Perceptrtal and Motor Skills, 1975, 40, 755-761. @ Perceptual and Motor Skills 1975
SHIFTS IN KINESTHESIS THROUGH TIME AND AFTER ACTIVE AND PASSIVE MOVEMENT1 BRIAN CRASKE AND MARTIN CRAWSHAW University o/ Southampton Summary.-The position sense of a stationary acm was investigated subsequent to an horizontally adductive movement with axis the shoulder joint. The right arm was the treated arm: it reached a test position actively, using minimal voluntary effort, or passively from each of 10 starting positions. The blindfolded S localized the index finger of the treated arm by attempting to touch it with the index finger of his left hand. The results indicate that subsequent to active movement the final position of a limb is more accurately known than a posicion resulting from passive movement. A second finding is that concomitant with both forms of limb placement there is a unidirectional drift of perceived limb position over trials.
It is now commonly accepted that receptors associated with the joints are responsible for the static position sense of limbs. The evidence in support of such a view has recently been marshalled by Skoglund (1973). A difficulty for this view may be inferred from a study by Collins (1971), who showed that when limb position was held constant and the muscles opposed a torque, subsequent judgments of the horizontal plane of the pronated hand were significantly different from control measures. Since the joint capsule is not rigid, such a procedure could produce an unusual yet adequate stimulus for the joint receptors: biased positional aftereffects due to this form of stimulation may still be mediated by the joint. If this does not provide adequate stimulus for the joint receptors, however, the experiment can be interpreted to suggest an involvement of muscle in position sense. Recent work on position sense in the eye (Skavenski, 1972) also suggests a muscular component. More dramatically, a study using vibration of the biceps tendon, a procedure which drives the muscle spindles, yielded estimates of arm position which were in error by as much as 8" (Goodwin, McCloskey, & Matthews, 1972). Although this situation is far from one which would occur under normal circumstances, it does suggest that positional information is being fed into the system. Its modus operandi, however, may be no more direct than a bias in positional analysis due to a massive afferent bombardment of movement information. Their work does imply, however, that structures other than joints might provide this information. If this is so, the normally functioning adult could be supplied with a positional signal arising from the joints, and additional positional information derived from non-joint sources. To this may be added a third source, positional information derived from monitoring the outflow of efference to the muscles, 'This work was supported by the Medical Research Council.
756
B. CRASKE
&
M. CRAWSHAW
for accurate movement and positioning can occur in the absence of afference from the joints (Lashley, 1917). Accuracy of determination of limb position has been the subject of a widely quoted study by Paillard and Brouchon ( 1968). They investigated the effects of active and passive placement, and active and passive maintenance of a position of a limb, on its localization by the contralateral limb. Their conclusion was that precision is significantly better with active rather than passive movement of the arm. This conclusion is not convincing, however. It appears that both the treated and the indicating arms were moved from the same starting position. It follows that in the active condition, S need only repeat the excursion of the target arm with the indicating arm. Thus judgments of excarsion were confounded with judgments of position. This methodological weakness casts some doubt upon their findings. Furthermore, the results of their second experiment, which manipulated elapsed time between placement and estimation, are at variance with their conclusions, for little difference between active and passive conditions was found. The existence of a difference in perceived limb position after active and passive movement is still unresolved. Possible changes in registered limb position are implied by a recent pilot experiment. In this, 12 Ss had to point to visual targets with both arms and without error feedback for 90 min. The trend of the errors of localization was consistent with the notion that with time the registered position of each limb was changing in opposite directions; to the left for the right arm and to the right for the left arm. In the experiment to be reported the accuracy of information available to S about the position of his stationary limb was investigated after passive transport of the limb and after active movement involving minimal effort in its execution. At the same time, the possibility of a systematic drift in kinesthesis was investigated.
METHOD Preliminary Tests Electromyographic recordings were taken from those muscles which are implicated in the performance of horizontal adduction of the arm. This was measured using 12 silver/silver chloride s ~ ~ r f a celectrodes e in conjunction with Cambridge electrode jelly. Raw EMG data were integrated by two methods, either by using a fixed epoch or by a fixed threshold. This established that passive movement of 0.1 rad/sec. produced EMG not measurably different from that of the resting and relaxed condition. These trials suggested that movements of this velociry would be suitable for the condition of passive movement in the experiment.
