Brain Research Bulletin, Vol. 34. No. 6. pp. 587-593, 1994 Copyright 0 1994 Elsevier Science Ltd Printed in the USA. All rights reserved

Pergamon

036l-9230/94$6.00 + .oO

0361-9230(94)EOO56-6

BRIEF C~~~UNICATI~N

Reciprocal Activation and Coactivatio~ in Antagonistic Muscles During Rapid Goal-directed Movements YOSHIHIKO

YAMAZAKI,*’

TETSUO

OHKUWA,”

HIROSHI

ITOH,”

AND

MASATAKA

SUZUKI?

~~epa~rnent of Hearth and Physical Educut~o~, ~ag~yu Znstitute of Technology, Nagoya 466, Japan fllepartment of General Edu&at~~~,Kinjo Gakuin Universi~, Nagoya 463, Japan Received

26 July 1993; Accepted

24 February

1994

YAMAZAKI, Y., T. OHKUWA, H. ITOH AND M. SUZUKI. Reciprocal activation and coactivafion in antagonistic muscles movements. BRAIN RES BULL 34(6) 587-593, 1994.-Seven normal subjects performed elbow extensions as rapidly as possible from an initial position to a visually defined target at 36” in amplitude. In electromyograms, the reciprocal activation of the agonist and then ~tagonist bursts was always followed by simult~eous activation of the ~~gonistic muscles, i.e., coactivation. Instructions added fo perform extensions “as rapidly as possible” changed coactivation; the command to “strongly fix tbe upper arm at the target” increased coactivation, whereas “relax immediately after the start of movement” made coactivation almost disappear. However, basic features of reciprocal activation remained the same. Other instructions given also changed coactivation on initiation and termination, while reciprocal activation was relatively unaltered. When subjects were encouraged to “relax immediately after the start of movement, but fix the upper arm quickly after awning the target,” coactivation initiated shortly after reaching the target (< 200 ms). Following the instruction to *‘relax the upper arm quickly after attaining the target,” coactivation terminated rapidly after reaching the target (< 280 ms). The results show that instructions serve to change amplitude and timing of coactivation while keeping reciprocal activation relatively unaltered, suggesting that coactivation is controlled independently of reciprocal activation during rapid goal-directed movements. during rapid goal-directed

Motor control

Cocontraction

Electromyogram

Elbow extension

Human

ing the rapid movements. The exa~nat~on of the reciprocal activation and coactivation provides evidence to identify the dependence/independence of the commands during rapid movements. For these reasons, we gave subjects various instructions that induced changes in coactivation, and studied electromyograms (EMGs) to determine whether or not coactivation was independent of reciprocal activation during rapid goal-directed movements. In the course of the experiments, we found that coactivation changes markedly with the ins~ctions, whereas reciprocal activation remains essentially the same, suggesting that there are separate control mechanisms for reciprocal activation and coactivation.

IN rapid goal-directed movements about a single joint, the ago-

nist and antagonist muscles are reciprocally activated, as was initially described by Wachholder and Altenburger (25). Generally, the reciprocal activation is followed by tonic, postmovement activation of antagonistic muscles, i.e., coactivation. The reciprocal activation mainly reflects a movement command that initiates and terminates movements (10). Although the origin of the coactivation is still unclear, it has been suggested that coactivation represents a postural command that addresses a final posture (14) or specifies joint stiffness (11,15). There is evidence documenting coactivation of antagonistic muscles in many circumstances (4,11,17- 19,22-24). In spite of a considerable body of literature dealing with reciprocal activation, very few studies have been done on coactivation during rapid goal-directed movements. Moreover, it has not been determined in muscular activation whether or not the movement command and the postural command are independently released dur-

METHOD

Seven male subjects (ages: 1%44), with no known history of neurological disorders, gave informed consent to participate in

’ To whom requests for reprints should be addressed.

