Differing control of neural activities during various portions of expiration in the cat.

1. Activities of the phrenic nerve, intercostal nerve and nerves innervating the thyroarytenoid (TA) muscle of the larynx and triangularis sterni (TS) muscle of the chest wall were recorded in decerebrate, vagotomized, paralysed and ventilated cats. 2. Neural inspiration was defined by the phase of phrenic activity. Neural expiration was divided into two phases with phase I corresponding to the duration of TA activity and phase II to TS activity: intercostal nerves discharged across both phases. 3. Phrenic activity was terminated prematurely by electrical stimulation of the superior laryngeal nerve or of the dorsolateral region of the rostral pons. Following stimulation, neural activities during phase I of expiration rose and those during phase II fell in most animals. 4. Stimulation of the superior laryngeal nerve during phase I caused augmentations of both TA and TS activity. At the termination of stimulation, a phase of TA discharge was recorded followed by a phase of TS activity. The durations of these post‐stimulation phases of TA and TS activities approximated those of cycles without stimulation. 5. Stimulation of the superior laryngeal nerve during phase II caused a resetting of neural expiration. Following stimulation, phases of TA and TS activity were recorded which had durations approximating those of cycles without stimulation. 6. The current required to induce a premature onset of phrenic activity by stimulation of the dorsolateral region of the rostral pons fell dramatically with the change from phase I to phase II of expiration. 7. We conclude that the control of neural activities differs markedly between phase I and phase II of expiration. The data support the hypothesis that post‐inspiratory medullary respiratory neurones play a fundamental role in the definition of the ventilatory cycle.