1. We studied the activity of 254 cells in the primary somatosensory cortex (SI) responding to inputs from peripheral proprioceptors in a variety of tasks requiring active reaching movements of the contralateral arm. 2. The majority of cells with receptive fields on the proximal arm (shoulder and elbow) were broadly and unimodally tuned for movement direction, often with approximately sinusoidal tuning curves similar to those seen in motor and parietal cortex. 3. The predominant temporal response profiles were directionally tuned phasic bursts during movement and tonic activity that varied with different arm postures. 4. Most cells showed both phasic and tonic response components to differing degrees, and the population formed a continuum from purely phasic to purely tonic cells with no evidence of separate distinct phasic and tonic populations. This indicates that the initial cortical neuronal correlates of the introspectively distinguishable sensations of movement and position are represented in an overlapping or distributed manner in SI. 5. The directional tuning of the phasic and tonic response components of most cells was generally similar, although rarely identical. 6. We tested 62 cells during similar active and passive arm movements. Many cells showed large differences in their responses in the two conditions, presumably due to changes in peripheral receptor discharge during active muscle contractions. 7. We tested 86 cells in a convergent movement task in which monkeys made reaching movements to a single central target from eight peripheral starting positions. A majority of the cells (46 of 86, 53.5%) showed a movement direction-related hysteresis in which their tonic activity after movement to the central target varied with the direction by which the arm moved to the target. The directionality of this hysteresis was coupled with the movement-related directional tuning of the cells. 8. We recorded the discharge of 93 cells as the monkeys performed the task while compensating for loads in different directions. The large majority of cells showed a statistically significant modulation of activity as a function of load direction, which was qualitatively similar to that seen in motor cortex under similar task conditions. Quantitatively, however, the sensitivity of SI proprioceptive cells to loads was less than that seen in motor cortex but greater than in parietal cortex. 9. We interpret these results in terms of their implications for the central representation of the spatiotemporal form (“kinematics”) of arm movements and postures. Most importantly, the results emphasize the important influence of muscle contractile activity on the central proprioceptive representation of active movements.
Primate motor cortex and free arm movements to visual targets in three- dimensional space. II. Coding of the direction of movement by a neuronal population