SHIFTS IN KINESTHESIS AFTER MOVEMENT
Subjects and Apparatas Sixteen male Ss were blindfolded, stripped to the waist and were seated in front of a two-tier platform. The lower tier supported a light-weight (365 gm.) free-running carriage. The carriage could be strapped to the arm which could be moved horizontally, either actively by S, or passively by a constant-speed electric motor controlled by E. The axis of movement was S s shoulder joint. The carriage could be altered in length to suit S and was designed to support his arm horizontally with the hand semi-pronated, index finger outstretched. Imrnediately above the hand and attached to the cradle was a horizontal scale, its zero coincident with the centre-line of the index finger. This was used to read off
FIG. 1. Subject pointing to contralateral limb which has been moved into position either actively or passively
errors of localization. The upper tier of the apparatus was of glass and was mounted immediately above the error scale on the carriage. A triangular section had been removed from the left side of the upper tier, S could therefore rest his left arm in a comfortable position on the lower tier, yet without obstruction he was able to raise it and then bring it down over the index finger of the right arm. A line was drawn on the nail of the left index finger to provide a datum. The upper tier prevented contact between the hands. A scale centered about S's right shoulder was marked in degrees on the lower tier, this enabled E to position the arm.
Procedure The design counterbalanced active and passive conditions. Trial excursions were made in whichever condition S was to perform first. If this was the active condition, then S adjusted the speed of his excursion to approximate closely the speed used throughout the passive condition (0.1 rad/sec.). During these practice excursions, S rehearsed stopping on E's instruction and indicating the position of his right index finger with his left. In the passive condition, S's arm was moved to one of 10 starting positions lying between 30°and 40' to the right of his median sagittal plane, thus prevent-
758
B. CRASKE
&
M. CRAWSHAW
ing S making judgments based on excursion. Each position was used twice in both halves of the experiment, the order of appearance being randomized. After 10 sec. S s arm was moved toward his left and stopped at one of 10 positions lying between 12' left and 5" right of his median sagittal plane. He was then instructed to point with his left hand. E noted rhe error, and the left arm was returned to its resting position. After 3 sec. S's right arm was returned at the same speed as before to the next starting position. In the active condition, the only difference in procedure was that the excursions were voluntary and were stopped on command. Twenty trials were given in the first condition and were followed, without break, by 20 trials in the second condition. RESULTS Only angular errors of pointing were measured, errors of depth being ignored. Errors were measured along the arc described by the right index finger. The data for 20 trials in both conditions for each of 16 Ss were subjected to a 3-way analysis of variance. The mean position indicated by all Ss for the active condition was 1'36' and for the passive condition, 2'57'; in both conditions to the left of his right index finger. The difference between these two means was significant ( F = 5.31, df = 1/15, p < 0.05 ). The differences between Ss were significant (F = 135.75, df = 15/285, P < 0.001), as was the Ss X conditions interaction ( F
TRIALS FIG. 2. Mean errors of localization
SHIFTS IN KINESTHESIS AFTER MOVEMENT
759
< 0.001). The effect of order of trials was such that there was a general increase in size of error toward the left as the experiment proceeded. This effect existed in both conditions and, for both conditions combined, was significant ( F = 2.59, df = 19/285, P < 0.01). The variances of the active and passive conditions were not significantly different ( t = 1.41, df = 1 5 ) and variance of the condition performed first was not significantly different from that performed second ( t = 0.3, d f = 15 ) . The pooled within-condition standard deviation was 5'07'. = 9.35, df = 15/285, p
DISCUSSION The order effect is such that Ss tended to make significantly greater errors of localization as the experiment ~ r o ~ r e s s e dThese . errors were of undershoot, i.e., the arms were separated, not crossed. As is shown in Fig. 2, this effect occurred with similar rate of increase over trials in both the active and passive conditions. This is consistent with the nonsignificant interaction of order X conditions. It seems that kinesthesis manifests an underlying lability when information about the accuracy of its operation is not available. That limb position sense is exceedingly labile is not in doubt; experiments with prisms (Harris, 1965; Craske, 1966) have shown that kinesthesis can be easily modified when vision and kinesthesis are discordant. The property of rapid modification of position sense in order to overcome prism-induced errors of reaching suggests that similar cross-modal checking of