mythologies, something must be said about the materials we possess. A strict criticism of the collections of tales offered to the inquirer, a strict avoidance of theory founded on hasty analogies is needful. We must try to distinguish as far as possible what is ancient and essential, from what is relatively modern and accidental in each tale. We must set apart scientific and exact collections from merely literary collections in which the traditional element is dressed up for the sake of amusement. Grimms' collection of Household Tales or Märchen is among the earliest of those which were made for scientific purposes. Sanskrit stories, Arab and Egyptian stories, Italian stories, French stories, had been gathered long before into the garners of Somadeva, The Thousand and One Nights, Straparola, the Queen of Navarre, Perrault, and others. But to bring together popular narratives merely to divert the reader is an aim which permits the collector to alter and adorn his materials almost as much as he pleases. Consequently the old compilations we have named, however delightful as literature, must be used with great caution for purposes of comparative science. Modern touches, as will be seen, occur freely even in such collections as the Grimms'. Science accepts these narratives (when it can get them unadulterated) as among the oldest productions of the human fancy, as living evidence to the character of the early imaginative faculty. But we must be quite certain that we do not interpret late additions to the tales, as if these incidents were of the primitive essence. An example of this error may be taken from Grimms' Legend (No. 5), "The Wolf and the Kids." Here a wolf deceives seven little kids, and eats them all except the youngest, who hides (like the hero of one of M. Fortuné du Boisgobey's novels) "in the clock-case." The bereaved old she-goat comes home; finds that only the youngest kid survives,
