As far as I can see, the division of a dream into two unequal portions does not always signify a causal relation between the thoughts of the two portions. It often seems as though in the two dreams the same material were presented from different points of view; this is certainly the case when a series of dreams, dreamed the same night, end in a seminal emission, the somatic need enforcing a more and more definite expression. Or the two dreams have proceeded from two separate centres in the dream-material, and they overlap one another in the content, so that the subject which in one dream constitutes the centre cooperates in the other as an allusion, and vice versa. But in a certain number of dreams the division into short preliminary dreams and long subsequent dreams actually signifies a causal relation between the two portions. The other method of representing the causal relation is employed with less comprehensive material, and consists in the transformation of an image in the dream into another image, whether it be of a person or a thing. Only where this transformation is actually seen occurring in the dream shall we seriously insist on the causal relation; not where we simply note that one thing has taken the place of another. I said that both methods of representing the causal relation are really reducible to the same method; in both cases causation is represented by succession, sometimes by the succession of dreams, sometimes by the immediate transformation of one image into another. In the great majority of cases, of course, the causal relation is not represented at all, but is effaced amidst the succession of elements that is unavoidable even in the dream-process.