position information may be used to ensure veridicality in limb position sense in the circumstances of everyday life. Since the rate of drift in the present experiment was the same in both active and passive conditions, the significant main effect cannot be attributed to this factor. Concerning the main effect, the findings support the notion that there is a small but real superiority of localization subsequent to active movement, even under these stringent conditions when voluntary effort is very small, and the mechanical loading of the limbs minimal. This suggests that mechanisms other than those associated with the joint can influence the registered position of a stationary limb, unless even these minima1 conditions of muscular involvement can influence the joint receptors via muscles inserting at the joint. The significant interaction between Ss and conditions is an indication that some individuals utilize the information made available by active movement more than others. As previous writers have suggested, it appears that muscular involvement does lead to greater accuracy, but there could be an alternative explanation. It might be suggested that the effect of active involvement of muscles on performance is due to the increased attention that this entails. However, if inattention were the source of greater errors in the passive condition, they would presumably
760
B. CRASKE
(i:
M. CRAWSHAW
be randomly distributed. Hence, the variance of localizations made in that condition would increase, but there would be no great difference in mean position between conditions. On the contrary, the results show that there was a significant difference in mean position, but no significant difference in variance. Nevertheless, attention may well be an important factor in other studies directed at differences between active and passive conditions. The issue of the necessity for active movement in adaptation to displaced vision is a case in point (Held & Hein, 1958; Held & Freedman, 1963). Although it has been shown that such adaptation calz take place in the absence of active movement (Howard, Craske, & Templeton, 1965; Kravitz & Wallach, 1966; Craske & Crawshaw, 1974), it is likely in these experiments that attention was maintained. In the first of these S was repeatedly prodded with a rod, in the second his limb was vibrated, and in the last he had to stand on the (stationary) feet he was inspecting. The lack of adaptation found in other studies where S was required merely to look at a passively moved limb may be due to his simply not attending to kinesthetic information. In the present case, however, such inattention is unlikely, for Ss were required to make judgments of position at frequent intervals, and to make active movements with their untreated arm on the basis of these judgments. Paillard and Brouchon (1968) found a great contrast between active and passive conditions, while in the present experiment the increase in accuracy was quite modest: perhaps this may be attributed to the smaller voluntary effort required in the latter. This suggests that under everyday conditions, where the limbs are lightly loaded, active movement would result in little increase in accuracy. REFERENCES COLLINS,J. K. Isolation of the muscular component in a proprioceptive spatial aftereffect. Jordmal o f Expe7jnental Psychology, 1971, 90, 287-299. CRASKE,B. Change in transfer function of joint receptor output. Nature, 1966, 210, 764-765. CRASKE, B., & CRAWSHAW,M. Adaptive changes of opposite sign in the oculomotor systems of the two eyes. Quarterly J O I N Mo~f Experimental Psychology, 1974, 26, 106-113. GOODWIN,G. M., MCCLOSKEY, D. I., & MATTHEWS, P. B. C. The contribution of muscle afferents to kinesthesia shown by vibration induced illusions and by the effects of paralysing joint afferencs. Brain, 1972. 95, 705-748. HARRIS.C. S. P e r c e p ~ a ladaptation to inverted, reversed and displaced vision. Psychological Review, 1965, 72, 419-444. HELD,R., & FREEDMAN, S. J.. Plasticity in human sensorimotor control. Science, 1963, 142, 455-462. HELD.R., & HEIN,A. Adaptation of disarranged hand-eye coordination contingent upon re-afferent stirnulation. Perceptual and Motor Skills. 1758, 8, 87-90. HOWARD, I. P., CRASKE, B., & TEMPLETON, W. B. Visuomotor adaptation to discordant exafferent stimulation. Journal o f Exfierimental Psychology, 1965, 70, 187-171. KRAVITZ, J. H., & WALLACH, H. Adaptation to displaced vision contingent upon vibrating stimulation. Psychononic Science, 1766, 6, 465-466.
SHIFTS IN KINESTHESIS AFTER MOVEMENT
761
LASHLEY,K. S. The accuracy of movement in the absence of excitation from the moving organ. American Jorirnal of Physiology, 1917, 43, 169-194. PAILLARD,J., & BROUCHON,M. Active and passive movement in the calibration of position sense. In S. J. Freedman (Ed.), The neuropsychology of spatially oriented behavior. Homewood, Ill.: Dorsey, 1968. Pp. 37-55. SKAVENSKI,A. A. Inflow as a source of extra-retinal eye position information. Vision Research, 1972, 12, 221-229. SKOGLUND,S. Joint receptors and kinesthesis. In A. Iggo (Ed.), Handbook of sensory physiology. New York: Springer, 1973. Pp. 50-93.