587

YAMAZAKI

588

the present experiments. All subjects had engaged in sports in which skillful movements of the dominant arm were required. Subjects were seated in a chair with their right arm (the dominant arm for all subjects) abducted 90” and with the forearm and hand placed in a molded splint with the forearm in a neutral position. The splint was fixed to a light aluminum manipulandum, to which a vertical shaft was connected to rotate horizontally. The elbow joint was positioned exactly above the vertical axis. In the first series of experiments, subjects learned to coactivate antagonistic muscles around the elbow joint while maintaining it at 90” (full elbow extension was defined as 180”) with the assistance of EMG visual feedback. The instruction, “fix the upper arm, ” was effective for teaching coactivation. After subjects learned to coactivate, we studied whether they were able to initiate coactivation and terminate preexerted coactivation quickly in response to a computer-produced beep. In the second series of experiments, subjects viewed a computer display, 1 m away, which showed the elbow angle as a movable line, and initial and final target levels as two stationary lines. Subjects had to align the elbow angle with the initial level of 63” at the beginning of each trial. At the sound of the beep, subjects made a single elbow extension from the initial level to a target angle of 99” (36” in amplitude). The beep lasted for 2 s, during which subjects maintained the target position or the commanded condition. Speed in responding to the beep was not important. In the first block, subjects were instructed to move to the target “as rapidly as possible, ” which was a common instruction throughout the second series of experiments (control condition). In subsequent blocks, subjects had an additional instruction that was intended to vary the amplitude and timing of coactivation. Additional instructions were “strongly fix the upper arm at the after the start of target” (fix condition), “relax immediately movement” (relax condition), “relax immediately after the start

ET AL.

of movement, but fix the upper arm quickly after attaining the target,” and “relax the upper arm quickly after attaining the target.” Some other instructions were tested, but the results are not included here. Before each new block, subjects were given the routine instruction and then the additional instruction. Subjects practiced about 10 times to become familiar with each new instruction. Each block consisted of 5- 11 movements. The second series of experiments was repeated on a different day. The latter experiments were used for the analysis. Angular acceleration of the elbow joint was measured by a strain-gauge accelerometer fixed to the manipulandum 40 cm from the rotational shaft. The elbow angle was measured by a potentiometer connected to the lower end of the shaft beneath the elbow. EMGs were recorded from the lateral head of the triceps brachii muscle (TL) and the biceps brachii muscle (BB), using surface electrodes (Ag/AgCl disposable electrode). A quasi-monopolar recording was utilized: one electrode was placed on the skin above the approximate motor point region of each muscle with another electrode 5 cm distally. This recording method differed from the one commonly used [e.g., Basmajian and DeLuca (l)]. In our previous paper, however, we suggested that a monopolar electrode arrangement was more appropriate than a bipolar one for detecting the repetitive, near-synchronous recruitment of many motor units in rapid contractions (26,27). Therefore, we chose the quasi-monopolar recording. EMGs were preamplified near the electrodes and then further amplified, with the overall band width ranging from 7 to 500 Hz. Crosstalk between the triceps and biceps EMGs was tested by rapid alternating movements of the elbow joint. In pilot studies, we recorded the surface EMGs from the triceps longus and brachioradialis muscles in addition to the TL and BB muscles. Because the triceps longus and brachioradialis EMGs were respectively represented by the

Relax

FIG. 1. Mechanical and EMG recordings during rapid elbow extension movements under three instructions. Mechanical recordings included angle, velocity, and acceleration of the elbow joint. EMGs were recorded from antagonist biceps and agonist triceps lateralis muscles. In A, the subject (sl) was instructed to move “as rapidly as possible” to a visually defined target at 36” in amplitude which is shown as dotted lines (control condition). An additional instruction was given in B and C. The additional instruction was “strongly fix the upper arm at the target” in B (fix condition), “relax immediately after the start of movement” in C (relax condition).

COACTIVATION

IN RAPID

589

MOVEMENTS

Angle J

Biceps

FIG. 2. Angle, acceleration,

and EMGs during rapid movements under three instructions. Figures (A-F) respectively correspond to recordings of different subjects (s2-~7). In each figure, single recordings under control, fix, and relax conditions are shown. Mechanical recordings shown as broken (control), thin (fix), and thick (relax) lines, are superimposed. EMGs of both antagonist biceps and agonist triceps lateralis muscles were fullwave rectified.

TL and BB EMGs, only the TL and BB EMGs were recorded in the main experiments. All data were stored on a hard disk through a micr~omputer after I2 Bit A/D conversion at a sampling rate of 1 kHz. Angular velocity of the elbow joint was calculated by integrating the acceleration. The peak value of the angular velocity was determined, and was shown as movement velocity, which was indicated by mean & SD (radfs). The onset interval between TL and BB EMGs was measured. Reciprocal activation was quantified by separately integrating TL and BB EMGs after fullwave rectification. EMGs were integrated from the onset of agonist TL to the end of the initial acceleration of the arm. With antagonist BB, EMGs were integrated from its onset to the end of the deceleration, which occurred immediately after the initial acceleration. Coactivation was also quantified after full-wave ratification, in which TL and BB EMGs were separately integrated for 200 ms from 100 ms after the end of the deceleration. The statistical difference between data was examined by Student’s r-test. Statistical significance was put at the 5% level.

