Focal epileptic activity (FEA) in amygdalohippocampal complexes (AHC) is mostly highly intensive (except in patient BUK where it is nearly missing). Unilateral FEA is hardly ever propagated to the superficial EEG electrodes, bilateral FEA only very rarely (patients BRY and DUS), in some patients solely in deep synchronous sleep (patient LOB). FEA intensity rises in relaxed vigilance and in superficial sleep while tending to decrease in deep synchronous sleep. FEA intensity tends to drop or even disappear in active vigilance during mental activity or paradoxical sleep. Epileptic activity generalized in all the superficial and deep-implanted leads is mostly accompanied by clinical manifestations (absences, twitching, motor automatisms), or to put it the other way round: if a clinical paroxysms is in progress, there is evidence of generalized epileptic activity in all the leads. The time parameter is of no consequence, the inconspicuous motion of the hand is due to a second-lasting discharge of polyspike and wave (patient BRY in sleep). If prolonged discharges remain localized there is subclinical paroxysm (patient IRL). Similar findings were reported by Lieb et al. (1976). All we can add is that the start of an attack depends not only on the amplitude and frequency of the spikes but also on the regularity of spike intervals. Superficial "neocortical" EEG and deep "paleocortical" SEEG exhibit equal sleep stages equally, i.e. either there is synchronization in all the leads (like in synchronous sleep), or there is desynchronization (such as in active vigilance or in paradoxical sleep); those two cortical structures are not antithetical such as in, e.g., rats or cats. The sleep stages show quantitative as well as qualitative changes. There is increasingly more wakefulness and superficial sleep at the expense of spindle and paradoxical sleep. EEG graphoelements often show little differentiation, e.g. the sleep spindles are short and irregular in shape, delta activity is low in amplitude and also irregular in shape, and paradoxical sleep shows insufficient desynchronization in EEG and preserved tonic muscular activity. Epileptic activity variability is often found helpful for the reliable identification of the sleep stage concerned.(ABSTRACT TRUNCATED AT 400 WORDS)