

18.0

1.2

2.4

1.6

4.6

5.6

1.2

3.S

2.6

2.4

10.8

2.7

2.0

27.6

.i ..

2.6

1.6

4.S
6.0
1.8

Postoperative Retention Tests

Pull Box

Crank Box

Hasp Box

2.8

2.8

63.6

31.0

1.6

2.2

7.4

57.6

2.2

4.4

■2M

3.8

Tests for visual discrimination were made. There was no error in 100
trials. At all times following the operation the animal was oriented in the
cage and room and. with the exception of the motor disturbance, gave no
indication of anv deterioration.

Fig. 1. — Extent of lesions in animal Xo. 1.
sketch and serial sections.

Reconstructed from camera

Extent of Lesions. — The areas destroyed are shown diagrammatically in
Figure 1 and in sections in Plate 1.

Left Hemisphere: Mediad, the lesion began about 1 mm. behind the end of
the fissure of Rolando and extended forward to the level of the knee of the
corpus callosum. The cortex of the precentral gyrus was destroyed to within
2 mm. of the edge of the longitudinal fissure, but that of the median surface
was uninjured. Caudad, the lesion extended slightly onto the postcentral gyrus
but did not involve all of the cortex within the fissure of Rolando. Laterad, it
extended to the upper border of the operculum. The parts of the stimulable
area left intact were the paracentral gyrus, the cortex within the fissure of
Rolando and the lateral part of the face area included on the operculum.

LASH LEY— CEREBRAL FUNCTION

259

Plate 1. — Fig. A: Outline of the dorsal aspect of the cerebrum. The broken
areas outlined indicate the extent of dural adhesions. The transverse broken
lines show the level of the sections designated by the corresponding numbers.

Figs. 1 to 5. — Camera sketches of sections showing extent of lesions. Blood
clots and scar tissue are indicated in solid black. Obviously degenerated cortex
is marked with coarse stippling. R, fissure of Rolando; S, fissure of Sylvius;
P, parieto-occipital fissure.

260 ARCHIVES OF NEUROLOGY AXD PSYCHIATRY

Right Hemisphere: The lesion was similar to that on the left but slightly
more extensive. The cortex within the rolandic fissure was destroyed, and the
lesion extended farther over the operculum.

After destruction of almost all of the arm areas of both sides and
of most of the areas of the legs and face, this animal, on recovery from
paralysis, showed perfect retention of visual and motor habits acquired
before injury. Except for the paralysis and later spasticity, no sig-
nificant change in his behavior could be noted.

Number 2. — This was a small male cebus too wild for training at the begin-
ning of the experiments. The motor area of the right hemisphere was exposed,
the arm, face and leg areas identified and cauterized to a depth of 5 mm.,
caudad to central fissure, laterad and cephalad to a line 5 mm. beyond the limits
of the excitable area.

Following operation, the left leg and arm were not used. The leg was
hyperextended and gave some support to the body in standing or sitting, but
made no stepping movements. No movements of the hand could be elicited.
This complete paralysis of the left hand persisted for two weeks.

Four weeks after operation, the left leg was used almost normally. The left
arm could be used to support the animal's weight, but tended to become rigid in
hyerpextension, and the left hand could not be used for grasping.

Eleven weeks after operation, the paralysis had almost disappeared; the
left hand was somewhat clumsy but could be used in grasping food. It was
not used when the right hand was unrestrained.

Training on the problem boxes was begun at this stage of recovery. During
the next three months, the problems were learned and retention tests given.

Seven months after the first operation, the left motor area was exposed and
similarly explored and destroyed. Paralysis of the right arm and leg followed. It
appeared to be as complete as that of the left side following the first lesion. It
improved more rapidly, however. Six weeks after operation the right hand was
used to pick up food, and, although still somewhat spastic, was judged capable
of manipulating the latch boxes. Retention tests were therefore begun. The
average time per trial for successive groups of five trials in training, pre-
liminary retention tests, and postoperative retention tests are given in Table 2.
Visual discrimination was unaffected by the operation.

Table 2. — Average Time in Seconds Per Trial Consumed in Opening Eaeh

Problem-Box in Each Day's Practice in Training, Preliminary Retention

Tests and Postoperative Retention Tests

Tra

ining Tests

Preliminary Retention

Tests

Pull Box

Crank Box

Hasp Box

Pull Box Crank Box

Hasp Box

24 brs.«

126.0

102.2+

2.2 28.8

8.8

24 hrs.*

105.0

lll.Ot

2.0 1.6

6.4

2,520 sec.

16.0

235.0

24 hrs.*

5.2

184.4

Postoperative Retention Tests

366 sec.
2.5

7.S
7.4

19.6
16.0

A_

Pull Box Crank Box

Hasp Box

1.4

1.4

55.6

9.8 14.8

102.5

4.5

1.2

6.0

3.6 11.4

93.4

20.4

2.8

6.6

1.2 7.2

35.4

1.6

60.6
22.8
15.4

* Failed to open box while under observation and was left in the cage over night.
+ Time with hasp left unfastened.

