1. This article presents the results from stimulation in 21 loci within the medullary reticular formation (MRF; between 0.5 and 2.5 mm from the midline) and in 5 loci in the medial longitudinal fasciculus (MLF) of four intact, unanesthetized cats during locomotion. Stimulus trains (I 1 pulses, 0.2-ms duration, 330 Hz, stimulus strength 35-mu-A) were applied at those loci in each track at which the most widespread effects in each of the four limbs were obtained with the cat at rest. Electromyograms were recorded from flexor and extensor muscles of each limb. 2. As previously reported, stimulation with the cat at rest generally evoked brief, short-latency, twitch responses in both flexor and extensor muscles of more than one limb. In contrast, stimulation during locomotion evoked a more complex pattern of activity in which responses were normally evoked in one or other of the muscle pairs and incorporated into the locomotor pattern. 3. In the majority of sites, the stimulation evoked excitatory responses in the flexor muscles of each of the four limbs during that period of the step cycle in which each respective muscle was naturally active; stimulation in the stance phase of locomotion, although less effective, was also capable of producing responses in these muscles. All three ipsilateral extensor muscles studied [long and lateral heads of triceps and vastus lateralis (Tri, TriL, and VL, respectively)] were normally inhibited during their phase of muscle activity, although excitatory responses were occasionally seen. Responses in the contralateral (co) Tri were invariably excitatory and were largest during the period of muscle activity, whereas responses during the period of activity of the coVL were mixed, with both excitatory and inhibitory responses being seen from any one locus. 4. Excitatory responses were normally largest when stimulation was applied during the time that the muscle was active during the locomotor cycle. Responses evoked at times when the muscle was inactive were sometimes larger than those evoked with the animal at rest; such responses were most commonly seen in the hindlimb flexors and in the co VL. 5. In both flexors and extensors of each of the four limbs, the latency of the responses was greatest when the cat was at rest and least for stimuli given during the period of activity of the respective muscle. Average latencies during the period of muscle activity ranged from a minimum of 9.0 +/- 2.6 (SD) ms for inhibitory responses in the ipsilateral Tri and TriL to a maximum of 17.1 +/- 3.0 ms for the responses evoked in the ipsilateral semitendinosus. 6. Stimuli delivered to different loci in the anteroposterior plane (P3.1-9.7) generally evoked responses that were similar with respect to the sign of the response (excitation or inhibition) and to the period of the step cycle in which the maximal effect in any one muscle was observed. Differences occurred, however, in the relative amplitude of the response evoked in any one muscle from different loci. 7. The responses evoked by stimulation of the MRF and the MLF were similar in most respects. The relative amplitude of the responses in flexor muscles, however, was always greater for stimuli delivered within the MLF than in the MRF. The amplitude of the responses in the extensor muscles were generally similar for the two different regions. 8. The results suggest that there is a change in the functional connections between the MRF and its target motoneurons when the cat walks. This appears to be due principally to differential changes in the level of excitability of interneurons which are probably either part of, or driven by, the central pattern generator (CPG) for locomotion. In some muscles the effects are suggested to be mediated through both excitatory and inhibitory interneurons, with the final expression of the stimulation being dependent on the phase of the step cycle in which the stimulus is applied. 9. It is suggested that the known branching pattern of individual reticulospinal axons provides a substrate whereby the nervous system may simultaneously affect the activity of muscles of more than one limb. The role of the CPG in coordinating these responses in the different limbs is discussed.
