- Email: [email protected]
Distribution of slow cortical potentials preceding self-paced hand and hindlimb movements in the premotor and motor areas of monkeys
Brain Research, 224 (1981) 247-259 Elesevier/North-Holland Biomedical Press
247
D I S T R I B U T I O N OF SLOW C O R T I C A L P O T E N T I A L S P R E C E D I N G SELFPACED HAND AND HINDLIMB MOVEMENTS IN THE PREMOTOR AND M O T O R A R E A S OF M O N K E Y S
SHUJI HASHIMOTO, HISAE GEMBA and KAZUO SASAKI* Department of Physiology, Institute for Brain Research, Faculty of Medicine, Kyoto University, 606 Kyoto (Japan)
(Accepted March 12th, 1981) Key words: premovement slow cortical potential - - hand movement - - hindlimb movement - premotor cortex - - motor cortex - - monkey
SUMMARY Surface negative-depth positive, slowly increasing potentials prior to self-paced hand and hindlimb movements were recorded in the dorsal aspect of the motor and premotor cortices with chronically implanted electrodes. It was shown that the potentials were recorded in the contralateral forelimb m o t o r area prior to hand movements but were hardly seen in the hindlimb motor area. On hindlimb movements, the contralateral hindlimb motor area showed the premovement potentials, whereas the forelimb m o t o r area revealed little or no prernovement potentials. The contralateral premotor cortex was shown to induce the premovement potentials in its wider areas and participate in both of hand and hindlimb movements in a similar fashion, with predominances in its dorsolateral portion for hand movements and in its dorsomedial portion for hindlimb movements respectively. In the hemisphere ipsilateral to the moving hand, the relatively large premovement slow potentials emerged frequently also in the premotor cortex, whereas only the small potential was obtained from the forelimb motor area. These results suggest that the premotor cortex (area 6) participates in the more general and associative organization of motor function than the motor cortex (area 4) which represents the specialized role in the motor performance.
* To whom correspondence should be addressed. 0006-8993/81/0000-0000/$02.50 © Elsevier/North-Holland Biomedical Press
248 INTRODUCTION We reported that slowly increasing cortical potentials preceding self-paced hand movements were recorded in the lateral portion of dorsal aspect of the premotor cortex and in the forelimb areas of the motor and somatosensory cortices contralateral to the moving hand of monkeys 5,6. The similar distribution of premovement potentials (motor potentials) was recently reported in monkeys and human subjects 9. Such premovement slow potentials started about one second before the hand movement elevating a lever and showed the transcortical reversal of electrical polarity; negativity at the cortical surface and positivity at a cortical level of 2.5-3 mm below the surface. They were interpreted to be due mainly to the currents of EPSPs generated in superficial parts of apical dendrites of cortical pyramidal neurones via certain thalamocortical (T-C) projections, and were supposed to participate in initiating and/or controlling the movement 8. It was further suggested that the slow potential in the motor cortex is at least partly induced by the activity of the hemispherical part of the cerebellum through the superficial T-C projections 13 (see ref. 12). The slow potentials in the premotor, motor and somatosensory cortices were also shown to relate closely to the control of required muscle force in expectation 7. The previous papers mainly dealt with the potentials in the lateral portion of dorsal aspect of the premotor cortex and in the forelimb area of the motor cortex, when monkeys were performing the hand movement elevating a lever. The present study was devoted to explore the range of responsive cortical areas in these cortices, particularly whether or not the dorsomedial portion of the premotor cortex and the hindlimb area of the motor cortex would be included as responsive areas for hand movements. It will be shown in this report that the hindlimb area of the motor cortex hardly shows premovement slow potentials on the lever-elevating hand movement. On the contrary, the premotor cortex in its wider area is active on the hand movement, and even the premotor cortex ipsilateral to the moving hand is also involved on the movement. The cortical potentials prior to self-paced hindlimb movements will also be briefly reported in order to collate such distribution of premovement cortical potentials for hand movements with that for hindlimb movements. MATERIALS AND METHODS Experiments were carried out on seven adult monkeys (Macacafuscata). All of them were trained to perform lever-elevating hand movements and three among them were trained to perform also hindlimb movements pressing a pedal at a time interval of more than two seconds without any instructing stimuli (see Fig. 1). A load opposing lever elevation was sometimes given in a series of daily experiments by a cablepulley system so that the premovement slow potentials prior to self-paced hand movements increase in their amplitude (see ref. 7). Elevating a lever or pressing a pedal by the correct movement delivered a small amount of juice about 600 ms after the movement. Monkeys usually performed the movements several hundred times in a series on each day.
249
EMG
S
D
' '\'
nn • ,
>
;,
BONE
'"' " { , .
DuRA
',,,:
CORTEX
Fig. 1. Diagrams illustrating forelimb and hindlimb movements, and the electrode array implanted in the cortex. Upper diagram: lever-elevating hand movement with electromyogram (EMG) record. M.S., microswitch generating the timing pulse. The lever is prevented from falling down by stopper (STOP.). Lower diagram, right: pedal-pressing hindlimb movement with EMG record from gastrocnemius-soleus muscles. Lower diagram, left: the recording electrode implanted in the cortex. The surface needle (S) placed on the cortical surface and the depth one (D) at about 3 mm below the surface.
Recording and data processing Most of recording procedures were previously described in detail 6. The surgery for implanting the electrodes in various parts of the cerebral cortices on both sides was performed under anesthesia with ketamine hydrochloride (7 mg/kg, i.m., every 1-2 h). Each electrode array consisted of two silver needles (0.25 m m in diameter) attached together, of which the tips were arranged so that they were on the cortical surface (S) and at a depth of 2.5-3 m m in the cortex (D) respectively when implanted (see Fig. 1, lower diagram to the left). The silver needles were insulated except for the chloridized tip. The surface and deep cortical layer potentials were led simultaneously from the needles respectively in referring to the indifferent electrode, which was composed of linked two Ag-AgCI electrodes buried in the bone behind the ear on both sides. They were respectively amplified by AC amplifiers with time constant of 2.0 s, and then stored in cassette tapes through multichannel data recorders together with the timing pulse generated at the moment of microswitch closure by lever-elevating or pedalpressing movements. 'Electrooculogram (EOG)' in this study was taken from the electrode implanted in the rostral edge of frontal bone above the orbita on each side in reference to the indifferent electrode mentioned above, and amplified by an AC amplifier with time constant of 2.0 s. Electromyogram (EMG) was led bipolarly from silver plate electrodes attached to the skin over wrist extensor muscles on the hand movement or over plantar flexor (gastrocnemius-soleus) muscles on the hindlimb movement and amplified with an AC amplifier, of which the time constant was 0.01 s. ' E O G ' and E M G were stored likewise in cassette tapes.
