application to the exogamic classes or phratries by Dr. Frazer with what I venture to think is unanswerable force. But that which is true of the exogamic classes may be true also of the totemic clans. They may owe their genesis to a similar but earlier movement. The new groups thus created would generally have attracted to themselves names, whether assumed from within or imposed from without. Such names would as a rule have been obtained from familiar objects, — animals or plants. Once a name was identified with a group, it would form a powerful aid to permanence. It would consolidate it ; it would present it as a tangible entity, not a mere concept. It would become a centre for superstitions, — some perhaps already developed, and others still vague and only half-defined. Where this was the case, these superstitions would coalesce about the new social unit, their evolution would be assisted, and their objects would become defined, belief and practice growing together and forming ultimately the complex organization we call totemism. Exogamy may thus be more vitally related, among some peoples at all events, to totemism than Dr. Frazer's researches appear to show.
Elsewhere than in Australia we find voluntary fission expressly initiated in order to provide mates for members of exogamous clans. In a certain district of Sumatra, where the population is organized in exogamous clans on the basis of matrilineal descent, when a clan has grown too big, as sometimes happens, and the prohibition of marriage within it has therefore become incon- venient, it is divided into two or more smaller exogamous clans, and thus the difficulty of want of spouses for its members is overcome.^ In East Africa, Mr. Ernest Haddon records a native tradition of the separation of the Nyori clan from the rest of the Bari in consequence of a dispute. It left the others, and journeying to the east entered the Nile valley. The Bari clans were exo- gamous ; but, as the other clans were now far distant, exogamous marriages were impossible to the Nyori. So the elders decided that the Nyori clan must itself be broken up into exogamous clans. This was done accordingly.- The result was what was
Globus, vol. xcvi., p. 263. '^/ou7-nalofihc Africa)! Society, vol. x., p. 468.