Abstract
The finding by Molinari and Foulkes (1969) that REM-burst sleep mentation is more sensory and less conceptual than that occurring during episodes of ocular quiescence within EEG stage REM was confirmed, but at a reduced level of absolute discriminability and only when spontaneous portions of S's dream interview were considered. Evidence was gathered to suggest that bursts of EEG sawtooth waves which often precede REMs are phasic events with mentation correlates comparable to those of REM bursts themselves and that sawtooth activity may, in fact, mark the point at which the most discontinuity is experienced in S's train of sleep thought. Implications of these findings for tonic-phasic and scanning-hypothesis models are discussed, and the possibility is presented of extending them to studies of non-REM sleep mentation.