Generation of scratching. II. Nonregular regimes of generation

1. The activity of muscle nerves and that of spinal interneurons from the L4 and L5 segments was recorded during fictitious scratching (5), which was evoked in decerebrate curarized cats by stimulation of the cervical spinal cord. In some experiments, rhythmical generation was disturbed by stimulation of the fifth lumbar dorsal root (DL5). 2. Excluding the very beginning of scratching, rhythmical generation was usually rather regular: fluctuations of the cycle duration were less than +/-5%. But changes in the stimulation strength, in the stimulating electrode position, and in the hindlimb position led to changes of the generation regime. In different regimes, the mean value of the cycle duration could differ by 20-30%. No correlation was found between mean durations of flexor and extensor phases for different regimes. 3. Rhythmical generation was possible only if the hindlimb was put to "scratch posture," i.e., deflected forward. Generation immediately stopped when the limb was deflected backward, and immediately started when it was returned to scratch posture. 4. In some experiments, stimulation of the cervical spinal cord first resulted in generation of slow oscillations with the temporal pattern typical of stepping (cycle duration about 500 ms, flexor and extensor phases being almost equal to each other). Then, during 5-20 cycles, gradual transition to a normal scratch cycle (about 250 ms) occurred mainly due to considerable shortening (5-10 times) of the extensor phase. In some experiments, considerable spontaneous variations of the flexor phase were observed, while the extensor phase was constant. 5. A single stimulus applied to DL5 considerably affected the cycle duration. Repetitive DL5 stimulation,with a rhythm close to that of scratching, resulted in synchronization of the spinal generator by the stimuli. 6. Spinal interneurons recorded during transition from slow oscillations to a normal scratch cycle only slightly changed phases of their activity in relation to the activity of motoneurons. 7. A hypothesis is advanced that generation of different kinds of limb movements is produced by one and the same central spinal mechanism which can operate in different regimes. The role of sensory input for operation of this mechanism is discussed.