On measuring the strength of conditioned reflexes during the period of spontaneously developing or experimentally induced sleep, it is found that, before conditioned reflexes disappear alto gether, they pass through a variety of stages in which the normal relation between the strength of response and the strength of stimulus is altered in some way. There is, for instance, a phase in which the animal responds to all conditioned stimuli, whether strong or weak, but the usually weak responses become some what stronger, and the strong ones become somewhat weaker. Reflexes which showed differences of about 00% in strength are now equalised. This phase is known as the phase of equalisation. With further development of sleep, it is found that the weak stimuli produce a definitely greater response (in some cases 00% greater) than the strong stimuli. This phase is known as the para doxical phase. Finally, with a still further intensification of sleep, all conditioned reflexes vanish completely, and this is followed sooner or later by the relaxation of the muscles and the develop ment of complete, sleep. The process of awakening seems to go
through the same stages.
Other changes in strength of conditioned reflexes have been observed. The question whether they can be arranged in a definite order, and if so in what order, must remain entirely open for the present. All these different states of activity of the hemispheres bear a strong resemblance to the different stages of what is gener ally known as hypnotism, which from the point of view of condi tioned reflexes is merely a transition stage between alertness and complete sleep, a stage of a greater or less extent and a greater or less intensity of sleep. Inhibition or localised sleep is a phe nomenon always present, under whatever conditions the animal or man may be. The state of alertness is only an expression of a predominance of excitation, while sleep is due to a predominance of inhibition. (G. AN.)