If this is the nature of sleep, then dreams do not come into its scheme at all, but seem rather to be an unwelcome supplement to it; and we do indeed believe that dreamless sleep is the best, the only proper sleep. There should be no mental activity during sleep; if any such activity bestirs itself, then in so far have we failed to reach the true pre-natal condition of peace; we have not been able to avoid altogether some remnants of mental activity, and the act of dreaming would represent these remnants. In that event it really does seem that dreams do not need to have meaning. With errors it was different, for they were at least activities manifested in waking life; but if I sleep and have altogether suspended mental activity, with the exception of certain remnants which I have not been able to suppress, there is no necessity whatever that they should have any meaning. In fact, I cannot even make use of any such meaning, seeing that the rest of my mind is asleep. It can really then be a matter of spasmodic reactions only, of such mental phenomena only as have their origin in physical stimulation. Hence, dreams must be remnants of the mental activity of waking life disturbing sleep, and we might as well make up our minds forthwith to abandon a theme so unsuited to the purposes of psycho-analysis. Superfluous as dreams may be, however, they do exist nevertheless, and we can try to account for their existence to ourselves. Why does not mental life go off to sleep? Probably because there is something that will not leave the mind in peace; stimuli are acting upon it and to these it is bound to react. Dreams therefore are the mode of reaction of the mind to stimuli acting upon it during sleep. We note here a possibility of access to comprehension of dreams. We can now endeavour to find out, in various dreams, what are the stimuli seeking to disturb sleep, the reaction to which takes the form of dreams. By doing this we should have worked out the first characteristic common to all dreams.