For brevity’s sake, results of a single subject (sl) will be presented as a typical example, and other subjects had the same results unless stated otherwise. RESULTS

Reciprocal Activation

and Coactivation

of TL and BB Muscles

Figure 1 shows angle, velocity, acceleration, and EMGs of BB and TL muscles during rapid goal-directed movements performed under varying instructions. In Fig. IA, the subject (sl) performed an elbow extension movement from the initial position to a 36” target “as rapidly as possible” without any additional instructions (control condition). The EMGs of the antagonistic muscles showed phasic, reciprocal activation; the first burst of activity in the agonist TL (AGl) was followed by the antagonist BB burst (ANT), then followed by the second agonist burst (AG2). The initial part of AG 1 and ANT showed an EMG volley that was characterized by a slow wave with initial negativity and ensuing positivity (26,27). ANT started with a certain delay after

590

YAMAZAIU

Oi

s7

Subject FIG. 3. Movement velocity. Peak angular velocity under control, fix, and relax conditions are shown as movement velocity by mean 2 SD (n = 5) for all subjects (sl-~7). AGI . Hereafter, we refer to reciprocal activation as a set of AG 1 and ANT, in which AG2 was excluded because it was often inseparable from the subsequent coactivation described next. After the reciprocal activation, TL and BB muscles showed tonic, simultaneous activation (i.e., coactivation), which was accompanied by some activation fluctuations in each muscle. The coactivation usually lasted for several seconds, then waned. Figure 2 shows angle, acceleration, and EMGs of the other subjects (s2s7), in which mechanical recordings in control, fix, and relax conditions were superimposed, and EMGs were shown after fullwave rectification. The other subjects also showed both the reciprocal activation and the coactivation in the control condition. Figures 3 and 4, respectively, show the movement velocity and the onset interval between AGI and ANT under control, fix, and relax conditions. Figure 5 shows the relationship in quantified EMGs between reciprocal activation and coactivation. In Fig. 5A, the quantity of agonist reciprocal activation (AGI quantity) is plotted against agonist coactivation quantity; in Fig. 5B, the quantity of antagonist reciprocal activation (ANT quantity) is plotted against antagonist coactivation quantity. Figures 3, 4, and 5 are referred to in the following. Coactivation

Changed ~ndepe~de~t~~

ofRecipmcal Activation

When instructed to “fix the upper arm” while maintaining the joint angle at 90”. subjects could easily coactivate TL and BB muscles. They could initiate or terminate the coactivation quickly in response to the beep, as well as vary the rate of the coactivation rise at initiation or of the fall at termination. They could also vary the coactivation even during rapid goal-directed movements. In Fig. 1B (fix condition), the subject was instructed to “strongly fix the upper arm at the target” in addition to the routine instruction, “as rapidly as possible.” Coactivation markedly increased in comparison with that foilowing a routine instruction (Fig. IA). In contrast, reciprocal activation preserved the basic features: AGl was followed by ANT with a certain delay after the AG 1 onset. The onset interval between AGl and ANT was almost preserved (Fig. 4, sl). In EMG quantification, reciprocal activation was relatively unaffected in comparison with the marked increase in coactivation (cf., control and fix conditions in Fig. SA and B, top figures for sl). Moreover, there was no difference in the movement ve-

ET AL.