LASH LEY— CEREBRAL FUNCTION 261

The short time required for opening the boxes in the postoperative
retention tests gives clear evidence for some retention of the habits.
Much of the delay apparent was due to the weakness and clumsiness
of the right hand. With each problem-box the attack in the postopera-
tive retention tests was directly on the latches. The methods employed
were at first those used before operation, and the movements were
definitely adapted to solving the problems, although lacking force and
accuracy; for example, efforts were all directed to turning the crank
counter-clockwise.

The weakness of the right arm led to a surprising readjustment on
the part of this animal. The operation on the right hemisphere made
the left arm weak and spastic during training. In all of the trials of
training and preliminary retention tests the left arm was used only as
a pro]), and the left hand was not once used in manipulating the latches
of any of the boxes. After the second operation, the right hand was
much more affected than the left, which had largely recovered, and
an almost immediate shift to the left hand in opening all of the boxes
occurred.

Pull Box : Postoperative retention tests. Trial 1 : Fumbled in hole with
right hand, removed hand, peered into hole; again inserted right hand and again
pulled lever ; twenty-seven seconds.

Trial 2 : Right hand, fairly accurately, seven seconds.

Trial 3: He fumbled with right, then inserted left, grasped lever and pulled:
seven seconds.

Trial 4 : He inserted left hand at once ; three seconds.

Trial 5: Left hand used at once; five seconds.

The right hand was used only four times in the succeeding forty trials.

Crank Box: Trial 1: He grasped the crank with his right hand at once
and turned counter-clockwise. The crank stuck in the third quadrant. He
pushed at it feebly, gave up, returned to the attack from the side of the box
and pulled it through final segment ; 160 seconds.

Trial 2: He grasped the crank with the right hand. Apparently, he was
unable to move it. He grasped it with both hands and swung it around ;
twenty-five seconds.

Trials 3, 4 and 5: He used only the right hand and turned with difficulty
moving to side of box and exerting direct pull instead of his former transverse-
rotary movement.

Trial 6: He grasped with the left hand and turned counter-clockwise;
six seconds. All later trials were made with the left hand only.

Hasp Box: On the first three days of the postoperative tests, he pulled
out the plug and disengaged the hasp promptly with his right hand, but lacked
strength to lift the lid. He gave up after a few attempts with his right hand.

On the fourth day, he drew out the plug and disengaged the hasp with his
right hand, then lifted the lid with his left foot. It fell back as he attempted
to reach the food. He lifted it again with his left hand, climbed to the edge
of the box, bringing his right side against the lid, so holding it up while he
grasped the food with his left hand. On the second trial, he lifted with the
left hand and held it up with his left hand, inserting his head for the food. In

262 ARCHIVES OF NEUROLOGY AXD PSYCHIATRY

all later trials, he lifted the lid with his left hand, sometimes holding it back
with his left knee or with his head while reaching into the box with his left
hand, or with the left hand while reaching with his head. On the eighth and
all later trials, he pulled the plug with his left hand and used the right only as
a prop.

Throughout the retention tests his activities were centered on the
plug, hasp, and lid. He never attempted to lift the lid until the hasp
was disengaged. When the lid was raised the next acts seemed defi-
nitely directed to holding it up, and though clumsy, the movements were
clearly not random.

EFFECTS OF SUBSEQUENT DESTRUCTION OF THE CORPUS STRIATUM

A hroad bladed cautery was next passed through the old lesion into
the corpus striatum and drawn back and forth through this nucleus.
The wound was closed, and the animal was kept under observation
until his death eight days later.

On recovery from anesthesia, the animal showed marked spasticity
of the left side. The left arm was usually hyperextended, although in
walking or clinging to a perch the arm and leg assumed a normal
posture. The left side was very weak, and when he walked the arm
and leg frequently collapsed suddenly. He took food with his right
hand and placed it in his mouth. When I held a bit of food, he drew
my fingers to his mouth with the palm of his left hand, but without
closing the fingers, which remained hyperextended.

On subsequent days, he used both right and left hands in walking,
climbing and grasping food. The left side was spastic and very weak
but capable of a variety of fine adaptive movements. The right side
-bowed a coarse tremor, and athetoid movements of the left arm
appeared when the right was used. There was marked paralysis of the
pharynx. He kept his mouth stuffed with food or shavings but was
unable to swallow.

At no time was the paralysis as marked as after the destruction of
die cortex. Indeed the condition showed no resemblance to hemiplegia,
but, except for the pharyngeal paralysis, was essentially that described
by Wilson 4L' for lesions of the striate nucleus without involvement of
the pyramidal tracts. The possibility that the recovery from the initial
paralysis was due to vicarious functioning of the striate nucleus seems
thus definitely to be ruled out. The animal did not recover sufficiently
for retention tests after this operation, but his behavior when released
in the laboratory -bowed that his general orientation was unaffected.

