one whose art appears to him to be the most complete. And how does one draw the dividing line between two plots? One does not know when one folktale ends and another begins—“Where one researcher sees a new theme, another will see a variant, and vice versa” (Propp, 1968:9). Finally, a large number of plots remain in Aarne’s collection (20–25%, which seems a great deal) where one does not know to which type they should be ascribed. One must conclude from these failures that a classification of folktales