locity (cf., control and fix conditions in Fig. 3, sl). Additionally, the distinction between AGl and AG2, and between AG2 and agonist coac!tivation, and between ANT and antagonist coactivation became less clear under the fix condition, which made it difficult to determine changes in timing of AGl, AG2, and ANT. The other subjects showed the same results (Figs. 2-5). These findings indicate that coactivation increases independently of reciprocal activation. In Fig. 1C (relax condition), the subject was instructed to “relax immediately after the start of movement,” in addition to being given the routine instruction. In comparison with the routine instruction (Fig. 1A), the basic features of reciprocal activation were preserved, in which both AGl and ANT remained as a burst activity with a certain delay between them (but the delay was slightly increased as shown Fig. 4, sl). By contrast, coactivation became negligible. EMG quantification confirmed the marked decrease in coactivation (cf., control and relax conditions in Fig. 5A and B, top figures for sl), although reciprocal activation also decreased in amplitude. The AGl duration was approximately preserved. The onset of AG2 delayed and the amplitude of AG2 decreased. The other subjects showed the same results with a few exceptions (Figs. 2-5). There was a decrease in movement velocity (Fig. 3, sl). For all subjects, movement velocity was 5.19.8 rad/s (median: 7.1 r&/s), against 7.4- 10.4 rad/s (median: 8.2 rad/s) with the routine instruction (Fig. 3). The decrease in movement velocity might have caused coactivation to be negligible. However, changes in coactivation were greater than those in reciprocal activation, especially in AGl; when EMG quantity in the relax condition was compared to that in the control one, AG 1 quantity was 72.4 c 13.4 (%), while agonist coactivation quantity was 14.8 -t- 15.9 (%), and ANT quantity was 39.0 t: 19.1 (%), while antagonist coactivation quantity was 11.8 + 13.8 (%). Furthermore, comparison of the relax condition with the fix condition (Fig. 5) reveals a marked decrease in coactivation despite relatively unaltered reciprocal activation. The greater changes in coactivation are due to the instructional manipulation rather than the decrease in movement velocity, because reciprocal activation, indicative of muscle activation which brings about force to drive and stop the limb, should reduce more than coactivation. These results show that coactivation decreases relatively independent of reciprocal activation.

k

120

T 0

2

ms

0 5

go.

= 608: z 8 30TJ L Qt goJ._,......

fdontro

zP g I f

sl

s2

f

s3

s4

%f

IIs6

s7

Subject FIG. 4. Onset interval between agonist triceps and antagonist biceps EMGs. The interval under control, fix, and relax conditions are shown by mean -t SD fn = 5) for all subjects (sl -s7).

IN RAPID

COACI’IVATION

100

-I

591

MOVEMENTS

I Control

Sl

Sl

s2

s2

$ Relax Fix r

n

150

g

100

---+

(E Control

1

Fix

s4

$

ss

a

O-’

150 loo

50 i

s5

ConJrol z IERelax

s6

s6

150,

lO

s7

Fix Relax

~ 50+ . 0

*Connvrl * 1

100

.

,

200

.

I

.

,

400

.

I

,

I

1

96 6oo

Agonist coactivation quantity

Antagonist

coactivation

&ant&y

FIG. 5. Quan@ied EMGs of

reciprocal activation and coactivation. EMG quantities of each subject (sl-s7) under control, fix, and relax conditions are shown. (A) Quanti~ of agonist reciprocal activation (A01 quantity) vs. agonist eoactivation quantity. (B) Quantity of antagonist reciprocal activation (ANT quantity) vs. antagonist coactivation quantity. All plots are shown as percentage of EMG quantity in the control condition. Mean r: SD (n = 5) is shown for all subjects (sl-~7).

Varying Onset and Termination of&activation With Instructions Independent of Reciprocal Activation In Fig. 6A, the subject was told to “relax immediately after the start of movement, but fix the upper arm quickly after attaining the target” in addition to the routine instruction. Shortly after a cessation of reciprocal activation in both ‘IL and BB muscles, coactivation initiated. Reciprocal activation remained relatively unchanged (cf. Fig. 1C). The other two subjects showed a similar result (Fig. 6C: s2, Fig. 6E: ~3). Subjects could initiate coactivation less than 200 ms after reaching the target position. In Fig. 6B, the subject was instructed to ‘Wax the upper arm quickly after attaining the target” in addition to the routine instruction. Coactivation rapidly decreased, whereas reciprocal activation remained the same when compared with the case for a routine instruction (Fig. IA). Recordings of two subjects (s2 and s3) were respectively shown in Fig. 6D and F. Coactivation di-

minished less than 280 ms after attaining the target position for all subjects. DISCUSSION The present experiments revealed that in rapid target-directed movements, coactivation of antagonistic muscles changed with the instructions, whereas the essential features of reciprocal activation were preserved. Moreover, the onset and end of coactivation varied relatively without affecting reciprocal activation. This decoupling between the coactivation and the reciprocal activation contrasts with the view that both activations are controlled as a unit through a motor control system. Therefore, our results suggest that, at a command level, coactivation is independent of reciprocal activation during rapid target-directed movements. In human motor unit recordings, independent control of reciprocal activation and coactivation has been suggested (5) and