42. Wilson, S. A. K. : An Experimental Research Into the Anatomy and
Physiology of the Corpus Striatum. Brain 36:427-492, 1913.

LASH LEV— CEREBRAL FUNCTION 263

When given an egg, he made efforts to break it by pounding it on the
floor, as he had done before the operation, and in the performance of
this habit both hands were used.

Extent of Lesions. — The extent of the destructions is indicated in Figure 2
and sections through the area are shown in Plate 2.

Right Hemisphere : The lesion extended cephalad from the median end of the
central fissure to the middle of the superior frontal gyrus, bordering the
longitudinal fissure but leaving the cortex of the median surface of the hemi-
sphere intact. Caudad it invaded the postcentral gyrus and completely oblit-
erated the fissure of Rolando. Laterad it extended well onto the operculum.
Only the paracentral gyrus and the lateral portion of the face area remained
intact.

The second operation destroyed all of the caudate nucleus and the greater
part of the lenticular, leaving only the posterior end of the puramen intact.

Fig. 2. — The extent of the lesions in animal No. 2. Reconstructed from
camera sketch and serial sections. The posterior border of the left precentral
gyrus escaped injury.

Left Hemisphere: The lesion was less extensive than that on the right. It
began 5 mm. in front of the median end of the central fissure and extended to
the frontal lobe. All of the cortex of the median surface to the callosomarginal
fissure was destroyed. The posterior edge of the precentral gyrus remained
intact, for a width of about 5 mm.

In this animal, practically all of the precentral gyrus of the right
hemisphere was destroyed. He was then trained in manipulative move-
ments of the right hand. This was followed by partial destruction of
the left precentral gyrus. ( )n recovery from paralysis, he gave clear
evidence of retention of the habits but owing to spasticity of the right
hand, made a direct transfer of the habits to the left hand. The right
precentral gyrus was almost completely destroyed, whereas a rather
large proportion of the left precentral gyrus escaped injury, but in spite

264 ARCH IlllS OF NEUROLOGY AND PSYCHIATRY

Plate 2. — Fig. A: Outline of the dorsal aspect of the cerehrum. The broken
areas outlined indicate the extent of dural adhesions. The transverse broken
lines show the level of the sections designated by the corresponding numbers.

Figs. 1 to 5. — Camera sketches of sections showing extent of lesions. Blood
clots and scar tissue are indicated in solid black. Obviously degenerated cortex
is marked with coarse stippling. R. fissure of Rolando: S, fissure of Sylvius:
/', parieto -occipital fissure.

LAS HLEY— CEREBRAL FUNCTION 265

of this, the animal shifted to the use of his left hand in opening the
problem-boxes. Subsequent destruction of the greater part of the
corpus striatum did not produce a recurrence of the hemiplegic
symptom s.

Number 3. — This was a large female rhesus trained on problem-boxes and
visual discrimination. After retention tests, the motor areas of both sides were
exposed, mapped and cauterized. Following the operation the legs and left
arm were completely paralyzed. The right arm made clumsy pawing move-
ments. Twelve hours after operation, the animal walked a few feet with stag-
gering gait, then collapsed with arms and legs widely extended, and for several
days made no further efforts to walk. The following day she grasped a grape
with her right hand and brought it to her mouth after several unsuccessful
trials. The movements were clumsy and slow.

Four weeks after operation she seemed sufficiently recovered for retention
tests, although still showing a general clumsiness and marked weakness of the
left limbs.

The average time per trial for successive groups of five trials in training,
preliminary retention tests and retention tests after operation is given in Table 3.
Visual discrimination was unaffected by the operation.

Table 3. — Average Time in Seconds Per Trial Required for Opening Each

Problem-Box in Each Day's Practice in Training, Preliminary Retention

Tests, and Postoperative Retention Tests

Training T* >t-

Preli:

minary

Retention

Tests

Pull Box

Crank Box

H

asp Box

Pull Box

Crank Box

Hasp Box

68.2

1,595.0

40.4-

1.0

4.2

24.0

2. 6

111.8

373.2

1.0

3.8

6.6

1.0

145.4

124.<i

1.6

5.0

1.0

304.2

24.2

1.2

3.6

140.0

17.0

1.0

2.2

16.S

2.8

Po>

'toperative

Retention Tests

Pull Box Crank Box Hasp Box
1.0 2.2 49.2

1.0 1.4 9.8

2.6 1.4 8.4

' Time with hasp unfastened.

