Now see how much information we have gained, and that with hardly any trouble, from our study of children’s dreams! We have learnt that the function of dreams is to protect sleep; that they arise out of two conflicting tendencies, of which the one, the desire for sleep, remains constant, whilst the other endeavours to satisfy some mental stimulus; that dreams are proved to be mental acts, rich in meaning; that they have two main characteristics, i.e., they are wish-fulfilments and hallucinatory experiences. And meanwhile we could almost have forgotten that we were studying psycho-analysis. Apart from the connection we have made between dreams and errors our work has not borne any specific stamp. Any psychologist knowing nothing of the assumptions of psycho-analysis could have given this explanation of children’s dreams. Why has no one done so? If only all dreams were of the infantile type the problem would be solved and our task already achieved, and that without questioning the dreamer, referring to the unconscious or having recourse to the process of free association. Clearly it is in this direction that we must continue our work. We have already repeatedly found that characteristics alleged to be universally valid have afterwards proved to hold good only for a certain kind and a limited number of dreams. So the question we now have to decide is whether the common characteristics revealed by children’s dreams are any more stable than these, and whether they hold also for those dreams whose meaning is not obvious and in whose manifest content we can recognize no reference to a wish remaining from the day before. Our idea is that these other dreams have undergone a good deal of distortion and on that account we must refrain from immediate judgement. We suspect too that to unravel this distortion we shall need the help of psycho-analytic technique, which we could dispense with while learning, as we have just now done, the meaning of children’s dreams.