Human subjects attempted to maintain a constant force by flexing their elbow against a spring which was attached to a force transducer at one end and the subject's wrist at the other. The tremor at 8-12 Hz which develops in this situation was enhanced in amplitude with negligible change of frequency by applying vibration at 100 Hz to the tendon either of an agonist muscle (biceps brachii) or of the antagonist (triceps brachii). The tremor was assessed by spectral analysis of the force records and measurement of the peaks in the spectra. The compliance of the spring was normally 2.8 N/mm and target forces of 40-120 N were used. 2. The percentage increase in the tremor on applying vibration was relatively independent of target force, although the absolute amounts of tremor increased markedly with increasing target force. The average increase was greater for vibration of triceps than for biceps (70% as opposed to 37%, averaged between subjects and over a range of forces). 3. When the spring was replaced by a rigid connexion there was usually no clear tremor peak either in the presence or absence of vibration. Vibration, however, tended to increase the general noisiness of the force signal. 4. Qualitatively similar effects were seen when the elbow exerted an extending force so that triceps became the agonist and biceps the antagonist. 5. The tremor peak present in the spectrum of the demodulated electromyogram during vigorous tremor increased in size when vibration made the tremor larger. 6. The effect of a rhythmic afferent input was studied by modulating the amplitude of the vibration at 8-9 Hz, to correspond to the tremor frequency, while the subject pulled against a rigid attachment. Both the e.m.g. and the tension spectra contained peaks at the modulation frequency. The raw force records showed that, with reference to the modulation, the effects of biceps and of triceps vibration were approximately 180 degrees out of phase with each other, as would occur if vibration of one were having an excitatory action, and vibration of the other an inhibitory action. Moreover, in each case the effect on force (whether excitatory or inhibitory) lagged about half a cycle on the vibration envelope, as required for such reflexes to help in the generation of tremor. 7. It is suggested that vibration increases the modulation of Ia firing elicited by a given movement tremor and this, by means of the stretch reflex arc, enhanced the tremor. The powerful action of vibration of the antagonist illustrates, it would seem, the functional effectiveness under normal conditions of a spinal inhibitory pathway, most probably the Ia disynaptic route. The findings are also discussed in relation to the increase in stretch reflex gain that occurs in association with increasing strength of voluntary contraction.