250
o
200
,I,
EOG
~ ~00pV EOG
-
-
~'~._._.--
"
S-D C s
D
$-: "
~
/
Z
D
t
+
$-D
EMG
1
t
L_
.
Fig. 2. Cortical potentials in the motor cortex on the side contralateral to the moving hand. The number of samples for averaging was 100. The left and central columns show simultaneous records from the hindlimb area (A) and the forelimb areas (B and C) of the motor cortex, together with EOG and E M G activities in one and the same monkey. The left (0) and central (200) columns present the potentials without and with the additional load of 200 g, respectively, being obtained from a series on the same day. Simultaneous records of cortical potentials from the forelimb areas (D and C) of the motor cortex in another monkey are shown in the right column. The surface (S) and depth (D) cortical potentials and the electronic subtraction of the depth record from the surface record (S-D) are shown in respective rows, together with EOG and EMG. The instant of microswitch closure by hand movements is indicated by upper and lower arrows in this and all subsequent figures. 500/iV calibration is for EOG potentials and 50/zV for the cortical potentials. Negativity is represented by upward deflection. 500 ms time scale is applied to all records. S.C., sulcus centralis. Note the premovement slow potential is not recorded in the hindlimb area of the motor cortex (A).
251 Cortical and 'EOG' potentials were usually averaged 100 times, or occasionally 20 times, in the period between 1.5 s before and 0.5 s after the timing pulse. The surface cortical potentials were electrically subtracted by the deep cortical layer potentials and the results (S-D potentials) served as an indication for true transcortical potentials without marked contamination of EOGs and of some other potentials generated in sites remote from the recording electrodes (see ref. 6). EMGs were passed through a full-wave rectifier circuit and then averaged likewise. After such daily experiments for several months, cortical sites of implanted electrodes in the cerebral cortex were checked at autopsy. RESULTS This report deals with the cortical potentials preceding movements, particularly surface negative-depth positive potentials that reach their summit 50-200 ms before movements, and will not treat the potentials thereafter (see ref. 6). The averaged cortical potentials in the motor cortex contralateral to the moving hand are shown in Fig. 2, which were obtained from two different monkeys performing hand movements. The left (0) and central (200) columns exemplify the slow potentials obtained from one and the same monkey performing lever-elevating hand movements, respectively without additional load and with additional 200 g-load opposing the lever elevation, in a series of the same day. The potentials in the right column were taken from another monkey. The potentials were recorded from the hindlimb area of the motor cortex (A) and the forelimb areas of the motor cortex (B and C in one monkey, D and C in another monkey). As shown in Fig. 2, surface negative-depth positive, slowly increasing potentials prior to the movement were recorded in the forelimb areas of the motor cortex (B, C and D) in the two monkeys, but they were not appreciably seen in the hindlimb area, except small surface positive-depth negative potentials just before the movements that would not be dealt with as mentioned above. Configurations of the potentials were similar within one subject, although their sizes differed from each other. The slow potentials in C were larger than those in B in the one monkey as shown in the left and central columns, and, in another monkey, the potentials in C, which was nearer to the central sulcus than D, were larger than those in D. This would correspond to the recent finding that antecedent cortical activity for wrist extension movements is maximal in the posterior margin of the precentral gyrus 1. With 200 g-load, the slow potentials in the forelimb area of the motor cortex were increased in accordance with the required muscle force to elevate the lever, as seen in the left and central columns (see also ref. 7). Even under such conditions, surface negative-depth positive slow potentials before movement appeared barely in the hindlimb area of the motor cortex (A, central column). Small negative shifts without transcortical reversal of electrical polarity as in S and D rows of A in the central column can be considered to be due to passive spread of some potentials generated in some remote sites from the recording electrode 14. No significant potential changes before hand movements were recorded in more medial sites to A (not shown in this figure). The results reveal that the premovement, surface negative-depth positive slow
252 potentials are recorded in the forelimb area of the motor cortex, but not in the hindlimb area in association with the lever-elevating hand movement. It was usually observed that the rostral part of the forelimb motor area produced a little smaller premovement potential than the caudal part near the central sulcus, as shown in the right column of Fig. 2. The potential in the rostral part was often even smaller than the largest one in the premotor cortex so that a small potential was interposed between the large potentials in the premotor cortex and in the caudal part of the motor cortex. Fig. 3 presents slow potentials preceding self-paced hand movements in the dorsal aspect of the contralateral premotor cortex (A, B, C and D), together with that in the forelimb area of the motor cortex (E). All records in the premotor cortex showed the surface negative-depth positive slow potentials, although the amplitude was larger in the lateral portion (C and D) than in the medial (A and B). Such relatively diffuse appearance of the potentials in the premotor cortex was in contrast to the distribution of premovement slow potentials in the motor cortex; such potentials could not be recorded in the dorsomedial portion of the motor cortex (hindlimb area). This figure also shows, as reported previously 7, that the contour of the premovement slow potentials in the premotor cortex differs from that in the motor cortex. The
C
E
i25 IJV
S-D ~ Fig. 3. Specimen records of cortical potentials in the premotor cortex (A-D) on the side contralateral to the moving hand, together with those in the forelimb area of motor cortex (E). The recording sites are indicated in the inset brain diagram with respective letters. All records were taken in one and the same monkey and all except B were simultaneously recorded. Averaged of 200 samples. 25/*V calibration and 500 ms time scale are applied to all cortical potentials. Abbreviations as in Fig. 2.
253 potentials in the p r e m o t o r cortex started earlier, usually by a few hundred ms, and reached the summit a b o u t 100-200 ms earlier than those in the m o t o r cortex. The premovement slow potentials in the hemisphere ipsilateral to the moving h a n d are exemplified in the left c o l u m n o f Fig. 4. F o r collation, the slow potentials in the contralateral hemisphere o f the s a m e m o n k e y are also shown in the right column o f Fig. 4, being obtained in a series o f the same day. Surface negative-depth positive, slowly increasing potentials with fairly large amplitudes were encountered in the ipsilateral p r e m o t o r cortex (A), while those in the ipsilateral m o t o r cortex (B) were considerably smaller in this case than in the contralateral m o t o r cortex. It was
EOG
A
s
S-
S-b
EMG
Fig. 4. Slow potentials recorded from the premotor (A) and motor (B) cortices on the side ipsilateral and contralateral to the moving hand. Average of 100 samples. All records were obtained from the same monkey, and in a series of the same day. The potentials in the premotor and motor cortices on each side were recorded simultaneously. Abbreviations and voltage calibrations are as in Fig. 2. Time scale is 250 ms.
