Diurnal- and behaviour-related activity of ventromedial hypothalamic neurones in freely behaving rats.

1. Long-term (25-93 h) single-neurone activity of seventy-eight ventromedial hypothalamic (v.m.h.) neurones was analysed in freely behaving rats. 2. The activity of fifty-seven (73%) of the seventy-eight neurones tested depended on arousal and sleep, determined by electroencephalogram (e.e.g.), and varied diurnally. The activity of 39.8% (31/78) increased during e.e.g. arousal, and decreased during slow-wave sleep. Activity of 33.3% (26/78) varied inversely. 3. The activity of 21.8% (17/78) was independent of e.e.g. arousal and sleep. The activity of all except one of the seventeen increased in the dark period and varied diurnally. Activity of one increased during light stimulation. 4. The activity of the remaining 5.1% (4/78) was less than 0.5 spikes/s, and was not related to e.e.g. arousal or sleep or the diurnal cycle. 5. The effects of bilateral lesions of the suprachiasmatic nuclei (s.c.n.) were investigated on six v.m.h. neurones. All of these lost their diurnal variation; five retained their dependence on e.e.g. arousal and sleep, and one retained its independence of e.e.g. arousal and sleep. 6. There was no discernible relation between v.m.h. recording site and any of the tested characteristics of these neurones. 7. Of seventy-eight v.m.h. neurones tested, seventy-four were classified as either: (1) diurnal, sleep-arousal related, or (2) diurnal, sleep-arousal independent. Because s.c.n. lesions abolished the circadian rhythm of v.m.h. activity in every neurone tested, it was concluded that at least some v.m.h. diurnal variation is driven from the s.c.n. as a biological clock. It is clear also that responses related to e.e.g. are driven from some source other than the s.c.n.