Page:Folk-lore - A Quarterly Review. Volume 11, 1900.djvu/242

 220 endeavoured to show in a discussion of Welsh superstitions, be justifiably included among the "categories," in spite of the apparent multiplicity of possible origins of this ominous character of animals. I include in this division further the use of the animal in (a) magic, (b) medicine, (c) divination.

III.— Annual Ceremonies.

I. Sacrifice; 2. Communion with the sacred animal.

This analysis is, I think, fairly exhaustive; it is in the main based on the superstitions actually found among totems. There are probably few, if any, European superstitions relating to animals that cannot be classified according to the above scheme. As evidence of the former existence of totemism in Europe, the sections are, however, of unequal value, and I propose to deal with the more important ones only.

This will naturally not permit me to show the cumulative character of the evidence with regard to single animals. More important, however, than this cumulative character is the local distribution of the superstitions with which I deal; and I prefer to emphasize this feature rather than to analyse all the superstitions relating to a small number of animals.

I lay down at the outset the principle that no theory of the origin of animal superstitions can disregard this local character of the beliefs; one parish respects an animal and will not kill it; their neighbours regard the same animal with indifference, and even aversion. Even where no such striking contrast is present, the local character of the superstition is always very prominent. Yet in spite of this we find such an agreement between distant regions, both in custom and belief, as to exclude the idea of a purely local origin of the superstitions. Special explanations, alluring enough when only the single case is considered, are manifestly impossible when we have to deal with a great mass of