So dreams become the object of psycho-analytic research—another of these ordinary, under-rated occurrences, apparently of no practical value, like “errors,” and sharing with them the characteristic of occurring in healthy persons. But in other respects the conditions of work are rather less favourable. Errors had only been neglected by science, people had not troubled their heads much about them, but at least it was no disgrace to occupy oneself with them. True, people said, there are things more important but still something may possibly come of it. To occupy oneself with dreams, however, is not merely unpractical and superfluous, but positively scandalous: it carries with it the taint of the unscientific and arouses the suspicion of personal leanings towards mysticism. The idea of a medical student troubling himself about dreams when there is so much in neuropathology and psychiatry itself that is more serious, tumours as large as apples compressing the organ of the mind, hæmorrhages, chronic inflammatory conditions in which the alterations in the tissues can be demonstrated under the microscope! No, dreams are far too unworthy and trivial to be objects of scientific research. There is yet another factor involved which, in itself, sets at defiance all the requirements of exact investigation. In investigating dreams even the object of research, the dream itself, is indefinite. A delusion, for example, presents clear and definite outlines. “I am the Emperor of China,” says your patient plainly. But a dream? For the most part it cannot be related at all. When a man tells a dream, has he any guarantee that he has told it correctly, and not perhaps altered it in the telling or been forced to invent part of it on account of the vagueness of his recollection? Most dreams cannot be remembered at all and are forgotten except for some tiny fragments. And is a scientific psychology or a method of treatment for the sick to be founded upon material such as this?