Accepted February 14, 1975.
SHIFTS IN KINESTHESIS THROUGH TIME AND AFTER ACTIVE AND PASSIVE MOVEMENT1 BRIAN CRASKE AND MARTIN CRAWSHAW University o/ Southampton Summary.-The position sense of a stationary acm was investigated subsequent to an horizontally adductive movement with axis the shoulder joint. The right arm was the treated arm: it reached a test position actively, using minimal voluntary effort, or passively from each of 10 starting positions. The blindfolded S localized the index finger of the treated arm by attempting to touch it with the index finger of his left hand. The results indicate that subsequent to active movement the final position of a limb is more accurately known than a posicion resulting from passive movement. A second finding is that concomitant with both forms of limb placement there is a unidirectional drift of perceived limb position over trials.
It is now commonly accepted that receptors associated with the joints are responsible for the static position sense of limbs. The evidence in support of such a view has recently been marshalled by Skoglund (1973). A difficulty for this view may be inferred from a study by Collins (1971), who showed that when limb position was held constant and the muscles opposed a torque, subsequent judgments of the horizontal plane of the pronated hand were significantly different from control measures. Since the joint capsule is not rigid, such a procedure could produce an unusual yet adequate stimulus for the joint receptors: biased positional aftereffects due to this form of stimulation may still be mediated by the joint. If this does not provide adequate stimulus for the joint receptors, however, the experiment can be interpreted to suggest an involvement of muscle in position sense. Recent work on position sense in the eye (Skavenski, 1972) also suggests a muscular component. More dramatically, a study using vibration of the biceps tendon, a procedure which drives the muscle spindles, yielded estimates of arm position which were in error by as much as 8" (Goodwin, McCloskey, & Matthews, 1972). Although this situation is far from one which would occur under normal circumstances, it does suggest that positional information is being fed into the system. Its modus operandi, however, may be no more direct than a bias in positional analysis due to a massive afferent bombardment of movement information. Their work does imply, however, that structures other than joints might provide this information. If this is so, the normally functioning adult could be supplied with a positional signal arising from the joints, and additional positional information derived from non-joint sources. To this may be added a third source, positional information derived from monitoring the outflow of efference to the muscles, 'This work was supported by the Medical Research Council.
756
B. CRASKE
&
M. CRAWSHAW
for accurate movement and positioning can occur in the absence of afference from the joints (Lashley, 1917). Accuracy of determination of limb position has been the subject of a widely quoted study by Paillard and Brouchon ( 1968). They investigated the effects of active and passive placement, and active and passive maintenance of a position of a limb, on its localization by the contralateral limb. Their conclusion was that precision is significantly better with active rather than passive movement of the arm. This conclusion is not convincing, however. It appears that both the treated and the indicating arms were moved from the same starting position. It follows that in the active condition, S need only repeat the excursion of the target arm with the indicating arm. Thus judgments of excarsion were confounded with judgments of position. This methodological weakness casts some doubt upon their findings. Furthermore, the results of their second experiment, which manipulated elapsed time between placement and estimation, are at variance with their conclusions, for little difference between active and passive conditions was found. The existence of a difference in perceived limb position after active and passive movement is still unresolved. Possible changes in registered limb position are implied by a recent pilot experiment. In this, 12 Ss had to point to visual targets with both arms and without error feedback for 90 min. The trend of the errors of localization was consistent with the notion that with time the registered position of each limb was changing in opposite directions; to the left for the right arm and to the right for the left arm. In the experiment to be reported the accuracy of information available to S about the position of his stationary limb was investigated after passive transport of the limb and after active movement involving minimal effort in its execution. At the same time, the possibility of a systematic drift in kinesthesis was investigated.
METHOD Preliminary Tests Electromyographic recordings were taken from those muscles which are implicated in the performance of horizontal adduction of the arm. This was measured using 12 silver/silver chloride s ~ ~ r f a celectrodes e in conjunction with Cambridge electrode jelly. Raw EMG data were integrated by two methods, either by using a fixed epoch or by a fixed threshold. This established that passive movement of 0.1 rad/sec. produced EMG not measurably different from that of the resting and relaxed condition. These trials suggested that movements of this velociry would be suitable for the condition of passive movement in the experiment.