The time required to open the problem-boxes in the postoperative
retention tests gives certain evidence of the retention of the habits.
An average of 678.8 seconds was consumed in each of the first five
trials of training in opening the boxes by the method of random activity.
( )nly 17.5 seconds' average were required for the first five trials of
the postoperative retention tests. The methods of opening the pull and
crank boxes were the same before and after the operation. The per-
sistent weakness of the left arm called for a change in method of open-
ing the hasp box. Before operation, the animal had used the same
method in twenty consecutive trials. The plug was pulled out of the
staple with the right hand. The hasp was lifted from the staple with
the right hand, turned back against the lid, and then transferred to the
left hand. The lid was lifted with the left hand and the right hand

266 ARCHIVES OF NEUROLOGY AXD PSYCHIATRY

thrust into the box tor the food. In the postoperative retention tests,
the left hand was not used at all. She first lifted the lid with her right
hand, then released it and attempted to grasp the food with the same
hand, but the lid falling back prevented this. On the first trial, she
finally held the lid back with her right hand, inserted her head in the
box and took the food in her teeth. On the second trial, she thrust
her head against the lid after raising it with the right hand and so held
it open while the hand was inserted in the box. The same method was
used on the third and fourth trials. On the fifth trial, she attempted
to hold up the lid with her left foot and finally succeeded in this after
overbalancing twice. In all later trials, she released the lid and allowed
it to fall against her right arm as this was thrust into the box. These

Fig. 3. — The extent of the lesions in animal Xo. J. Reconstructed from
camera sketch and serial sections. Practically the whole of both precentral
gyri destroyed, including the paracentral gyrus and cortex within the central
sulcus.

various acts were carried out with definite adaptation to the contour
of the box and had none of the elements of random pulling and thrust-
ing which characterize the initial stages of learning. As in the case of
Number 2, there seemed to be an immediate adaptation of movements
to opening the box, which had not been employed at any time in the
previous practice.

Extent of Lesions. — The injured areas are shown in Figure 3 and sections in
Plate 3.

Right Hemisphere: The lesion began at the posterior median end of the
central fissure and included practically all of the precentral gyrus. On the
median surface, all of the cortex above the callosomarginal fissure was destroyed.
Practically all of the gyrus within the rolandic fissure was involved. Laterad.

I. ASHLEY— CEREBRAL FUNCTION

267

Plate 3. — Fig. A: Outline of the dorsal aspect of the cerebrum. The broken
areas outlined indicate the extent of dural adhesions. In a preliminary exam-
ination, the left hemisphere was cut through along the longitudinal dotted line.
The sections of its two halves are therefore from somewhat different levels.
The transverse broken lines show the level of the sections designated by the
corresponding numbers.

Figs. 1 to 5. — Camera sketches of sections showing extent of lesions. Blood
clots and scar tissue are indicated in solid black. Obviously degenerated cortex
is marked with coarse stippling. R, fissure of Rolando; S, fissure of Sylvius:
P , parieto-occipital fissure.

268 ARCHIVES OF NEUROLOGY AND PSYCHIATRY

the lesion included the upper half of the operculum. At most, only the lateral
facial area escaped destruction.

Left Hemisphere: The lesion was almost coextensive with that on the right.
More of the paracentral gyrus was destroyed and somewhat less of the
i iperculum.

After practically complete destruction of both precentral gyri, this
animal gave evidence of perfect retention of visual habits and habits
of manipulation. Direct adaptive changes in behavior were made to
compensate for weakness of the left arm.

DISCUSSION OF EXPERIMENTS

After extensive lesions to the precentral gyri of both hemispheres,
each of tbe animals studied gave clear evidence of the retention of
patterns of movement which had been acquired before the operative
destructions. Evidence of this retention was obtained from a com-
parison of the time required for opening the problem boxes in initial
training with that for the postoperative retention tests, from the restric-
tion of reactions in the postoperative retention tests to the catches of the
problem boxes, and from the persistence of individual peculiarities of
opening the boxes.

The average time required by all animals for opening each box in
the first five trials of training was 584 43 seconds. An average of only
30.7 seconds was required in the first rive trials of the postoperative
tests. The animals all failed the hasp box in the preliminary training
until first trained without the plug in the staple. In the postoperative
tests, they all opened this box promptly with the hasp closed by the
plug. At the beginning of the postoperative tests, each animal (except
Number 3 with the hasp box) used the same hand or hands for each
part of the manipulation as he had before the operation, and attacked
the latches in his former manner, although the methods of attack were
modified rapidly to compensate for the persistent motor difficulties. All
the animals showed perfect retention in the visual discrimination test.

These results establish conclusively that the cerebral areas destroyed
were not essential to the performance of the habits studied, and observa-
tions on the general behavior of the animals following recovery from
the paralysis justify the further conclusion that the areas are not essen-
tial to the performance of any type of complex adaptive or habitual
activity.