254 characteristic of the premovement slow potentials on the lpsilateral side that the potential in the premotor cortex was relatively larger than in the motor cortex, although the slow potentials on the ipsilateral side were in general rather variable in size. In contrast, peak amplitude of the premovement slow potential in the premotor cortex was always smaller than that of the motor cortex on the contralateral side in all monkeys tested, as shown in the right column of Fig. 4 (see ref. 5). The upper part of Fig. 5 (hindlimb) shows potentials preceding self-paced hindlimb movements in the motor cortex on the side contralateral to the movements. As shown in A (recorded in the hindlimb area of the motor cortex), surface negative-depth positive potentials, although much shorter in duration than those on hand movements, were recorded before the movements. On the other hand, the potentials obtained from the electrode implanted in B (forelimb area of the motor cortex) showed no obvious transcortical reversal of electrical polarity on the hindlimb movement. Traces of S-D in the lower part of Fig. 5 (forelimb) reveal potentials preceding hand movements obtained from A and B with the same electrodes in the same monkey as mentioned above. As has been described, the pre-
S
-
~
/
/
50 EMG FOREL1RB
pV
A___ t
-,,) Fig. 5. Records obtained on hindlimb movements (upper part) and on hand movements (lower parts). Upper: potentials in the hindlimb areas (A and C) and the forelimb area (B) of the motor cortex contralateral to the moving hindlimb are shown in one and the same monkey. A and B were on the same but C on the opposite side of hemispheres. The hindlimbs pressing a pedal were respectively on the contralateral side to the hemispheres under recording. The traces in A and B were concurrently recorded. Lower: potentials in the same hindlimb area (A) and forelimb area (B) of the motor cortex contralateral to the moving hand. They were simultaneously recorded with the same electrodes (A and B) as in the upper part, being averaged for 100 times. Abbreviations and calibrations as in Fig. 2.
255 movement slow potential was encountered in the forelimb area of the motor cortex (B) with hand movements, while such a potential was not recorded in the hindlimb m o t o r area (the lower part in column A). Records in C are the slow cortical potentials in the hindlimb motor area on the contralateral hindlimb movements in another hemisphere, presenting more predominant surface negative-depth positive potentials with relatively short duration. These results show that only the hindlimb area within the motor cortex becomes active to produce the potentials prior to hindlimb movements. The contours of premovement potentials in the forelimb and hindlimb areas on respective movements are different from each other. Surface negative-depth positive potentials are much shorter for the hindlimb activity than the forelimb one. The physiological significance of such difference can not be explained at present. The surface negative~iepth positive potentials in the hindlimb motor area were occasionally preceded by slowly increasing, surface positive-depth negative potentials, which lasted for nearly 1 s, as seen in A and C of Fig. 5. They were very small, if present, in size, and were inconstantly observed even in the same recording site on different days. It is not certain whether the slow potentials would be active responses in this area of the cortex or not. Further studies will be required to clarify the nature of the potentials.
B
s-D C
s-B
Fig. 6. Slow potentials recorded on contralateral hindlimb movements with the same electrodes in the same monkey as in Fig. 3. All records except in C were obtained concurrently; 200 samples were averaged. Calibrations as in Fig. 3; abbreviations as in Fig. 2.
256 In contrast to the motor cortex, surface negative-depth positive slow potentials could be recorded in wide portions of the premotor cortex prior to hindlimb movements. Fig. 6 gives the potentials in the premotor cortex contralateral to the moving hindlimb in the same monkeys as Fig. 3. Premovement potentials appeared, in the premotor cortex, with a relatively large amplitude and a considerable duration on hindlimb movements. Their peaks emerged at the almost same moment before the movement, as in the case of forelimb movement shown in Fig. 3. The sizes of the potentials in the sites of A and D were occasionally smaller than in B, although the sizes in A, B and D were nearly same in this case as revealed by S-D traces. The results indicate that the premovement slow potentials in the premotor cortex appear in a similar fashion on both hindlimb and forelimb movements, and that preponderance shifts to the medial portion of dorsal aspect of the premotor cortex for hindlimb movements. In contrast, the forelimb motor area (E) hardly showed such transcortical reversal of electrical polarity as seen in the premotor cortex on hindlimb movements (cf. Fig. 3E). Fig. 7 summarizes the results of the present study. Prior to self-paced hand movements, the forelimb area of the motor cortex on the side contralateral to the moving hand became active and the hindlimb area remained inactive, whereas both of the lateral and medial portions of dorsal aspect of the premotor cortex revealed obvious premovement potentials with predominance in the lateral portion. In the ipsilateral hemisphere, relatively large potentials with transcortical reversal of polarity often appeared in the premotor cortex, while the forelimb area of the motor cortex gave merely small potentials. In association with hindlimb movements, the hindlimb area of the motor cortex contralateral to the movements showed marked premovement potentials, but the forelimb area revealed scarcely transcortical reversal of electrical polarity. The premotor cortex showed, prior to hindlimb movements, the slowly increasing potentials in both of the lateral and medial portions of its dorsal aspect with some predominance in the medial portion. DISCUSSION The present results revealed that the forelimb area of the motor cortex becomes active to produce the surface negative--depth positive, slow potentials before hand movements on the contralateral side, and that the hindlimb area of the motor cortex remains inactive on the movements. The hindlimb motor area induced the premovement slow potentials in association with hindlimb movements which were hardly observed in the forelimb motor area. Such somatotopical representation in the motor cortex was previously reported on slow potentials prior to voluntary movements recorded from human scalp 15. On the other hand, the premotor cortex on the side contralateral to movements was shown to participate in both of the forelimb and hindlimb movements in a similar fashion, although some quantitative difference was noted between the slow potentials preceding forelimb and hindlimb movements. Relatively large potentials emerged in the lateral portion of the dorsal aspect of the premotor cortex prior to hand movements and those in the medial portion before
257 hindlimb movements. The fairly large premovement slow potentials were frequently encountered also in the premotor cortex ipsilateral to the moving hand, whereas the forelimb area of the motor cortex in the ipsilateral hemisphere gave only small potentials prior to hand movements. It can be summarized as follow; as for the premovement slow potentials, the motor cortex has the somatotopical relationship to contracting muscles on the contralateral side, while the premotor cortex participates relatively in a diffuse manner in forelimb and hindlimb movements, and also bilater-
FORELIMB
I PS! L A T E R ~
HINDLIMB
CONTRA~
"" ~"
Fig. 7. Distribution of premovement slow cortical potentials in the dorsal aspect of the premotor and motor cortices. Upper and middle diagrams show the distribution in the contralateral and ipsilateral hemispheres on the forelimb movements. The lower diagram gives the distribution in the contralateral hemisphere on hindlimb movements. Large filled circles represent relatively large potential difference (S-D of usually more than 50/*V) with transcortical reversal of electrical polarity; small circles indicate small differences (usually less than 50 #V); asterisks show inactive points, s.a., sulcus arcuatus; s.c., suicus centralis; s.i., sulcus intraparietalis.
