The sleeping style of each organism is exquisitelyadapted to the ecology of the animal. It is conceivable that animals who are too stupid to be quiet on their own initiative are, during periods of high risk, immobilized by the implacable arm of sleep.

However, in part for reasons of organizationalconvenience, modern societies are structured as if all humans had the same sleep requirements; and in many parts of the world there is a satisfying sense of moral rectitude in rising early. The amount of sleep required for buffer dumping would then depend on how much we have both thought and experienced since the last sleep period.

This looks very much as if the integration of the day's experience into our memory, the forging of new neural links, is either an easier or a more urgent task. As the night wears on and this function is completed, the more affecting dreams, the more bizarrematerial, the fears and lusts and other powerful emotions of thedream material emerge.

There is some evidence that dreaming is necessary. When people or other mammals are deprived of REM sleep (by awakening them as soon as the characteristic REM and EEG dream patterns emerge), the number of initiations of the dream state per night goes up, and, in severe cases, daytime hallucinations-that is, waking dreams-occur.

There is another connection between infancy and dreams: both are followed by amnesia. When we emerge from either state, we have great difficulty remembering what we have experienced. In both cases, I would suggest, the left hemisphere of the neocortex, which is responsible for analyticrecollection, has been functioning ineffectively.

Thus, an inhibition center developed below what in humans is the temporal lobe, to turn off much of the functioning of the reptilian brain; and an activation center evolved in the pons to turn on the R-complex, but harmlessly, during sleep.