Four possible explanations of the results must be considered:

1. In no case did the operation destroy the entire precentral gyrus
of both sides. The parts remaining intact may have contained a suffi-

-13. This does not include the failures with the has]) box or the times when
Number 1 was left in the case over night.

LASHLEY— CEREBRAL FUNCTION 269

cient number of fibers previously integrated in tbe habit to produce the
conditioned reflexes, in spite of the great destruction of other fibers of
equivalent function. Such a possibility is supported by data on other
functional areas in which the various parts seem equipotential ( Franz, 4::"
Lashley34) and by the apparent equipotentiality of parts of tbe motor
area revealed by electrical stimulation (Lashley44). but several facts
speak strongly against this explanation.

Partial destructions usually entail a certain confusion in the per-
formance of all the functions of an area, which seems to exceed any-
thing of the sort noted in these animals.4"'

In Number 3, the destruction on both sides was so nearly complete
that only a part of the face areas could have remained functional.
If we attempt to explain the survival of habits as being due to the
activity of undestroyed parts of the motor area, we must assume that
a part of the face area is capable of performing all the functions of the
entire motor cortex — an assumption which is as far from the accepted
views of localization as is the denial of all habit function to tbe motor
areas.

2. It might be urged that in the recovery from the motor paralysis,
the vicarious functions assumed by other areas included the movements
involved in the problem-box habits ; that the habits were relearned dur-
ing the period of recovery from paralysis. The habits, howrever, con-
sist of particular patterns of movement associated with the stimuli
presented by the latch boxes. During the postoperative period, there
was no occasion for the animals to reacquire these particular patterns
of movement and no opportunity for the movements to be associated
with the latch boxes.

3. The long controversy concerning the sensorimotor function of
both the precentral and postcentral gyri suggests that the two may both
include centrifugal cells for the performance of habits. The literature
cited in the first part of this paper seems to establish the differential
function of the two areas, however, and the lack of paralysis after
lesions to the postcenral gyrus makes the hypothesis untenable.

43a. Franz, S. I.: On the Functions of the Cerebrum: The Frontal Lobes,
Arch. Psychol, 1907. No. 2, pp. 1-64.

44. Lashley, K. S. : Temporal Variation in the Function of the (iyrus Pre-
centralis in Primates, Am. J. Physiol. 65:585-602, 1923.

45. I am collecting data on this question at present. The evidence is not
complete, but there is indication that. e. g., any extensive but incomplete
destruction of the visual areas of both hemispheres in the rat is followed by
inaccuracy of brightness discrimination, with great variability from day to
day, such as has been reported by Franz in 1916 for aphasia, yet without any
complete loss of any phase of the visual function. Such loss as appeared in
the motor hahits of the monkeys was almost certainly ascribable to the simple
motor weakness, and gave no indication ol an apraxia.

270 ARCHIVES OF NEUROLOGY AND PSYCHIATRY

Nanagas 46 found a few islands of large pyramidal cells in the postcen-
tral gyrus, but the great mass of them was restricted to the precentral.
Finally, Brown 36 reported that the destruction of the postcentral gyrus
did not abolish learning ability or interfere seriously with habits formed
before the operation in the chimpanzee.

4. The only remaining possibility seems to be that the electro-
stimulable areas do not include the centrifugal elements of conditioned
reflex arcs of any sort. ( It is of course possible that they contain
some such elements, but these cannot comprise any significant propor-
tion of the total number of centrifugal cells, since their destruction
leaves the habits completely unaffected.) In this, the experiments con-
firm for primates the results previously reported for the rat. The
neural impulses involved in conditioned reactions do not pass from
sensory projection areas to the precentral gyrus and thence to lower
centers, but must be conducted by centrifugal cells lying outside of the
pyramidal system. In the rat. the evidence points to the view that the
centrifugal fibers of the sensory projection area itself are primarily
involved in this motor function, since the destruction of any fourth of the
cerebrum exclusive of the visual areas does not affect the performance
of visual habits.47 Whether or not the same lack of important tran>-
cortical conduction holds true for the monkey is questionable in view of
the greater proportionate development of the transcortical association
tracts in this animal, but it seems established that the "motor areas" are
not concerned in the initiation of habitual movements.

THE CORPUS STRIATUM AND VICARIOUS FUNCTION

All recent students of the question agree that recovery from cerebral
paralysis is not due to the assumption of the function of the destroyed
motor cortex by the corresponding area of the opposite side. On the
contrary, the simultaneous destruction of the areas in both hemispheres
seems to be followed by a rather more rapid recovery than follows the
destruction of either alone. The fact has been noted by Grunbaum and
Sherrington 20 and by Wagner -\ It was apparent in the slower recov-
ery of Number 2 from the first operation than from the second. The

46. Nanagas, J. C. : Anatomical Studies on the Motor Cortex of Macacus
rhesus, J. Comp. Neurol. 35:67-96. 1922.

47. Experiments now in progress, which indicate that extensive fronto-
parieto-temporal lesions may also abolish visual habits without producing
a general deterioriation of learning ability, indicate that a mass action of the
cerebrum is also somehow involved, but they do not seem to invalidate the con-
clusion that the efferent fibers of the sensory projection area are primarilv
concerned in the subcortical initiation of movements associated with the receptor
for that area.

