1. In cats anaesthetized with sodium pentobarbital 160 external intercostal muscle spindle afferents were identified by their pause in response to ventral root stimulation; the internal intercostal muscle was denervated. 2. In order to assignate the afferents to either primary or secondary endings they were tested for their responsiveness to vibration (Bianconi & Van Der Meulen, 1963). The maximal frequency which they were able to follow regularly for at least four cycles, termed 'critical frequency', was determined. 3. The endings fell into two groups: low-f (frequency) sensitive endings with critical frequencies below 400 c/s and high-f sensitive endings with critical frequencies above 400 c/s. The latter were regarded as primary endings and the former as secondary ones. 4. The manner in which the spindle endings resumed activity after a pause produced by shocks to the ventral root, i.e. whether 'phasic' or 'tonic' (Granit & Van Der Meulen, 1962) was studied in all the spindle afferents. 5. All the secondary (low-f sensitive) endings were 'tonic' except three for which the determination of critical frequency was questionable. Both 'tonic' and 'phasic' properties were found among the primary (high-f sensitive) endings. 6. The majority of the secondary endings (74%) showed inspiratory ythmic fusimotor activation in parallel with the skeletomotor contracton as did the primary endings (79 %). 7. Fifty-seven spindle endings which all showed marked rhythmic inspiratory gamma activation were tested for respiratory variations in their dynamic responses to steady stretch and length changes introduced at low repetition rates. 8. The results indicate that both 'dynamic' and 'static' gamma fibres are represented among the rhythmic gamma fibres controlling primary muscle spindle endings, whereas rhythmic activation of secondary endings seems to be mediated only by 'static' fibres.