SUMMARYThe regions of the gill neural plexus responsible for habituation of the withdrawal reflex were studied using surgical and electrophysiological techniques. Waterdrops were applied to the gill pinnule, and testing for habituation was carried out in each preparation before and after ganglionectomy of the large ganglion in the gill and pinnule isolation. Three parameters of habituation, response decrement, spontaneous recovery, and dishabituation, were used to evaluate the function of the ganglion and the plexus in the pinnule in habituation of the reflex. Disruption of response decrement, of spontaneous recovery, and of dishabituation were caused by either isolating the stimulated pinnule from the gill ganglion by severing the branchial nerve, or by extirpating the gill ganglion. We conclude that: a) the gill ganglion is necessary for the withdrawal reflex to exhibit the three parameters which characterize hibituation; and b) the plexus in the pinnule mediates the withdrawal reflex. Extracellular recordings from the pinnule showed that sensory adaptation cannot be contributing to habituation of the reflex since the apparent sensory activity evoked by waterdrops in the pinnule was unaltered by repeated stimulation. In gill ganglion neurons, biphasic potentials were evoked by waterdrops and consisted of an initial depolarization followed by a longer lasting hyperpolarization whose duration was increased with repeated stimulation. I n three spontaneously active neurons the rate of activity decreased with repeated stimulation.It is felt that the decreased rate is brought about by potentiation of the hyperpolarization. The rate of spike activity is considered to control the reflex amplitude since the increased rate with intracellular stimulation, in one neuron, resulted in an increased amplitude and since the absence of spontaneous discharge, after ganglionectomy, resulted in a reduced amplitude. A neural network mediating habituation of the withdrawal reflex is proposed.191