258 ally in the movements. The premovement slow potentials in the premotor and motor cortices showed the different time courses as mentioned in Results (see also ref. 7). These differences between the premotor and motor cortices may suggest that the premotor cortex is a different functional entity from the motor cortex for movement initiation and/or motor performance. Woolsey 16 placed the anterior boundary of 'precentral motor cortex' in the premotor cortex (area 6) by stimulation experiments; the axial musculature was reported to be represented in the caudal portion of area 6 (Bucy's excitable portion of area 62). However, the present results show that the rostral portion of dorsal aspect of the premotor cortex (A and C in Figs. 3 and 6), which is not involved in Woolsey's 'precentral motor cortex', has given the similar premovement slow potentials to those in the more caudal portion of area 6 (B and D in Figs. 3 and 6), which may be included in Woolsey's 'precentral motor cortex'. In addition, the contours of premovement slow potentials obtained from the electrodes within the motor cortex (area 4) were similar to one another, but differed from those in area 6. A smaller premovement potential was occasionally noted in the border area between areas 6 and 4 than premovement potentials in areas 6 and 4 as mentioned in Results. These data may suggest that the boundary for functional entities based on the premovement slow cortical potentials is more reasonably placed between areas 6 and 4, following the cytoarchitectonic studies 3, than on the anterior border of Woolsey's 'precentral motor cortex'. However, further careful examination should be made for the possibility of simultaneous activities in truncal and proximal limb muscles supporting the hand and hindlimb movement, although sampling E M G records from these muscles has so far revealed no marked activities. The dorsal aspect of the premotor cortex (area 6) represented the considerably wide responsive area of premovement slow potentials, while the motor cortex showed the fairly definitive somatotopical representation. Recently, measurement of regional blood flow in human subjects revealed that simple repeated finger movements were associated with the activation (increase of blood flow) of the contralateral forelimb motor area, and that the finger movement for a maze test activated the contralateral forelimb motor area and the bilateral premotor areas 11. Furthermore, programming a sequence of fast isolated finger movements and the maze test was reported to be accompanied with a marked increase of regional I blood flow in the supplementary motor areal0, H. The activation of the contralateral motor cortex and the bilateral premotor cortices might be related to the appearance of premovement slow potentials in these cortices shown in the present report, as the potentials are considered to be due to massive long-lasting EPSP currents. However, this study did not seem to suggest marked activities in the supplementary motor area. Although no electrode was implanted directly into the supplementary motor area, the electrode in the dorsomedial part of the premotor cortex (e.g. electrode B in Figs. 3 and 6) could have picked up some influences of an electrical dipole generated in the supplementary motor area (see ref. 14). However, one should be cautious in this respect, since the method of the present study, particularly transcortical potentials (S-D), ignores the activities of stellate-type neurons, which induce only closed electrical fields in the cortex (see refs.
259
8 and 14). Premovement potentials of the supplementary motor area should be carefully investigated in monkeys in the future. The present study suggests that the premotor cortex (area 6) participates in the general and associated organization of motor function, e.g. the organization of movement patterns as proposed by Fulton 4, when compared with the motor cortex (area 4), which may play the main role in the more specialized motor function, i.e. the final common station for the execution of muscle contraction.
REFERENCES 1 Arezzo, J. and Vaughan, H. G., Jr., Intracortical sources and surface topography of the motor potential and somatosensory evoked potential in the monkey. In H. H. Kornhuber and L. Deecke, (Eds.), Motivation, Motor and Sensory Processes of the Brain: Electrical Potentials, Behaviour and Clinical Use, Progress in Brain Research, Vol. 54, Elsevier/North-Holland, Amsterdam, 1980, pp. 77-83. 2 Bucy, P. C., Electrical excitability and cyto-architecture of the premotor cortex in monkeys, Arch. Neurol. Psychiat., 30 (1933) 1205-1225. 3 Bucy, P. C., A comparative cytoarchitectonic study of the motor and premotor areas in the primate cortex, J. comp. Neurol., 62 (1935) 293-332. 4 Fulton, J. F., Physiology of the Nervous System, 3rd edn., Oxford University Press, New York, 1949, pp. 421-446. 5 Gemba, H., Hashimoto, S. and Sasaki, K., Slow potentials preceding self-paced hand movements in the parietal cortex of monkeys, Neurosci. Lett., 15 (1979) 87-92. 6 Hashimoto, S., Gemba, H. and Sasaki, K., Analysis of slow cortical potentials preceding selfpaced hand movements in the monkey, Exp. Neurol., 65 (1979) 218-229. 7 Hashimoto, S., Gemba, H. and Sasaki, K., Premovement slow cortical potentials and required muscle force in self-paced hand movements in the monkey, Brain Research, 197 (1980) 415-423. 8 Lorente de N6, R., A study of nerve physiology, Stud. Rockefeller Inst. reed. Res., 132 (1947) Ch. 16. 9 Pieper, C. F., Goldring, S., Jenny, A. B. and McMahon, J. P., Comparative study of cerebral cortical potentials associated with voluntary movements in monkey and man, Electroenceph. clin. Neurophysiol., 48 (1980) 266-292. 10 Roland, P. E., Larsen, B., Lassen, N. A. and Skinhoj, E., Supplementary motor area and other cortical areas in organization of voluntary movements in man, J. Neurophysiol., 43 (1980) 118136. 11 Roland, P. E., Skinhoj, E., Lassen, N. A. and Larsen, B., Different cortical areas in man in organization of voluntary movements in extrapersonal space, J. Neurophysiol., 43 (1980) 137150. 12 Sasaki, K., Electrophysiological studies on the cerebellothalamocortical projections, Appl. Neurophysiol., 39 (1976/77) 239-250. 13 Sasaki, K., Gemba, H., Hashimoto, S. and Mizuno, N., Influences of cerebellar hemispherectomy on slow potentials in the motor cortex preceding self-paced hand movements in the monkey, Neurosci. Lett., 15 (1979) 23-28. 14 Sasaki, K., Staunton, H. P. and Dieckmann, G., Characteristic features of augmenting and recruiting responses in the cerebral cortex, Exp. Neurol., 26 (1970) 369-392. 15 Vaughan, H. G., Jr., Costa, L. D. and Ritter, W., Topography of the human motor potential, Electroenceph. clin. Neurophysiol., 25 (1968) 1-10. 16 Woolsey, C. N., Settlage, P. H., Meyer, D. R., Sencer, W., Pinto Hamuy, T. and Travis, A. M., Patterns of localization in precentral and 'supplementary' motor areas and their relation to the concept of a premotor area, Res. Publ. Ass. nerv. ment. Dis., 30 (1951) 238-264.
