first suggested by Balzac that Perrault's pantoffles de verre was due to his misunderstanding of the pantoffles de vair, or fur (the word vair is still used to indicate this in heraldry), which he had heard from his nurse or other folk-tale informant. But the step-sisters would not have been compelled to hack their heels to get inside a fur slipper, and, from this point of view, the glass shoe would be preferable. I have had, however, to reject it because it occurs in only six of the variants obviously derived directly, or indirectly, from Perrault. The majority of the versions prefer gold (see Miss Roalfe Cox's enumeration p. 342).
The Shoe Marriage Test again involves the previous meetings of the high-born lover and the menial heroine, transformed for the nonce by her dress into a dame of equal standing. In some of the variants these meetings are in church and not at a ball, royal or otherwise. But the Shoe Marriage Test involves a highly desirable parti who can practically command any wife he desires; this points to some super-chief or king. I have, therefore, reserved the church meetings for the Catskin type of story in which the heroine is scullery-maid in the young lord's own household. The obtaining of the dresses needed for the Royal Balls involves some animal or supernatural aid (in Perrault it is, of course, a fairy god-mother, unknown to the folk mind), while the menial condition of the heroine is best explained in the usual folk-tale manner by the envious step-mother or sisters.
I have pointed out in English Fairy Tales (Note to "Childe Rowland") that in most folk-tales of a romantic type the mode of telling is by prose narrative interspersed with rhyming formulas analogous to the cante-fable as in "Aucassin and Nicolete." The Cinderella formula shows clear traces of such rhymes, especially at the stages of the narrative where incidents are repeated—the appeal for aid at the mother's grave (Dress Rhyme), the avoidance of