592

YAMAZAKI

ET AL

Angle

A

and EMGs during rapid movements under two instrucitons. Subjects were instructed to “relax immediately after the start of movement, but fix the upper arm quickly after attaining the target” in A (sl), C (s2), and E (s3), and “relax the upper arm quickly after attaining the target” in B (sl), D (sZ), and F (s3), in addition to the common instruction to move “ as rapidly FIG. 6. Angle, acceleration,

as possible”

to a 36” target in amplitude.

corroborated (2,3). Furthermore, it was suggested that the central nervous system affects human spinal intememons differently during coactivation and reciprocal activation, because central effects on Ia inhibitory interneurons were different between the two activations (16). In experiments on monkeys, independent control was documented in corticospinal cells (9,X2) and cerebellar cortex celts (21). Simons and Zuniga (20) demonstrated that activation of the elbow flexion muscles differs when the forearm is spinated or pronated. Jongen et al. (13) showed that, during coactivation, the brachioradialis muscle exhibits different changes in activation quantity, whereas in isometric flexion, the EMG quantity of the two muscles is correlated. Thus, one may suppose that reciprocal activation and coactivation change in upper arm muscles under different instructions. In the present experiments, however, the activation of the elbow flexors was unaltered, because the forearm position was constant throughout the experiments. Furthermore, insofar as we recorded EMGs of the brachioradialis and triceps longus muscles in the pilot studies, the two muscles showed almost identical activation of BB and TL muscles, respectively, both in reciprocal activation and coactivation.

Feldman (6,7) and Levin et al. (15) suggested that the central nervous system inde~ndently releases reciprocal and coactivation commands during movement. Levin et al. (15) studied the torque-angle relationship when rapid wrist movements were unexpectedly perturbed by a spring-like opposing load or an assisting load, or when the movements met unexpected removal of these preexisting loads. Their subjects were instructed “not to correct errors arising from perturbation.” The positional shift in the relationship and the slope of the torque-angle relationship, which they regarded as indications of reciprocal and coactivation commands, respectively, showed independent variations when perturbed. Movement kinematics and EMGs of relevant muscles evidenced variable changes with the perturbation. The investigators concluded that reciprocal and coactivation commands were mutually independent, whereas neither kinematics nor EMG was representative of a central command. Setting aside theoretical implications depending on an equilibrium point hypothesis (S), our EMG findings reflect a more direct observation of a central command than the results based on the torque-angle relationship. The torque-angle relationship is a complex function of plural antagonistic muscle forces, each of which also includes

COACTIVATION

IN RAPID MOVEMENTS

593

complexity in force generation. Our results, however, are not influenced by the intricacies of the torque-angle relationship. The present results suggested that at a command level, coactivation operates inde~ndendy of recipti activation during rapid target-direct4 movements. However, we do not exclude possible interaction between them. At least, at a spinal level, two activation commands generating reciprocal activation and coactivation a~? likely to converge on a motoneuron pool and even the same motoneuron. Indeed, in the relax condition AGI and ANT EMGs showed a decrease in quantity (Fig. 5), which would cause the decrease in movement velocity. The merging of the two commands at the motoneuron pool may be a cause of such a decrease. Although

the essential pattern of reciprocal activation is generated independently of coactivation, the magnitude of reciprocal activation would be affected by the coactivation command. However, su~~sition of reciprocal activation on coactivation prevents el~~rnyo~~cal determination of the exact onset of coactivation as well as its amplitude variation. When coactivation is initiated and how its magnitude varies during the period dominated by reciprocal activation remain to be determined.

This work was partly supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan.

