The action of single static gamma-axons in 'single fibre' ventral root filaments was tested on the responses of from two to eight spindles in the same tenuissimus muscle in anaesthetized cats. The type of intrafusal fibre activated in each spindle by a particular axon was diagnosed from its effect on the discharge of primary and secondary sensory endings, frequently in the same spindle. One or two spindles were subsequently isolated in most experiments and the type of intrafusal fibre activated by the axon was observed directly. Except in the case of a few weak actions which could not be interpreted, the diagnosis made from the afferent recordings was always correct. Twenty-four static gamma-axons activated the chain fibres in every spindle they supplied on which their action was tested; the static bag2 fibre was also involved in about 20% of the spindles though its contraction was quite often weak. Eleven static gamma-axons activated the static bag2 fibre in every spindle they supplied on which their action was tested; some chain fibres were activated as well in about 20% of the spindles. There were only two possible exceptions to the general rule that a static gamma-axon activates either the chain fibres or the static bag2 fibre in every spindle it supplies and never activates only the chain fibres in one spindle, and only the static bag2 fibre in another spindle. Direct observation of isolated spindles provided preliminary evidence that some non-selective connexions of static gamma-axons have a pronounced physiological action by way of their terminals on one type of intrafusal fibre while the terminal(s) on the other type of fibre produce such a weak contraction that the sensory endings are not affected. Other non-selective connexions do, however, have a significant physiological effect by way of their terminals on both the static bag2 fibre and the chain fibres. It is proposed that there are two types of static gamma-motoneurone. 'Static bag gamma-motoneurones', through their terminals on the static bag2 fibre in every spindle, bias the discharge from primary endings (and to a lesser extent from some secondary endings) but leave their length sensitivity largely unaltered; connexions to chain fibres in some spindles have a weak action on primary endings but may greatly increase the length sensitivity of secondary endings.(ABSTRACT TRUNCATED AT 400 WORDS)