LASHLEY— CEREBRAL FUNCTION 271

explanation is probably to be found in the forced practice which
diplegia imposes on the paralyzed limbs (Odin and Franz4").

Other restricted cerebral areas have also been rather definitely
excluded from participation in the vicarious function of the stimulable
areas, by the work of Leyton and Sherrington 2,i ( Lashley 49) . Luciani :"'
has suggested that recovery may be due to the activity of the corpus
striatum, which has homologies with the stimulable cortex. This was
tested in animal Number 2 of the present series by destruction of the
right striatum after recovery from diplegia. Hemiplegia symptoms did
not recur, so that we may conclude that the recovery had not been due
to the vicarious activity of the striate nucleus.

THE FUNCTION OF THE ELECTROSTIMULABLE AREAS

The conclusions which may be drawn from these experiments are
wholly negative. They seem to prove that the precentral gyrus does
not include the efferent paths for learned activities ; in the current
localization terminology, it is not the center for "voluntary movements/'
as is almost universally assumed. But if this is true, what is the
significance of the movements elicitated by electrical stimulation? How
may we interpret the cerebral paralyses, and why do they especially
affect the finer manipulative movements? A number of lines of evi-
dence may help to answer these questions and clear up tbe function of
the precentral gyri.

The Postural Function of the Stimulable Areas. — Many investi-
gators have pointed out the similarity between the movements elicited
by cortical stimulation and "voluntary movements." I am convinced
that this is an error due to contrasting these movements with those
which are elicited by stimulation of motor nerves or spinal cord. In
the latter cases, the movements are wholly incoordinated. whereas the
movements following cortical stimulation involve synergic groups of
muscles. But in all cortical stimulation experiments which I have seen,
the movements have been slow and rather massive, i. e., chiefly involving
the larger musculature of the limbs. When smaller segments are
moved, the movements are never coordinated as they are. for example,
in grasping small objects. They never show the fineness of gradation
and accuracy of adjustment which is characteristic of the movements
of the intact animal. This has been observed by various investigators
and interpreted as showing that the finer adjustments are integrated

48. Odin, R., and Franz, S. I : On Cerebral Motor Control: The Recovery
from Experimentally Produced Hemiplegia, Psychobiol. 1:33-50, 1917.

49. Lashley, K. S.: Studies of Cerebral Function in Learning. Vicarious
Function After Destruction of the Visual areas, Am. 1. Physiol. 59:44-71. 1922.

50. Luciani, L. : Human Physiology, London 3. 1915.

272 ARCHIVES OF NEUROLOGY AXD PSYCHIATRY

at some higher level and imposed through it on the motor area. But
there is no direct evidence that this is the case. The movements fol-
lowing excitation are far more like the gross changes of posture which
one may observe in the intact animal — the raising of an arm preparatory
to snatching at food, bracing against a pressure, or the like. It seems
significant that coordinated movements of the eyes are among the most
easilv elicited movements on electrical stimulation ( although their stimu-
late points lie outside of the precentral areas), and that these move-
ments in the intact man or animal are almost always a reflex fixation
(postural adjustment) called out directly by exterostimulation and, in
fact, can not be accurately performed in the absence of such stimulation,
as with lids closed.

Wilson 51 has pointed out the similarity of the contractures in cere-
bral paralysis to the postural reflexes of decerebrate rigidity, and from
this it seems certain that a part of the function of the stimulable areas
is the regulation of these spinal and cerebellar postures. It seems rather
probable that the movements obtained on electrical stimulation are only
a further exhibition of this postural activity and are unrelated to the
finer coordinations of conditioned motor reflexes, or motor habits.

The Dynamic Function of the Stimulable Areas. — The condition
following lesions to the precentral gyrus or internal capsule, even in
man, should be described rather as an enormous difficulty in making
movements than as an absolute paralysis of movement. The degree of
paralysis varies somewhat from day to day. Excitement seems to
increase motor control (Minkowski.52 Lashley37), and the paralvsis
may in part or wholly disappear during emotional disturbance, only to
recur when the disturbing situation is past. If we may judge from the
tonic condition of the muscles, there must be in excitement a ireneral
facilitation of lower motor centers which temporarily reinstates cerebral
control. Further, if this is the case, in the intact animal in the absence
of emotional stimulation cerebral control must likewise be conditioned
by some such facilitation derived from the precentral gvrus. The work
of Brown 53 and of Leyton and Sherrington 26 has shown that stimula-
tion of the motor area does facilitate the centrifugal paths of other

51. Wilson, S. A. K. : On Decerebrate Rigidity in Man and the Occurrence
of Tonic Fits, Brain 43:220-268, 1920.

52. Minkowski, M. : Etude physiologique des circonvolutions rolandique et
parietal. Arch. Suisse de Neurol, et Psychiat. 1:389-459, 1917.

