characteristic of the subiculum as a projection system is the fact that it also operates as an important integrative center for the limbic system as a whole. The neurons of this structure receive axons from pyramidal cells in all fields of the hippocampus, from neurons of the amygdala, and from neurons in the entorhinal cortex [4, 7]. Data on the role of this structure in regulating the neurosecretory process at the level of the neurosecretory nuclei of the hypothalamus and in organizing limbic-hypothalamic projections are very limited. However, it has been shown that damage to the ventral subiculum leads to spontaneous release of oxytocin, and also facilitates the milkreflex [8]; electrical stimulation of this structure results in inhibition of background spike activity in neurosecretory cells (NSC) in the supraoptic nucleus, identified by an antidromal stimulation method [1]. Additionally, morphological data indicate that the quantity of extrahypothalamic projections addressed directly by oxytocinergic and vasopressinergic cells of the supraoptic nucleus is small [9]. Direct projections of the ventral subiculum were seen only in the region adjacent to the nucleus. This region is currently being studied by many groups, using both morphological and electrophysiological approaches. It is perfectly possible that cells in this area may function as mediators, relaying the transmission of the inhibitory effects of the subiculum to NSC in the nucleus. In fact, the region adjacent to the supraoptic nucleus has been found to contain GABAergic neurons which could function as inhibitory interneurons [6]. It has also been demonstrated that electrical stimulation of the areas located dorsally and dorsolaterally to the supraoptic nucleus and in its immediate vicinity leads to activation or inhibition of neurons in this nucleus [5]. The aim of the present work was to use electrophysiological methods to investigate the involvement of neurons in the perinuclear zone in the transmission of the inhibitory influences of the subiculum on NSC of the supraoptic nucleus. Neuron spike activity was recorded extracellularly using glass microelectrodes. The margins of the nucleus were defined in terms of the detection of NSC, which were identified from the characteristics of their responses to stimulation of the peduncle of the hypophysis. Neuron recording was started 300-500/~m above the dorsal margin of the supraoptic nucleus. Additionally, neurons were recorded in parallel tracks lacking NSC, by repeated insertion of the same microelectrode, assuming that the electrode moved only minimally. The subiculum and the peduncle of the hypophysis were stimulated using concentric metal electrodes and monophasic square-wave currents. Stimulation was performed using occasionalsingle impulses at frequencies of 0.1-0.5 per sec. Electrodes were inserted using coordinates from a stereotaxic atlas of the rat brain [3] with subsequent checking of the position by histological investigation. Electrical stimulation of the subiculum produced activation in 47 % of the neurons recorded in the juxtanuclear region; 15% of cells showed reductions in the background spike frequency, while the remaining cells did not respond to the stimulus. Excitation and inhibition responses were mostly phasic, i.e., there was a small latent period and responses were of short duration (see Fig. I, A). The latent period varied from 4 to 40 msec in different neurons, and responses could consist of one or a series of action potentials. The data obtained here suggest that the projection of the subiculum to NSC in the supraoptic nucleus are not monosynaptic, but are at least oligosynaptic. The probability of seeing both activatory and inhibitory responses approached unity. There were no significant differences in the numbers of responding neurons or in the latent periods of their responses at different sites of cell recording (in tracks containing NSC or in parallel tracks).