SHIFTS IN KINESTHESIS AFTER MOVEMENT
Subjects and Apparatas Sixteen male Ss were blindfolded, stripped to the waist and were seated in front of a two-tier platform. The lower tier supported a light-weight (365 gm.) free-running carriage. The carriage could be strapped to the arm which could be moved horizontally, either actively by S, or passively by a constant-speed electric motor controlled by E. The axis of movement was S s shoulder joint. The carriage could be altered in length to suit S and was designed to support his arm horizontally with the hand semi-pronated, index finger outstretched. Imrnediately above the hand and attached to the cradle was a horizontal scale, its zero coincident with the centre-line of the index finger. This was used to read off
FIG. 1. Subject pointing to contralateral limb which has been moved into position either actively or passively
errors of localization. The upper tier of the apparatus was of glass and was mounted immediately above the error scale on the carriage. A triangular section had been removed from the left side of the upper tier, S could therefore rest his left arm in a comfortable position on the lower tier, yet without obstruction he was able to raise it and then bring it down over the index finger of the right arm. A line was drawn on the nail of the left index finger to provide a datum. The upper tier prevented contact between the hands. A scale centered about S's right shoulder was marked in degrees on the lower tier, this enabled E to position the arm.
Procedure The design counterbalanced active and passive conditions. Trial excursions were made in whichever condition S was to perform first. If this was the active condition, then S adjusted the speed of his excursion to approximate closely the speed used throughout the passive condition (0.1 rad/sec.). During these practice excursions, S rehearsed stopping on E's instruction and indicating the position of his right index finger with his left. In the passive condition, S's arm was moved to one of 10 starting positions lying between 30°and 40' to the right of his median sagittal plane, thus prevent-
758
B. CRASKE
&
M. CRAWSHAW
ing S making judgments based on excursion. Each position was used twice in both halves of the experiment, the order of appearance being randomized. After 10 sec. S s arm was moved toward his left and stopped at one of 10 positions lying between 12' left and 5" right of his median sagittal plane. He was then instructed to point with his left hand. E noted rhe error, and the left arm was returned to its resting position. After 3 sec. S's right arm was returned at the same speed as before to the next starting position. In the active condition, the only difference in procedure was that the excursions were voluntary and were stopped on command. Twenty trials were given in the first condition and were followed, without break, by 20 trials in the second condition. RESULTS Only angular errors of pointing were measured, errors of depth being ignored. Errors were measured along the arc described by the right index finger. The data for 20 trials in both conditions for each of 16 Ss were subjected to a 3-way analysis of variance. The mean position indicated by all Ss for the active condition was 1'36' and for the passive condition, 2'57'; in both conditions to the left of his right index finger. The difference between these two means was significant ( F = 5.31, df = 1/15, p < 0.05 ). The differences between Ss were significant (F = 135.75, df = 15/285, P < 0.001), as was the Ss X conditions interaction ( F
TRIALS FIG. 2. Mean errors of localization
SHIFTS IN KINESTHESIS AFTER MOVEMENT
759
< 0.001). The effect of order of trials was such that there was a general increase in size of error toward the left as the experiment proceeded. This effect existed in both conditions and, for both conditions combined, was significant ( F = 2.59, df = 19/285, P < 0.01). The variances of the active and passive conditions were not significantly different ( t = 1.41, df = 1 5 ) and variance of the condition performed first was not significantly different from that performed second ( t = 0.3, d f = 15 ) . The pooled within-condition standard deviation was 5'07'. = 9.35, df = 15/285, p
DISCUSSION The order effect is such that Ss tended to make significantly greater errors of localization as the experiment ~ r o ~ r e s s e dThese . errors were of undershoot, i.e., the arms were separated, not crossed. As is shown in Fig. 2, this effect occurred with similar rate of increase over trials in both the active and passive conditions. This is consistent with the nonsignificant interaction of order X conditions. It seems that kinesthesis manifests an underlying lability when information about the accuracy of its operation is not available. That limb position sense is exceedingly labile is not in doubt; experiments with prisms (Harris, 1965; Craske, 1966) have shown that kinesthesis can be easily modified when vision and kinesthesis are discordant. The property of rapid modification of position sense in order to overcome prism-induced errors of reaching suggests that similar cross-modal checking of position information