53. Brown, T. G. : Studies in the Physiology of the Nervous System. XXV.
On the Phenomenon of Facilitation. 4. Its Occurrence in the Subcortical
Mechanism by the Action of Which Motor Effects Are Produced on Artificial
Stimulation of the "Motor" Cortex. J. Physiol. 9:131-145. 1915; also Footnote 36.

LASHLEY— CEREBRAL FUNCTION 273

areas, either at cortical or subcortical levels, since stimulation of the
precentral gyrus renders the otherwise inexcitable postcentral gyrus
excitable for corresponding movements.

After partial recovery from cerebral paralysis, the most prominent
symptom is the weakness of the formerly paralyzed limbs. The greater
part of their repertoire of movements may be restored, speed may be
nearly normal, but only a slight force can be exerted, and fatigue occurs
readily. What is lacking in this condition is not an adequate integra-
tion of the motor impulses, but a sufficient mass of neural impulses to
maintain muscular activity. This may be ascribed either to a reduction
in the number of functional nerve cells, or to inadequate facilitation.
The ready fatigability is evidence for the latter.54

There is evidence that the withdrawal of facilitation derived from
other sources will produce similar weakness and fatigability and a ten-
dency not to use the affected parts. Thus Munk 55 has shown that
denervation of a limb has such effects, and Sherrington's 50 work has
demonstrated that they are due to the withdrawal of impulses derived
largely from the denervated muscles.

The importance of such facilitating systems has been emphasized
by a number of recent investigators (Monakow,57 Wilson,42 Sherring-
ton,58 Tournay,59 Hunt00). The general conception of these investi-
gators is of a series of hierarchies of motor reflexes, all exerting a
facilitating influence on the final common path. These involve at least
the following elements.

54. The all or nothing principle of nerve and muscle activity requires the
assumption that strength of muscular contraction is dependent upon the number
of motor fibers involved and the rate of succession of propagated disturbance-.
Piper's work ( Electrophysiologic menschlicher Muskeln, Berlin, 1912) indi-
cates that fatigue involves a decrease in this rate rather than a reduction in the
total number of muscle cells activated.

55. Munk, H. : Ueber die Folgen des Sensibilitatsverlustes der Extremist
fur deren Motilitat, Sitzungsber. d. Berlin Akad. Wiss., pp. 1038-1077, 1903.

56. Sherrington, C. S. : The Integrative Action of the Nervous System.
London, 1911.

57. Monakow, C. von: Aufbau und Lokalisati.m der Bewegungen brim
Menschen, Ber. uber d. iv. Kongress f. exp. Psychol, in Innsbruck. 1910.

58. Sherrington. C. S. : Postural Activity of Muscle and Nerve, Brain 38:
191-234, 1915.

59. Tournay, A.: Conception actuelle des grande fonctions motrice, J. de
Psychol. 17:904-930, 1920.

60. Hunt. R. : The Static and Kinetic Systems of Motility. Arch. Neurol.
& Psychiat. 4:353, 1920.

274 ARCHIVES OF NEUROLOGY AXD PSYCHIATRY

1. Excitation of the motor cells supplying a muscle by impulses
derived from the receptors in the muscle itself (Sherrington56).

2. Long spinal reflexes from synergic muscles (Sherrington,56
Magnus 61).

3. Other proprioceptive and general exteroceptive facilitation whose
central mechanism is as yet rather obscure (Yerkes,62 Richter63).

4. Vestibular and proprioceptive influences exerted through the
mechanisms of the cerebellum.

5. Probably facilitation derived from thalamic mechanisms in emo-
tional excitement (Head'14).

6. Kinetic influences of obscure origin integrated in the corpus
striatum (Wilson,42 Hunt60).

Interference with any of these mechanisms is able to produce a
change in the excitability of the final common path, and in the intact
organism it seems certain that every act involves the participation of
all of them, both by excitation and inhibition.