247
D I S T R I B U T I O N OF SLOW C O R T I C A L P O T E N T I A L S P R E C E D I N G SELFPACED HAND AND HINDLIMB MOVEMENTS IN THE PREMOTOR AND M O T O R A R E A S OF M O N K E Y S
SHUJI HASHIMOTO, HISAE GEMBA and KAZUO SASAKI* Department of Physiology, Institute for Brain Research, Faculty of Medicine, Kyoto University, 606 Kyoto (Japan)
(Accepted March 12th, 1981) Key words: premovement slow cortical potential - - hand movement - - hindlimb movement - premotor cortex - - motor cortex - - monkey
SUMMARY Surface negative-depth positive, slowly increasing potentials prior to self-paced hand and hindlimb movements were recorded in the dorsal aspect of the motor and premotor cortices with chronically implanted electrodes. It was shown that the potentials were recorded in the contralateral forelimb m o t o r area prior to hand movements but were hardly seen in the hindlimb motor area. On hindlimb movements, the contralateral hindlimb motor area showed the premovement potentials, whereas the forelimb m o t o r area revealed little or no prernovement potentials. The contralateral premotor cortex was shown to induce the premovement potentials in its wider areas and participate in both of hand and hindlimb movements in a similar fashion, with predominances in its dorsolateral portion for hand movements and in its dorsomedial portion for hindlimb movements respectively. In the hemisphere ipsilateral to the moving hand, the relatively large premovement slow potentials emerged frequently also in the premotor cortex, whereas only the small potential was obtained from the forelimb motor area. These results suggest that the premotor cortex (area 6) participates in the more general and associative organization of motor function than the motor cortex (area 4) which represents the specialized role in the motor performance.
* To whom correspondence should be addressed. 0006-8993/81/0000-0000/$02.50 © Elsevier/North-Holland Biomedical Press
248 INTRODUCTION We reported that slowly increasing cortical potentials preceding self-paced hand movements were recorded in the lateral portion of dorsal aspect of the premotor cortex and in the forelimb areas of the motor and somatosensory cortices contralateral to the moving hand of monkeys 5,6. The similar distribution of premovement potentials (motor potentials) was recently reported in monkeys and human subjects 9. Such premovement slow potentials started about one second before the hand movement elevating a lever and showed the transcortical reversal of electrical polarity; negativity at the cortical surface and positivity at a cortical level of 2.5-3 mm below the surface. They were interpreted to be due mainly to the currents of EPSPs generated in superficial parts of apical dendrites of cortical pyramidal neurones via certain thalamocortical (T-C) projections, and were supposed to participate in initiating and/or controlling the movement 8. It was further suggested that the slow potential in the motor cortex is at least partly induced by the activity of the hemispherical part of the cerebellum through the superficial T-C projections 13 (see ref. 12). The slow potentials in the premotor, motor and somatosensory cortices were also shown to relate closely to the control of required muscle force in expectation 7. The previous papers mainly dealt with the potentials in the lateral portion of dorsal aspect of the premotor cortex and in the forelimb area of the motor cortex, when monkeys were performing the hand movement elevating a lever. The present study was devoted to explore the range of responsive cortical areas in these cortices, particularly whether or not the dorsomedial portion of the premotor cortex and the hindlimb area of the motor cortex would be included as responsive areas for hand movements. It will be shown in this report that the hindlimb area of the motor cortex hardly shows premovement slow potentials on the lever-elevating hand movement. On the contrary, the premotor cortex in its wider area is active on the hand movement, and even the premotor cortex ipsilateral to the moving hand is also involved on the movement. The cortical potentials prior to self-paced hindlimb movements will also be briefly reported in order to collate such distribution of premovement cortical potentials for hand movements with that for hindlimb movements. MATERIALS AND METHODS Experiments were carried out on seven adult monkeys (Macacafuscata). All of them were trained to perform lever-elevating hand movements and three among them were trained to perform also hindlimb movements pressing a pedal at a time interval of more than two seconds without any instructing stimuli (see Fig. 1). A load opposing lever elevation was sometimes given in a series of daily experiments by a cablepulley system so that the premovement slow potentials prior to self-paced hand movements increase in their amplitude (see ref. 7). Elevating a lever or pressing a pedal by the correct movement delivered a small amount of juice about 600 ms after the movement. Monkeys usually performed the movements several hundred times in a series on each day.
249
EMG
S
D
' '\'
nn • ,
>
;,
BONE
'"' " { , .
DuRA
',,,:
CORTEX
Fig. 1. Diagrams illustrating forelimb and hindlimb movements, and the electrode array implanted in the cortex. Upper diagram: lever-elevating hand movement with electromyogram (EMG) record. M.S., microswitch generating the timing pulse. The lever is prevented from falling down by stopper (STOP.). Lower diagram, right: pedal-pressing hindlimb movement with EMG record from gastrocnemius-soleus muscles. Lower diagram, left: the recording electrode implanted in the cortex. The surface needle (S) placed on the cortical surface and the depth one (D) at about 3 mm below the surface.
Recording and data processing Most of recording procedures were previously described in detail 6. The surgery for implanting the electrodes in various parts of the cerebral cortices on both sides was performed under anesthesia with ketamine hydrochloride (7 mg/kg, i.m., every 1-2 h). Each electrode array consisted of two silver needles (0.25 m m in diameter) attached together, of which the tips were arranged so that they were on the cortical surface (S) and at a depth of 2.5-3 m m in the cortex (D) respectively when implanted (see Fig. 1, lower diagram to the left). The silver needles were insulated except for the chloridized tip. The surface and deep cortical layer potentials were led simultaneously from the needles respectively in referring to the indifferent electrode, which was composed of linked two Ag-AgCI electrodes buried in the bone behind the ear on both sides. They were respectively amplified by AC amplifiers with time constant of 2.0 s, and then stored in cassette tapes through multichannel data recorders together with the timing pulse generated at the moment of microswitch closure by lever-elevating or pedalpressing movements. 'Electrooculogram (EOG)' in this study was taken from the electrode implanted in the rostral edge of frontal bone above the orbita on each side in reference to the indifferent electrode mentioned above, and amplified by an AC amplifier with time constant of 2.0 s. Electromyogram (EMG) was led bipolarly from silver plate electrodes attached to the skin over wrist extensor muscles on the hand movement or over plantar flexor (gastrocnemius-soleus) muscles on the hindlimb movement and amplified with an AC amplifier, of which the time constant was 0.01 s. ' E O G ' and E M G were stored likewise in cassette tapes.
250
o
200
,I,
EOG
~ ~00pV EOG
-
-
~'~._._.--
"
S-D C s
D
$-: "
~
/
Z
D
t
+
$-D
EMG
1
t
L_
.
Fig. 2. Cortical potentials in the motor cortex on the side contralateral to the moving hand. The number of samples for averaging was 100. The left and central columns show simultaneous records from the hindlimb area (A) and the forelimb areas (B and C) of the motor cortex, together with EOG and E M G activities in one and the same monkey. The left (0) and central (200) columns present the potentials without and with the additional load of 200 g, respectively, being obtained from a series on the same day. Simultaneous records of cortical potentials from the forelimb areas (D and C) of the motor cortex in another monkey are shown in the right column. The surface (S) and depth (D) cortical potentials and the electronic subtraction of the depth record from the surface record (S-D) are shown in respective rows, together with EOG and EMG. The instant of microswitch closure by hand movements is indicated by upper and lower arrows in this and all subsequent figures. 500/iV calibration is for EOG potentials and 50/zV for the cortical potentials. Negativity is represented by upward deflection. 500 ms time scale is applied to all records. S.C., sulcus centralis. Note the premovement slow potential is not recorded in the hindlimb area of the motor cortex (A).
251 Cortical and 'EOG' potentials were usually averaged 100 times, or occasionally 20 times, in the period between 1.5 s before and 0.5 s after the timing pulse. The surface cortical potentials were electrically subtracted by the deep cortical layer potentials and the results (S-D potentials) served as an indication for true transcortical potentials without marked contamination of EOGs and of some other potentials generated in sites remote from the recording electrodes (see ref. 6). EMGs were passed through a full-wave rectifier circuit and then averaged likewise. After such daily experiments for several months, cortical sites of implanted electrodes in the cerebral cortex were checked at autopsy. RESULTS This report deals with the cortical potentials preceding movements, particularly surface negative-depth positive potentials that reach their summit 50-200 ms before movements, and will not treat the potentials thereafter (see ref. 6). The averaged cortical potentials in the motor cortex contralateral to the moving hand are shown in Fig. 2, which were obtained from two different monkeys performing hand movements. The left (0) and central (200) columns exemplify the slow potentials obtained from one and the same monkey performing lever-elevating hand movements, respectively without additional load and with additional 200 g-load opposing the lever elevation, in a series of the same day. The potentials in the right column were taken from another monkey. The potentials were recorded from the hindlimb area of the motor cortex (A) and the forelimb areas of the motor cortex (B and C in one monkey, D and C in another monkey). As shown in Fig. 2, surface negative-depth positive, slowly increasing potentials prior to the movement were recorded in the forelimb areas of the motor cortex (B, C and D) in the two monkeys, but they were not appreciably seen in the hindlimb area, except small surface positive-depth negative potentials just before the movements that would not be dealt with as mentioned above. Configurations of the potentials were similar within one subject, although their sizes differed from each other. The slow potentials in C were larger than those in B in the one monkey as shown in the left and central columns, and, in another monkey, the potentials in C, which was nearer to the central sulcus than D, were larger than those in D. This would correspond to the recent finding that antecedent cortical activity for wrist extension movements is maximal in the posterior margin of the precentral gyrus 1. With 200 g-load, the slow potentials in the forelimb area of the motor cortex were increased in accordance with the required muscle force to elevate the lever, as seen in the left and central columns (see also ref. 7). Even under such conditions, surface negative-depth positive slow potentials before movement appeared barely in the hindlimb area of the motor cortex (A, central column). Small negative shifts without transcortical reversal of electrical polarity as in S and D rows of A in the central column can be considered to be due to passive spread of some potentials generated in some remote sites from the recording electrode 14. No significant potential changes before hand movements were recorded in more medial sites to A (not shown in this figure). The results reveal that the premovement, surface negative-depth positive slow
252 potentials are recorded in the forelimb area of the motor cortex, but not in the hindlimb area in association with the lever-elevating hand movement. It was usually observed that the rostral part of the forelimb motor area produced a little smaller premovement potential than the caudal part near the central sulcus, as shown in the right column of Fig. 2. The potential in the rostral part was often even smaller than the largest one in the premotor cortex so that a small potential was interposed between the large potentials in the premotor cortex and in the caudal part of the motor cortex. Fig. 3 presents slow potentials preceding self-paced hand movements in the dorsal aspect of the contralateral premotor cortex (A, B, C and D), together with that in the forelimb area of the motor cortex (E). All records in the premotor cortex showed the surface negative-depth positive slow potentials, although the amplitude was larger in the lateral portion (C and D) than in the medial (A and B). Such relatively diffuse appearance of the potentials in the premotor cortex was in contrast to the distribution of premovement slow potentials in the motor cortex; such potentials could not be recorded in the dorsomedial portion of the motor cortex (hindlimb area). This figure also shows, as reported previously 7, that the contour of the premovement slow potentials in the premotor cortex differs from that in the motor cortex. The
C
E
i25 IJV
S-D ~ Fig. 3. Specimen records of cortical potentials in the premotor cortex (A-D) on the side contralateral to the moving hand, together with those in the forelimb area of motor cortex (E). The recording sites are indicated in the inset brain diagram with respective letters. All records were taken in one and the same monkey and all except B were simultaneously recorded. Averaged of 200 samples. 25/*V calibration and 500 ms time scale are applied to all cortical potentials. Abbreviations as in Fig. 2.
253 potentials in the p r e m o t o r cortex started earlier, usually by a few hundred ms, and reached the summit a b o u t 100-200 ms earlier than those in the m o t o r cortex. The premovement slow potentials in the hemisphere ipsilateral to the moving h a n d are exemplified in the left c o l u m n o f Fig. 4. F o r collation, the slow potentials in the contralateral hemisphere o f the s a m e m o n k e y are also shown in the right column o f Fig. 4, being obtained in a series o f the same day. Surface negative-depth positive, slowly increasing potentials with fairly large amplitudes were encountered in the ipsilateral p r e m o t o r cortex (A), while those in the ipsilateral m o t o r cortex (B) were considerably smaller in this case than in the contralateral m o t o r cortex. It was
EOG
A
s
S-
S-b
EMG
Fig. 4. Slow potentials recorded from the premotor (A) and motor (B) cortices on the side ipsilateral and contralateral to the moving hand. Average of 100 samples. All records were obtained from the same monkey, and in a series of the same day. The potentials in the premotor and motor cortices on each side were recorded simultaneously. Abbreviations and voltage calibrations are as in Fig. 2. Time scale is 250 ms.
254 characteristic of the premovement slow potentials on the lpsilateral side that the potential in the premotor cortex was relatively larger than in the motor cortex, although the slow potentials on the ipsilateral side were in general rather variable in size. In contrast, peak amplitude of the premovement slow potential in the premotor cortex was always smaller than that of the motor cortex on the contralateral side in all monkeys tested, as shown in the right column of Fig. 4 (see ref. 5). The upper part of Fig. 5 (hindlimb) shows potentials preceding self-paced hindlimb movements in the motor cortex on the side contralateral to the movements. As shown in A (recorded in the hindlimb area of the motor cortex), surface negative-depth positive potentials, although much shorter in duration than those on hand movements, were recorded before the movements. On the other hand, the potentials obtained from the electrode implanted in B (forelimb area of the motor cortex) showed no obvious transcortical reversal of electrical polarity on the hindlimb movement. Traces of S-D in the lower part of Fig. 5 (forelimb) reveal potentials preceding hand movements obtained from A and B with the same electrodes in the same monkey as mentioned above. As has been described, the pre-
S
-
~
/
/
50 EMG FOREL1RB
pV
A___ t
-,,) Fig. 5. Records obtained on hindlimb movements (upper part) and on hand movements (lower parts). Upper: potentials in the hindlimb areas (A and C) and the forelimb area (B) of the motor cortex contralateral to the moving hindlimb are shown in one and the same monkey. A and B were on the same but C on the opposite side of hemispheres. The hindlimbs pressing a pedal were respectively on the contralateral side to the hemispheres under recording. The traces in A and B were concurrently recorded. Lower: potentials in the same hindlimb area (A) and forelimb area (B) of the motor cortex contralateral to the moving hand. They were simultaneously recorded with the same electrodes (A and B) as in the upper part, being averaged for 100 times. Abbreviations and calibrations as in Fig. 2.
255 movement slow potential was encountered in the forelimb area of the motor cortex (B) with hand movements, while such a potential was not recorded in the hindlimb m o t o r area (the lower part in column A). Records in C are the slow cortical potentials in the hindlimb motor area on the contralateral hindlimb movements in another hemisphere, presenting more predominant surface negative-depth positive potentials with relatively short duration. These results show that only the hindlimb area within the motor cortex becomes active to produce the potentials prior to hindlimb movements. The contours of premovement potentials in the forelimb and hindlimb areas on respective movements are different from each other. Surface negative-depth positive potentials are much shorter for the hindlimb activity than the forelimb one. The physiological significance of such difference can not be explained at present. The surface negative~iepth positive potentials in the hindlimb motor area were occasionally preceded by slowly increasing, surface positive-depth negative potentials, which lasted for nearly 1 s, as seen in A and C of Fig. 5. They were very small, if present, in size, and were inconstantly observed even in the same recording site on different days. It is not certain whether the slow potentials would be active responses in this area of the cortex or not. Further studies will be required to clarify the nature of the potentials.
B
s-D C
s-B
Fig. 6. Slow potentials recorded on contralateral hindlimb movements with the same electrodes in the same monkey as in Fig. 3. All records except in C were obtained concurrently; 200 samples were averaged. Calibrations as in Fig. 3; abbreviations as in Fig. 2.
256 In contrast to the motor cortex, surface negative-depth positive slow potentials could be recorded in wide portions of the premotor cortex prior to hindlimb movements. Fig. 6 gives the potentials in the premotor cortex contralateral to the moving hindlimb in the same monkeys as Fig. 3. Premovement potentials appeared, in the premotor cortex, with a relatively large amplitude and a considerable duration on hindlimb movements. Their peaks emerged at the almost same moment before the movement, as in the case of forelimb movement shown in Fig. 3. The sizes of the potentials in the sites of A and D were occasionally smaller than in B, although the sizes in A, B and D were nearly same in this case as revealed by S-D traces. The results indicate that the premovement slow potentials in the premotor cortex appear in a similar fashion on both hindlimb and forelimb movements, and that preponderance shifts to the medial portion of dorsal aspect of the premotor cortex for hindlimb movements. In contrast, the forelimb motor area (E) hardly showed such transcortical reversal of electrical polarity as seen in the premotor cortex on hindlimb movements (cf. Fig. 3E). Fig. 7 summarizes the results of the present study. Prior to self-paced hand movements, the forelimb area of the motor cortex on the side contralateral to the moving hand became active and the hindlimb area remained inactive, whereas both of the lateral and medial portions of dorsal aspect of the premotor cortex revealed obvious premovement potentials with predominance in the lateral portion. In the ipsilateral hemisphere, relatively large potentials with transcortical reversal of polarity often appeared in the premotor cortex, while the forelimb area of the motor cortex gave merely small potentials. In association with hindlimb movements, the hindlimb area of the motor cortex contralateral to the movements showed marked premovement potentials, but the forelimb area revealed scarcely transcortical reversal of electrical polarity. The premotor cortex showed, prior to hindlimb movements, the slowly increasing potentials in both of the lateral and medial portions of its dorsal aspect with some predominance in the medial portion. DISCUSSION The present results revealed that the forelimb area of the motor cortex becomes active to produce the surface negative--depth positive, slow potentials before hand movements on the contralateral side, and that the hindlimb area of the motor cortex remains inactive on the movements. The hindlimb motor area induced the premovement slow potentials in association with hindlimb movements which were hardly observed in the forelimb motor area. Such somatotopical representation in the motor cortex was previously reported on slow potentials prior to voluntary movements recorded from human scalp 15. On the other hand, the premotor cortex on the side contralateral to movements was shown to participate in both of the forelimb and hindlimb movements in a similar fashion, although some quantitative difference was noted between the slow potentials preceding forelimb and hindlimb movements. Relatively large potentials emerged in the lateral portion of the dorsal aspect of the premotor cortex prior to hand movements and those in the medial portion before
257 hindlimb movements. The fairly large premovement slow potentials were frequently encountered also in the premotor cortex ipsilateral to the moving hand, whereas the forelimb area of the motor cortex in the ipsilateral hemisphere gave only small potentials prior to hand movements. It can be summarized as follow; as for the premovement slow potentials, the motor cortex has the somatotopical relationship to contracting muscles on the contralateral side, while the premotor cortex participates relatively in a diffuse manner in forelimb and hindlimb movements, and also bilater-
FORELIMB
I PS! L A T E R ~
HINDLIMB
CONTRA~
"" ~"
Fig. 7. Distribution of premovement slow cortical potentials in the dorsal aspect of the premotor and motor cortices. Upper and middle diagrams show the distribution in the contralateral and ipsilateral hemispheres on the forelimb movements. The lower diagram gives the distribution in the contralateral hemisphere on hindlimb movements. Large filled circles represent relatively large potential difference (S-D of usually more than 50/*V) with transcortical reversal of electrical polarity; small circles indicate small differences (usually less than 50 #V); asterisks show inactive points, s.a., sulcus arcuatus; s.c., suicus centralis; s.i., sulcus intraparietalis.
