oven or boiling him in a kettle; in others by sending him into an overheated sweat-lodge or placing him near a large fire. More important differences may be observed in the general setting of the test tales, which in some areas are tests of the son-in-law; in others, matches between the inhabitants of a village and their visitors.
Other examples of the local development of the plot of a story by the introduction of specific incidents occur, as in the north Pacific coast story of raven killing the deer, whom, according to the Alaskan tale, he strikes with a hammer, while in the more southern form he pushes him over a precipice. Similarly, in a story of a rejected lover who is made beautiful by a supernatural being the magic transformation is accomplished in the northern versions by bathing the youth in the bathtub of the supernatural being, while in the south he is given a new head.
In other cases the geographical differentiation of the tales is not quite so evident, because different types of stories overlap. This is the case in the widely spread story of the deserted child. Tales in which a youth gives offense by being lazy or by wasting food belong to Alaska. Another tpe, in which a girl is deserted because she has married a dog, belongs to British Columbia; but the two types overlap in distribution. This particular theme occurs in a much wider area on the American Continent, and other types may easily be recognized in the stories of the Plains Indians.
Tales of marriages with supernatural beings or animals are often found in the form of the abduction of a girl who has unwittingly offended an animal. This type seems to belong primarily to Alaska, while the theme of helpful animals that succor unfortunate and innocent sufferers is much more frequent among the tribes of British Columbia.
All these examples illustrate that there are a number of simple plots, which have a wide distribution, and which are elaborated by a number of incidents that must be interpreted as literary devices peculiar to each area. In all these cases the incidents obtain their peculiar significance by being worked into different plots.
On the other hand, we find also certain incidents that have a very wide distribution and occur in a variety of plots. Many examples of these are given in the annotations to all the more important recent collections of folk-tales. The local character of folk-tales is largely determined by typical associations between incidents and definite plots.
In most of the cases here discussed the plot has a general human character, so that the processes of invention and diffusion of plots must be looked at from a point of view entirely different from that to be applied in the study of invention and diffusion of incidents. The latter are, on the whole, fantastic modifications of every-day