REFERENCES 1. Basmajian, J. V.; DeLuca, C. J. Muscle alive. Their functions revealed by el~~rnyo~phy, 5th edition. Baltimore: Witliams & Wilkins Co.; 1985. 2. Bremner, F. D.; Baker, .I. R.; Stephens, J. A. Correlation between the discharges of motor units recorded from the same and from different finger muscles in man. J. Physiol. (Land.) 432355-380; 1991. 3. Bremner, F. D.; Baker, J. R.; Stephens, J. A. Variation in the degree of synchronization exhibited by motor units lying in different finger muscles in man. J. Physiol. (Lend.) 432:381-399; 1991. 4. Carolan, B.; Cafarelli, E. Adaptations in coactivation after isometric resistance training. J. Appl. Physiol. 73:91 t-917; 1992. 5. DeLuca, C. J.; Mambrito, B. Voluntary control of motor units in human antagonist muscles: Coactivation and reciprocal activation. J. Physiol. (Lond.) 58525-542; 1987. 6. Feldman, A. G. Superposition of motor programs--I. Rhythmic forearm movements in man. Neuroscience 5:81-90; 1980. 7. Feldman, A. G. Su~~sition of motor programs-II. Rapid forearm flexion in man. Neuroscience 5:91-95; 1980. 8. Feldman, A. G. Once more on the equilibrium-point hypothesis (X model) for motor control. J. Motiv. Behav. l&17-54; 1986. 9. Fetz, E. E.; Cheney, P. D. Functional relations between primate motor cortex cells and muscles: Fixed and flexible. In: Bock, G.; O’Connor, M.; Marsh, J., eds. Motor areas of the cerebral cortex. Chichester: John Wiley & Sons Ltd.; 1987:98- 117. 10. Hallett, M.; Shahani, B. T.; Young, R. R. EMG analysis of stereotyped voluntary movements in man. J. Neurol. Neurosurg. Psychiatry 38:1154-1162; 1975. 11. Hogan, N. Adaptive control of mechanical impedance by coactivation of antagonist muscles. IEEE Trans. Auto. Control AC-29:681690; 1984. 12. Humphrey, D. R.; Reed, D. J. Separate cortical systems for control of joint movement and joint stiffness: Reciprocal activation and coactivation of antagonist muscles. In: Desmedt, J. E., ed. Motor control mechanisms in health and disease. New York: Raven Press; 1983:347-372. 13. Jongen, A. H.; Denier van der Gon, J. J.; Gielen, C. C. A. M. Inhomogeneous activation of motoneurone pools as revealed by cocontraction of antagonistic human arm muscles. Exp. Brain Res. 75:555-562; 1989.

14. Lestienne, F.; Polit, A.; Bizzi, E. Functional organization of the motor process underlying the transition from movement to posture. Brain Res. 230:121-131; 1981. 15. Levin, M. F.; Feldman, A. G.; Milner, T. E.; Lamarre, Y. Reciprocal and coactivation commands for fast wrist movements. Exp. Brain Res. 89:669-677; 1992. 16. Nielsen, J.; Kagamihara, Y. The regulation of disynaptic reciprocal la inhibition during co-contraction of antagonistic muscles in man. J. Physiol. (Lond.) 456:373-391; 1992. 17. Patton, N. J.; Mortensen, 0. A. An elec~omyographic study of reciprocai activity of muscles. Anat. Rec. 170:255-268; 1971. 18. Serres, S. J. D.; Milner, T. E. Wrist muscle activation patterns and stiffness associated with stable and unstable mechanical loads. Exp. Brain Res. 86:451-458; 1991. 19. Simmons, R. W.; Richardson, C. Peripheral regulation of stiffness during arm movements by coactivation of the antagonist muscles. Brain Res. 473:134- 140; 1988. 20. Simons, D. G.; Zuniga, E. N. Effect of wrist rotation on the XY plot of averaged biceps EMG and isometric tension. Am. J. Phys. Med. 49:253-256; 1970. 21. Smith, A. M. The coactivation of antagonist muscles. Can. J. Physiol. Phannacol. 59:733-747; 1981. 22. Solomonow, M.; Baratta, R.; Zhou, B. H.; D’Ambrosia, R. Electromyogram coactivation patterns of the elbow antagonist muscles during slow isokinetic movement. Exp. Neurol. 100:470477; 1988. 23. Tilney, F.; Pike, F. H. Muscular coordination experimentally studied in its relation to the cerebellum. Arch. Neurol. Psychiatry 13:289334; 1925. 24. Tyler, A. E.; Hutton, R. S. Was Sherrington right about co-contractions? Brain Res. 370:171-175; 1986. 25. Wachholder, K.; Altenburger, H. Beitrage zur Physiologie der willk&lichen Bewegung. X. Mitteilung. Ein~ibewegungen. Pflugers Arch. 214642-661; 1926. 26. Yamazaki, Y.; Suzuki, M.; Mano, T. Control of rapid elbow extension movement. Brain Res. Bull. 30: 1 l- 19; 1993. 27. Yamazaki, Y.; Suzuki, M.; Mano, T. An electromyographic volley at initiation of rapid isometric contractions of the elbow. Brain Res. Bull. 30:181-187; 1993.