These considerations make it possible to form a tentative hypothesis
concerning the function of the precentral gyrus. Its demonstrated
facilitating effects, and its lack of direct participation in the conditioned
reflex arc seem to throw it into a class with these other postural and
tonic systems. Cerebral paralysis, is, I believe, to be interpreted as
showing that a normal function of the stimulable cortex is to supply a
substratum of facilitating impulses which act in some way to render
the final common paths excitable by the more finely graduated impulses,
descending from the cortex by extrapyramidal paths and producing the
finer shades of adaptive movement. In other words, impulses descend-
ing from the precentral gyrus do not initiate the finer adaptive move-
ments through the lower motor neurons, but only "prime" these cells
so that they may be excited bv impulses from other sources. The
source of this activity and the probable interrelations of the stimulable
areas with other parts of the motor system and with sensory projection
areas present problems too complex for discussion here. Unquestion-

61. Magnus. R., and de Kleijn, A.: Die Abhangigkeit des Tonus der Extre-
mitatenmuskeln von der Kopfstellung, Arch. f. d. ges. Physiol. 145:455-548, 1912.
Magnus, R. : Welche Teile des Centralnervensystems mussen f iir das Zustande-
kommen der tonischen Hals- und Labyrinthreflexe auf die Korpermusku-
latur vorhanden sein ? Arch. f. d. ges. Physiol. 153:224-250, 1914.

62. Yerkes, R. M. : Inhibition and Reinforcement of Reactions in the Frog,
J. Comp. Xeurol. & Psychol. 14:124, 1904.

63. Richter, C. P.: A Behavioristic Study of the Activity of the Rat, Comp.
Psychol. Monogr. 1:1-55. 1922.

64. Head, H.: Studies in Neurology, London, 1920; Release of Functions in
the Nervous System, Proc. Roy. Soc. 92 B: 184-209, 1921.

LASHLEY— CEREBRAL FUNCTION 275

ably, the areas receive excitations from other parts of the cerebrum 65
and it is probable that all parts of the kinetic system ,irt are capable of
mutual influence. Postural facilitation and inhibition may themselves
be habitual responses, but the present experiments indicate that they
are rather generalized and not independently organized for each specific
manipulative habit.

Recent work in general tends to emphasize the complexity of neural
functions. We must hesitate to ascribe an exclusive or precise function
to any neural structure, for the evidence points rather to the view that
observable behavior is always the product of the interaction of many
neural systems and that the function of any system is dependent on its
temporary physiologic relation to other systems. This is particularly
true of the finer adaptive responses of the intact animal which are
subject to inhibition and facilitation by innumerable factors. Their
execution depends on preparatory postural adjustments, emotional and
other dynamic facilitation, as well as integration of impulses from many
exteroceptors.67 The total mass of excitation is effective both through
the specific efferent patterns activated ar.d also through the general
dynamic effects which alone a?e incapable of producing the overt motor
reactions elicited. The experiments reported here indicate that the
dc'vtrostimulable areas are rather more concerned with the maintenance
of excitability and the regulation of postural reflexes than with the
excitation and control of finely integrated adaptive movements.

65. I have made several attempts to isolate the area from other parts of the
cortex by circumsection but have not yet been successful. The literature on this
point is conflicting. Marique (Brain 8:536-538, 1885) reported the same results
from circumsection as from excision of the area. Exner and Panetli (Arch. f.
d. ges. Physiol. 44:544-555. 1889) found similar results but were inclined to
ascrihe them to interference with the blood supply of the area. Schafer (Jour.
Physiol. 26:23-25. 1901) reported one case of complete circumsection without
paralysis. He does not report histologic examination of the lesion, however,
and in view of the difficulty of the operation there is not sufficient evidence
that the isolation was complete.

66. In this discussion, I have disregarded the important conception of static
and kinetic functions advanced by Hunt (Arch. Neurol. & Psychiat. 4:353. 1920)
because the evidence does not show clearly to which of his systems the elect n>-
stimulable cortex is to be referred. The postural influences of the area would
indicate a static function. What I have called the dynamic or "priming" func-
tion is rather a kinetic function, but is more primitive than the activities implied
in Hunt's conception of the neokinetic system.

67. The statement that every act of the intact organism involves the par-
ticipation of every neuron within the central nervous system is probably no
more of an exaggeration than are the extreme theories of precise localization of
function or of isolated conditioned reflex paths.

276 ARCHIVES OF XEUROLOGY AND PSYCHIATRY

SUMMARY

The greater part of the precentral gyrus of both hemispheres was
destroyed in monkeys which had been trained previously in habits of
manipulation and visual discrimination. When the animals recovered
from paralysis, it was found that they showed perfect retention of these
habits. From this it is concluded that the so-called motor areas are not
directly concerned with the performance of complex learned activities.
The motor impulses of conditioned reflexes must descend from other
areas of the cerebral cortex than the precentral gyri, and the latter
cannot be regarded as the source of impulses to "voluntary movements.'*

Destruction of the corpus striatum subsequent to recovery from
diplegia produced only the usual symptoms of striate lesion without
recurrence of the symptoms of cerebral paralysis. Recovery from
paralysis was therefore not due to vicarious function of this nucleus.

The evidence for considering the precentral gyrus as a part of the
kinetic mechanism for reflex control of spinal posture and for mainte-
nance of excitability of lower motor centers is summarized.

I

QL
785
L37
v.5

Lash ley, Karl Spencer
Studies of cerebral
function in learning

Biological

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY