the journey, as actually happened in the morning when I woke with the sensation of a bodily need. I suppose one might be inclined to credit this sensation with being the actual stimulus of the dream; I should, however, prefer a different explanation, namely, that the dream- thoughts first gave rise to the desire to urinate. It is quite unusual for me to be disturbed in sleep by any physical need, least of all at the time when I woke on this occasion- a quarter to four in the morning. I would forestall a further objection by remarking that I have hardly ever felt a desire to urinate after waking early on other journeys made under more comfortable circumstances. However, I can leave this point undecided without weakening my argument.

Further, since experience in dream-analysis has drawn my attention to the fact that even from dreams the interpretation of which seems at first sight complete, because the dream-sources and the wish- stimuli are easily demonstrable, important trains of thought proceed which reach back into the earliest years of childhood, I had to ask myself whether this characteristic does not even constitute an essential condition of dreaming. If it were permissible to generalize this notion, I should say that every dream is connected through its manifest content with recent experiences, while through its latent content it is connected with the most remote experiences; and I can actually show in the analysis of hysteria that these remote experiences have in a very real sense remained recent right up to the present. But I still find it very difficult to prove this conjecture; I shall have to return to the probable role in dream-formation of the earliest experiences of our childhood in another connection (chapter VII).

Of the three peculiarities of the dream-memory considered above, one- the preference for the unimportant in the dream-content- has been satisfactorily explained by tracing it back to dream distortion. We have succeeded in establishing the existence of the other two peculiarities- the preferential selection of recent and also of infantile material- but we have found it impossible to derive them from the motives of the dream. Let us keep in mind these two characteristics, which we still have to explain or evaluate; a place will have to be found for them elsewhere, either in the discussion of the psychology of the sleeping state, or in the consideration of the structure of the psychic apparatus- which we shall undertake later after we have seen that by means of dream-interpretation we are able to glance as through an inspection- hole into the interior of this apparatus.

But here and now I will emphasize another result of the last few dream-analyses. The dream often appears to have several meanings; not only may several wish-fulfilments be combined in it, as our examples show, but one meaning or one wish-fulfilment may conceal another. until in the lowest stratum one comes upon the fulfilment of a wish from the earliest period of childhood; and here again it may be questioned whether the word often at the beginning of this sentence may not more correctly be replaced by constantly. * -

*The stratification of the meanings of dreams is one of the most delicate but also one of the most fruitful problems of dream interpretation. Whoever forgets the possibility of such stratification is likely to go astray and to make untenable assertions concerning the nature of dreams. But hitherto this subject has been only too imperfectly investigated. So far, a fairly orderly stratification of symbols in dreams due to urinary stimulus has been subjected to a thorough evaluation only by Otto Rank.

C. The Somatic Sources of Dreams

In the introductory chapter we examined at length the opinion of scientific writers on the role of somatic stimuli in the formation of dreams, so that here we need only recall the results of this inquiry. We have