Hammond reports another case where coitus maritalis could be performed only by the help of a certain costume; and Dr. Moll mentions several similar cases in individuals of hetero- and homo-sexuality. The cause may often be shown to be an early association, and such may always be assumed. It is only in this way that one can explain why a certain costume cannot be resisted by such individuals, no matter what person wears the fetich. Thus one can understand why, as Coffignon (_op. cit._) relates, men at brothels demand that the women with whom they are concerned put on certain costumes, such as that of a ballet-dancer, or nun, etc.; and why these houses are furnished with a complete wardrobe for such purposes. Binet (_op. cit._) relates the case of a judge who was exclusively in love with Italian girls who came to Paris as artists’ models, and their peculiar costume. The cause was here demonstrably an impression made at the time of the awakening of the sexual instinct. A third form of dress-fetichism, having a much higher degree of pathological significance, is by far the most frequent. In this form it is no longer the woman herself, dressed, or even dressed in a particular fashion, that constitutes the principal sexual stimulus, but the sexual interest is so concentrated on some certain article of female attire that the lustful idea of this object is entirely separated from the idea of woman, and thus obtains an independent value. This is the real domain of dress-fetichism, where an inanimate object—an isolated article of wearing-apparel—is alone used for the excitation and satisfaction of the sexual instinct. This third form of dress-fetichism is also the one that is important forensically. In a large number of these cases the fetiches are articles of female underwear, which, owing to their private use, are suited to occasion such associations.