258 ally in the movements. The premovement slow potentials in the premotor and motor cortices showed the different time courses as mentioned in Results (see also ref. 7). These differences between the premotor and motor cortices may suggest that the premotor cortex is a different functional entity from the motor cortex for movement initiation and/or motor performance. Woolsey 16 placed the anterior boundary of 'precentral motor cortex' in the premotor cortex (area 6) by stimulation experiments; the axial musculature was reported to be represented in the caudal portion of area 6 (Bucy's excitable portion of area 62). However, the present results show that the rostral portion of dorsal aspect of the premotor cortex (A and C in Figs. 3 and 6), which is not involved in Woolsey's 'precentral motor cortex', has given the similar premovement slow potentials to those in the more caudal portion of area 6 (B and D in Figs. 3 and 6), which may be included in Woolsey's 'precentral motor cortex'. In addition, the contours of premovement slow potentials obtained from the electrodes within the motor cortex (area 4) were similar to one another, but differed from those in area 6. A smaller premovement potential was occasionally noted in the border area between areas 6 and 4 than premovement potentials in areas 6 and 4 as mentioned in Results. These data may suggest that the boundary for functional entities based on the premovement slow cortical potentials is more reasonably placed between areas 6 and 4, following the cytoarchitectonic studies 3, than on the anterior border of Woolsey's 'precentral motor cortex'. However, further careful examination should be made for the possibility of simultaneous activities in truncal and proximal limb muscles supporting the hand and hindlimb movement, although sampling E M G records from these muscles has so far revealed no marked activities. The dorsal aspect of the premotor cortex (area 6) represented the considerably wide responsive area of premovement slow potentials, while the motor cortex showed the fairly definitive somatotopical representation. Recently, measurement of regional blood flow in human subjects revealed that simple repeated finger movements were associated with the activation (increase of blood flow) of the contralateral forelimb motor area, and that the finger movement for a maze test activated the contralateral forelimb motor area and the bilateral premotor areas 11. Furthermore, programming a sequence of fast isolated finger movements and the maze test was reported to be accompanied with a marked increase of regional I blood flow in the supplementary motor areal0, H. The activation of the contralateral motor cortex and the bilateral premotor cortices might be related to the appearance of premovement slow potentials in these cortices shown in the present report, as the potentials are considered to be due to massive long-lasting EPSP currents. However, this study did not seem to suggest marked activities in the supplementary motor area. Although no electrode was implanted directly into the supplementary motor area, the electrode in the dorsomedial part of the premotor cortex (e.g. electrode B in Figs. 3 and 6) could have picked up some influences of an electrical dipole generated in the supplementary motor area (see ref. 14). However, one should be cautious in this respect, since the method of the present study, particularly transcortical potentials (S-D), ignores the activities of stellate-type neurons, which induce only closed electrical fields in the cortex (see refs.
259
8 and 14). Premovement potentials of the supplementary motor area should be carefully investigated in monkeys in the future. The present study suggests that the premotor cortex (area 6) participates in the general and associated organization of motor function, e.g. the organization of movement patterns as proposed by Fulton 4, when compared with the motor cortex (area 4), which may play the main role in the more specialized motor function, i.e. the final common station for the execution of muscle contraction.
REFERENCES 1 Arezzo, J. and Vaughan, H. G., Jr., Intracortical sources and surface topography of the motor potential and somatosensory evoked potential in the monkey. In H. H. Kornhuber and L. Deecke, (Eds.), Motivation, Motor and Sensory Processes of the Brain: Electrical Potentials, Behaviour and Clinical Use, Progress in Brain Research, Vol. 54, Elsevier/North-Holland, Amsterdam, 1980, pp. 77-83. 2 Bucy, P. C., Electrical excitability and cyto-architecture of the premotor cortex in monkeys, Arch. Neurol. Psychiat., 30 (1933) 1205-1225. 3 Bucy, P. C., A comparative cytoarchitectonic study of the motor and premotor areas in the primate cortex, J. comp. Neurol., 62 (1935) 293-332. 4 Fulton, J. F., Physiology of the Nervous System, 3rd edn., Oxford University Press, New York, 1949, pp. 421-446. 5 Gemba, H., Hashimoto, S. and Sasaki, K., Slow potentials preceding self-paced hand movements in the parietal cortex of monkeys, Neurosci. Lett., 15 (1979) 87-92. 6 Hashimoto, S., Gemba, H. and Sasaki, K., Analysis of slow cortical potentials preceding selfpaced hand movements in the monkey, Exp. Neurol., 65 (1979) 218-229. 7 Hashimoto, S., Gemba, H. and Sasaki, K., Premovement slow cortical potentials and required muscle force in self-paced hand movements in the monkey, Brain Research, 197 (1980) 415-423. 8 Lorente de N6, R., A study of nerve physiology, Stud. Rockefeller Inst. reed. Res., 132 (1947) Ch. 16. 9 Pieper, C. F., Goldring, S., Jenny, A. B. and McMahon, J. P., Comparative study of cerebral cortical potentials associated with voluntary movements in monkey and man, Electroenceph. clin. Neurophysiol., 48 (1980) 266-292. 10 Roland, P. E., Larsen, B., Lassen, N. A. and Skinhoj, E., Supplementary motor area and other cortical areas in organization of voluntary movements in man, J. Neurophysiol., 43 (1980) 118136. 11 Roland, P. E., Skinhoj, E., Lassen, N. A. and Larsen, B., Different cortical areas in man in organization of voluntary movements in extrapersonal space, J. Neurophysiol., 43 (1980) 137150. 12 Sasaki, K., Electrophysiological studies on the cerebellothalamocortical projections, Appl. Neurophysiol., 39 (1976/77) 239-250. 13 Sasaki, K., Gemba, H., Hashimoto, S. and Mizuno, N., Influences of cerebellar hemispherectomy on slow potentials in the motor cortex preceding self-paced hand movements in the monkey, Neurosci. Lett., 15 (1979) 23-28. 14 Sasaki, K., Staunton, H. P. and Dieckmann, G., Characteristic features of augmenting and recruiting responses in the cerebral cortex, Exp. Neurol., 26 (1970) 369-392. 15 Vaughan, H. G., Jr., Costa, L. D. and Ritter, W., Topography of the human motor potential, Electroenceph. clin. Neurophysiol., 25 (1968) 1-10. 16 Woolsey, C. N., Settlage, P. H., Meyer, D. R., Sencer, W., Pinto Hamuy, T. and Travis, A. M., Patterns of localization in precentral and 'supplementary' motor areas and their relation to the concept of a premotor area, Res. Publ. Ass. nerv. ment. Dis., 30 (1951) 238-264.