The distribution of fusimotor axons to bag1, bag2 and chain muscle fibres in cat tenuissimus spindles has been studied using a modification of the glycogen-depletion technique of Edstrrom & Kugelberg (1968). Single fusimotor axons were stimulated intermittently at 40-100/sec for long periods (30-90 sec) during blood occlusion. Portions of muscle containing the activated spindles were quick-frozen, fixed in absolute ethanol during freeze-substitution, and then embedded in paraffin wax. Serial transverse sections were stained for glycogen using the periodic acid-Schiff method, and examined for depletion. 2. Dynamic gamma axons (i.e. those that increase the dynamic index of primary-ending responses to ramp stretches of large amplitude) depleted bag1 fibres almost exclusively. 3. Static gamma axons (i.e. those that reduce or abolish the dynamic index) depleted both bag and chain fibres. Bag1 and bag2 fibres were depleted about equally. 4. A single static gamma axon may activate both bag and chain fibres in one spindle (the most common pattern), chain fibres only in another, and bag fibres only in a third spindle. 5. Static gamma axons with conduction velocities less than 25 m/sec also had a non-selective distribution, but no depletion was observed in bag2 fibres. 6. The zones of depletion produced by dynamic gamma axons were distributed more or less equally in the intra- and extracapsular parts of spindle poles, whereas those produced by static gamma axons were mainly intracapsular. 7. The results are compared with the glycogen-depletion studies of Brown & Butler (1973, 1975) and our own study of the distribution of static gamma axons to spindles in which all other motor axons had degenerated (Barker, Emonet-Dénand, Laporte, Proske & Stacey, 1973). The implications of the finding that both static gamma and dynamic gamma axons activate bag1 fibres are discussed.
