Society Publication for 1893; see notes on "Cinder-Maid"). These variants of incidents obviously resemble the variæ lectiones of MSS. and naturally suggest the possibility of getting what may be termed the original readings. In 1889 the following suggestion was made by Mr. (now Sir) James G. Frazer in an essay on the "Language of Animals," in the Archælogical Review, i., p. 84:
"In the case of authors who wrote before the invention of printing, scholars are familiar with the process of comparing the various MSS. of a single work in order from such a comparison to reconstruct the archetype or original MS., from which the various existing MSS. are derived. Similarly in Folk-Lore, by comparing the different versions of a single tale, it may be possible to arrive, with tolerable certainty, at the original story, of which the different versions are more or less imperfect and incorrect representations."
Independently of Sir James Frazer's suggestion, which I have only recently come across, I have endeavoured in the present book to carry it out as applied to a considerable number of the common formulæ of European folk-tales, and I hope in a succeeding volume to complete the task and thus give to the students of the folk-tale as close approach as possible to the original form of the common folk-tales of Europe as the materials at our disposal permit.
My procedure has been entirely similar to that of an editor of a text. Having collected together all the variants, I have reduced them to families of types and from these families have conjectured the original concatenation of incidents into plot. I have assumed that the original teller of the tale was animated by the same artistic logic as the contemporary writers of Contes (see notes on "Cinder-Maid," "Language of Animals"), and have thus occasionally introduced an incident which seemed vital to the plot, though it occurs only in some of the families of the variants. My procedure can only be justified by the success of my
