SUMMARY1. Stretching a frog muscle spindle evoked a discharge ofaction potentials in its sensory axon. As the rate of this discharge decreased during the adaptation that followed the dynamic phase of a stretch, the variability of the interspike intervals of the impulse train increased.2. Adaptation occurred in two phases. At first the impulse train was almost regular and adapted rapidly, but later this gave way to a phase of slower adaptation where the variability of the discharge was much increased. In the second phase ofadaptation the interspike intervals increased in length less than half as quickly as in the first phase.3. When the rate of adaptation changed from the more rapid to the slower phase there was often an abrupt change in the character of the discharge and the relationship between the mean interspike interval and the variability changed. The interspike interval at which this change-over occurred was relatively constant in records of the discharge from one afferent fibre even though stretches of different amplitude were employed, though it differed from one afferent fibre to another.4. These features of the discharge during adaptation suggest that the two sections of the impulse trains were derived from different spike generators by a process of probabilistic mixing.

SUMMARY1. Extension of a frog muscle spindle evoked a discharge of impulses in its sensory axon. The adaptation of the discharge after the dynamic phase of stretching occurred in two phases. At first the impulse train was almost regular and adapted quickly, but later this gave place to a phase of slower adaptation in which the variability of discharge was much increased.2. The discharge of action potentials by the muscle spindle depressed the response of the receptor to a subsequent extension. This was true whether they were elicited antidromically by afferent stimulation or orthodromically by longitudinal vibration. This depression had its most marked effect on the first phase of adaptation where it greatly increased the rate of adaptation. The second, slower, phase of adaptation was little influenced.3. The depression increased with the duration and with the frequency of afferent stimulation. It persisted for about 300 msec. 4. The results show that the properties of the spike generating mechanisms in the muscle spindle are modified by previous discharge and that this modification may influence the overall properties of the receptor.5. The fact that afferent stimulation has different effects on the two phases of adaptation is consistent with the suggestion that the impulse train evoked by extension is derived from more than one spike generator within the muscle spindle.