may be used to ensure veridicality in limb position sense in the circumstances of everyday life. Since the rate of drift in the present experiment was the same in both active and passive conditions, the significant main effect cannot be attributed to this factor. Concerning the main effect, the findings support the notion that there is a small but real superiority of localization subsequent to active movement, even under these stringent conditions when voluntary effort is very small, and the mechanical loading of the limbs minimal. This suggests that mechanisms other than those associated with the joint can influence the registered position of a stationary limb, unless even these minima1 conditions of muscular involvement can influence the joint receptors via muscles inserting at the joint. The significant interaction between Ss and conditions is an indication that some individuals utilize the information made available by active movement more than others. As previous writers have suggested, it appears that muscular involvement does lead to greater accuracy, but there could be an alternative explanation. It might be suggested that the effect of active involvement of muscles on performance is due to the increased attention that this entails. However, if inattention were the source of greater errors in the passive condition, they would presumably
760
B. CRASKE
(i:
M. CRAWSHAW
be randomly distributed. Hence, the variance of localizations made in that condition would increase, but there would be no great difference in mean position between conditions. On the contrary, the results show that there was a significant difference in mean position, but no significant difference in variance. Nevertheless, attention may well be an important factor in other studies directed at differences between active and passive conditions. The issue of the necessity for active movement in adaptation to displaced vision is a case in point (Held & Hein, 1958; Held & Freedman, 1963). Although it has been shown that such adaptation calz take place in the absence of active movement (Howard, Craske, & Templeton, 1965; Kravitz & Wallach, 1966; Craske & Crawshaw, 1974), it is likely in these experiments that attention was maintained. In the first of these S was repeatedly prodded with a rod, in the second his limb was vibrated, and in the last he had to stand on the (stationary) feet he was inspecting. The lack of adaptation found in other studies where S was required merely to look at a passively moved limb may be due to his simply not attending to kinesthetic information. In the present case, however, such inattention is unlikely, for Ss were required to make judgments of position at frequent intervals, and to make active movements with their untreated arm on the basis of these judgments. Paillard and Brouchon (1968) found a great contrast between active and passive conditions, while in the present experiment the increase in accuracy was quite modest: perhaps this may be attributed to the smaller voluntary effort required in the latter. This suggests that under everyday conditions, where the limbs are lightly loaded, active movement would result in little increase in accuracy. REFERENCES COLLINS,J. K. Isolation of the muscular component in a proprioceptive spatial aftereffect. Jordmal o f Expe7jnental Psychology, 1971, 90, 287-299. CRASKE,B. Change in transfer function of joint receptor output. Nature, 1966, 210, 764-765. CRASKE, B., & CRAWSHAW,M. Adaptive changes of opposite sign in the oculomotor systems of the two eyes. Quarterly J O I N Mo~f Experimental Psychology, 1974, 26, 106-113. GOODWIN,G. M., MCCLOSKEY, D. I., & MATTHEWS, P. B. C. The contribution of muscle afferents to kinesthesia shown by vibration induced illusions and by the effects of paralysing joint afferencs. Brain, 1972. 95, 705-748. HARRIS.C. S. P e r c e p ~ a ladaptation to inverted, reversed and displaced vision. Psychological Review, 1965, 72, 419-444. HELD,R., & FREEDMAN, S. J.. Plasticity in human sensorimotor control. Science, 1963, 142, 455-462. HELD.R., & HEIN,A. Adaptation of disarranged hand-eye coordination contingent upon re-afferent stirnulation. Perceptual and Motor Skills. 1758, 8, 87-90. HOWARD, I. P., CRASKE, B., & TEMPLETON, W. B. Visuomotor adaptation to discordant exafferent stimulation. Journal o f Exfierimental Psychology, 1965, 70, 187-171. KRAVITZ, J. H., & WALLACH, H. Adaptation to displaced vision contingent upon vibrating stimulation. Psychononic Science, 1766, 6, 465-466.
SHIFTS IN KINESTHESIS AFTER MOVEMENT
761
LASHLEY,K. S. The accuracy of movement in the absence of excitation from the moving organ. American Jorirnal of Physiology, 1917, 43, 169-194. PAILLARD,J., & BROUCHON,M. Active and passive movement in the calibration of position sense. In S. J. Freedman (Ed.), The neuropsychology of spatially oriented behavior. Homewood, Ill.: Dorsey, 1968. Pp. 37-55. SKAVENSKI,A. A. Inflow as a source of extra-retinal eye position information. Vision Research, 1972, 12, 221-229. SKOGLUND,S. Joint receptors and kinesthesis. In A. Iggo (Ed.), Handbook of sensory physiology. New York: Springer, 1973. Pp. 50-93.
Accepted February 14, 1975.
