The story as he wrote it certainly conveys the impression, not of invented scenes and happenings, but rather of things that had actually been witnessed by the narrator. This may, however, of course, be due, not to the curious manner in which it had its origin, but to the graphic power of the novelist. Very similar to his experience were the dream creations which were so fully described by Robert Louis Stevenson in his essay called "A Chapter on Dreams."1 In this essay, which gives a most lucid account of his whole dream life, he described the process of inventive dreaming from which many of his stories originated. So completely did these dreams seem to him to be an inspiration from outside himself, the operation of faculties apart from the workings of his normal mind and working at a higher level, that he speaks of them as being the handiwork of the "Little People," Brownies of the mind, who - whilst he slept - bestirred themselves to construct and elaborate for him the plots of his stories, far better tales, he declared, than any that he could invent for himself by day. He gives in this essay the outlines of one such story, of which he says truly that it would be hard to better the dramatic effectiveness of its situations.

The plot of the dream story hinged upon the hidden motive of the woman who played the leading part in the little drama; and until its very end that secret was kept.

1 R. L. Stevenson, "Across the Plains".

"The dreamer . . . had no guess whatever at this motive - the hinge of the whole well-invented plot - until the instant of its highly dramatic declaration. It was not his tale; it was the Little People's! And observe; not only was the secret kept, the story was told with really guileful craftsmanship. ... I am awake now, and I know this trade; and yet I cannot better it . . . the more I think of it, the more I am moved to press upon the world my question, who are the Little People? They are near connections of the dreamer's, beyond doubt. . . . They share plainly in his training: they have plainly learned like him to build the scheme of a consistent story and to arrange emotion in progressive order; only I think they have more talent; and one thing is beyond doubt, they can tell him a story piece by piece, like a serial, and keep him all the while in ignorance of where they aim".

What indeed are these "dream-builders"? If we could but answer this question satisfactorily, we should solve the most baffling problems of our dreams. This power of the dream mind not only to construct a dramatic story, but to conceal from us till the very end the denouement to which the story led up, what a mystery it reveals! How does this thing happen? We know that it does happen, for, though our own dreams are far from being such remarkable ones as those described by Stevenson, and though they lack all the craftsmanship and finish of his, yet we too have experienced the same thrill of wonder when a secret carefully hidden from us till the end has suddenly been disclosed.

The suggestion made in the earlier part of this chapter as to the main difference between the ordinary dreams and super-dreams does not here suffice. It is inadequate to account for mental processes such as these. For here, far more even than in our dreams of argument, a dual consciousness or personality seems to be present; here again the curious sense is felt that some one not quite oneself, some one with rather different faculties from one's own, but yet an integral part of self, is at work, taking a hand in the business. Facts that we do not know are in his possession, and the answer to the riddle of the dream story is within the knowledge of this other self though it is hidden from ourselves. Stevenson realised that this other self was intimately related to the dreamer, trained as he himself was trained, but able, he believed, to do something which he himself could not do, or to do it better; and to this same mysterious other self I imagine that the mathematician also owes the dream solution of his problem.

These and many equally curious and interesting experiences which are nowadays occupying the attention of men of science seem to require the as-suniption of a secondary consciousness existing side by side with our ordinary personal conscious-ness, and indeed, unless we can assume the presence of such a divided personality or consciousness, it seems almost impossible to conceive how certain processes of the mind are carried out in the dream state or in the hypnotic state.

The consideration of this most difficult aspect of the subject belongs to a later chapter, and, although it can only be dealt with imperfectly by an observer who has not the necessary knowledge of psychology, it would be impossible to deal at all thoroughly with the question of dreams, unless this aspect is considered.

Before passing on to other questions, it is curious to note the attitude that many people, writers and others, adopt with regard to dream experiences, which, like those described above, are outside the common range of experience, and which do not chance to have come under their own immediate observation. Dreamers who have actually had these uncommon experiences know from first-hand knowledge that, strange as they may seem, such dreams occur and must be taken into account when problems concerning the activity of the mind in sleep, or the possibility of a divided consciousness or dual personality in man, are considered. But whereas a man's careful and straightforward statement about the processes of his waking thoughts would be accepted without questioning, it is more than likely that if he should make an equally careful statement about his dreams he will find that this is looked upon very doubtfully by his fellows. Even a philosopher like M. Bergson, having stated the theory that in general "dreams create nothing," finds it necessary to explain away the case of such creative dreams as Stevenson's by saying that the dreamer was probably in a psychical state in which it would have been difficult to say whether he were asleep or awake, for "when the mind creates, when it is capable of making the effort to organise and synthesise which is necessary in order to triumph over a difficulty, to solve a problem, or to produce a work of imagination, we are not really asleep."1 This summary way of dealing with facts or records which clash awkwardly with theories is noted here only because this attitude is rather often found in books about dreaming, but it is a curious attitude to adopt towards Stevenson's very deliberate and very careful analysis of his dream life over a period of years.

If in pursuing another study we found that carefully recorded facts did not conform to a theory that we had formed, we should probably concede that the theory might be either incorrect or not sufficiently elastic.

"The question whether anything can be known is to be settled not by arguing but by trying,"2 and the inductive method of arriving at truth by means of experiment rather than by logic advocated by the great philosopher is still the method that it is safest to follow if our conclusions are to be sound. It is, of course, the absence of sufficient accurately recorded facts concerning dreams that has made it natural that philosophers should build up their theories concerning them without an adequately wide foundation. And so once again we are brought round to the need for a clearing house of dreams, whose widely gathered stores of observation would be available to correct or to confirm the theories about the working of the dream mind that science may hereafter form.

1 Henri Bergson, "Revue Scientifique," Paris, June, 1001. 2 Bacon, "Novum Organon".