Let us take dreams as a type. The conscious dream may be made up of fantastic imagery and apparently absurd thoughts without apparent logical meaning. Now from general experience we may believe that the dream is a cryptic symbolic expression of a logical subconscious process—perhaps a wish. The question is, what wish? The symbolism cannot be deciphered on its face. Now, by the analytic method associative memories pertaining to each element of the dream are recovered in abstraction. When a memory of antecedent thoughts of which the dream element is a logical symbolism or synonym and which give an intelligent meaning to the dream is recovered, we infer that these antecedent thoughts are contained in the determining subconscious process. Further, as it is found that certain objects or actions (e. g., snakes, flying, etc.) frequently occur in the dreams of different people as symbolisms of the same thoughts, it is inferred that whenever these objects or actions appear in the dream they are always symbolisms of the same underlying thoughts. Obviously the mere fact of an antecedent experience arising as an associative memory is not of itself evidence of its being the causal factor. Hundreds of such memories might be obtained. To have evidential value the memory must give logical meaning to the dream or dream element under investigation. Now, as a matter of fact, more than one memory can often be obtained which answers these conditions. Consequently it becomes a matter of selection from memories, or interpretation, as to which is the correct solution of a given dream problem—and _mutatis mutandis_ of a pathological phenomenon. Naturally the selection is largely determined by personal views and _a priori_ concepts. It also follows that if one accepts the universality of a given symbolism and is committed to a given theory one can, by going far enough, find associations in vast numbers of dreams that will support that theory. The correct solution of a dream problem, that is, the correct determination of the specific underlying process, depends upon the correct determination of the causal factor and this must be inferred. The inferential nature of the latter factor therefore introduces a possible source of error. There must frequently be considerable latitude in the interpretation. This is not to gainsay that in a large number of instances the logical relation between antecedent experiences (recovered by associative memories) and the dream is so close and obtrusive that doubt as to the true subconscious process can scarcely be entertained. An example of a condensed _analysis of a dream_ will illustrate the practical difficulty often presented in determining by clinical methods the correct causal factor and subconscious process of a dream. I select a simple one which consists of two scenes:[112]