as it does not especially concern our subject, it will be discussed in strict proportion to its use here. Nothing is easier than limiting ourselves. Our task is reducible to a very clear and very brief question: What are the forms of association that give rise to new combinations and under what influences do they arise? All other forms of association, those that are only repetitions, should be eliminated. Consequently, this subject can not be treated in one single effort; it must be studied, in turn, in its relations to our three factors—intellectual, emotional, unconscious.

It is generally admitted that the expression "association of ideas" is faulty.[8] It is not comprehensive enough, association being active also in psychic states other than ideas. It seems indicative rather of mere juxtaposition, whereas associated states modify one another by the very fact of their being connected. But, as it has been confirmed by long usage, it would be difficult to eliminate the phrase.

On the other hand, psychologists are not at all agreed as regards the determination of the principal laws or forms of association. Without taking sides in the debate, I adopt the most generally accepted classification, the one most suitable for our subject—the one that reduces everything to the two fundamental laws of contiguity and resemblance. In recent years various attempts have been made to reduce these two laws to one, some reducing resemblance to contiguity; others, contiguity to resemblance. Putting aside the ground of this discussion, which seems to me very useless, and which perhaps is due to excessive zeal for unity, we must nevertheless recognize that this discussion is not without interest for the study of the creative imagination, because it has well shown that each of the two fundamental laws has a characteristic mechanism.

Association by contiguity (or continuity), which Wundt calls external, is simple and homogeneous. It reproduces the order and connection of things; it reduces itself to habits contracted by our nervous system.

Is association by resemblance, which Wundt calls internal, strictly speaking, an elementary law? Many doubt it. Without entering into the long and frequently confused discussions to which this subject has given rise, we may sum up their results as follows: In so-called association by resemblance it is necessary to distinguish three moments—(a) That of the presentation; a state A is given in perception or association-by-contiguity, and forms the starting point. (b) That of the work of assimilation; A is recognized as more or less like a state a previously experienced. (c) As a consequence of the coëxistence of A and a in consciousness, they can later be recalled reciprocally, although the two original occurrences A and a have previously never existed together, and sometimes, indeed, may not possibly have existed together. It is evident that the crucial moment is the second, and that it consists of an act of active assimilation. Thus James maintains that "it is a relation that the mind perceives after the fact, just as it may perceive the relations of superiority, of distance, of causality, of container and content, of substance and accident, or of contrast between an object, and some second object which the associative machinery calls up."[9]

Association by resemblance presupposes a joint labor of association and dissociation—it is an active form. Consequently it is the principal source of the material of the creative imagination, as the sequel of this work will sufficiently show.

After this rather long but necessary preface, we come to the intellectual factor rightly so termed, which we have been little by little approaching. The essential, fundamental element of the creative imagination in the intellectual sphere is the capacity of thinking by analogy; that is, by partial and often accidental resemblance. By analogy we mean an imperfect kind of resemblance: like is a genus of which analogue is a species.

Let us examine in some detail the mechanism of this mode of thought in order that we may understand how analogy is, by its very nature, an almost inexhaustible instrument of creation.

1. Analogy may be based solely on the number of attributes compared. Let a b c d e f and r s t u d v be two beings or objects, each letter representing symbolically one of the constitutive attributes. It is evident that the analogy between the two is very weak, since there is only one common element, d. If the number of the elements common to both increases, the analogy will grow in the same proportion. But the agreement represented above is not infrequent among minds unused to a somewhat severe discipline. A child sees in the moon and stars a mother surrounded by her daughters. The aborigines of Australia called a book "mussel," merely because it opens and shuts like the valves of a shellfish.[10]

2. Analogy may have for its basis the quality or value of the compound attributes. It rests on a variable element, which oscillates from the essential to the accidental, from the reality to the appearance. To the layman, the likeness between cetacians and fishes are great; to the scientist, slight. Here, again, numerous agreements are possible, provided one take no account either of their solidity or their frailty.

3. Lastly, in minds without power, there occurs a semi-unconscious operation that we may call a transfer through the omission of the middle term. There is analogy between a b c d e and g h a i f through the common letter a; between g h a i f and x y f z q through the common letter f; and finally an analogy becomes established between a b c d e and x y f z q for no other reason than that of their common analogy with g h a i f. In the realm of the affective states, transfers of this sort are not at all rare.

Analogy, an unstable process, undulating and multiform, gives rise to the most unforeseen and novel groupings. Through its pliability, which is almost unlimited, it produces in equal measure absurd comparisons and very original inventions.

After these remarks on the mechanism of thinking by analogy, let us glance at the processes it employs in its creative work. The problem is, apparently, inextricable. Analogies are so numerous, so various, so arbitrary, that we may despair of finding any regularity whatever in creative work. Despite this it seems, however, reducible to two principal types or processes, which are personification, and transformation or metamorphosis.

Personification is the earlier process. It is radical, always identical with itself, but transitory. It goes out from ourselves toward other things. It consists in attributing life to everything, in supposing in everything that shows signs of life—and even in inanimate objects—desires, passions, and acts of will analogous to ours, acting like ourselves in view of definite ends. This state of mind is incomprehensible to an adult civilized man; but it must be admitted, since there are facts without number that show its existence. We do not need to cite them—they are too well known. They