of which the members are of four or five different totems. As this is the case, how can local totemism prevail? In ten families in a given region there might be people of at least twenty different totems. Are we to suppose that sires of one totem choose to flock together, for purposes of working magic, and that their totem is the local totem, while those of their wives and children are negligible? We are not told by Mr. Spencer that all this occurs, but, if it does not, how have the Arunta managed to possess local totem groups?
I am not sure that the difficulty of accounting for the existence of local totem groups among a people whose totems come by accident has been remarked upon. The fact has always puzzled me, for how can there be a local wild cat or emu group in a district of which the inhabitants are necessarily of many different totems? If we can suppose that several adult men of the wild cat or emu totem have congregated in the locality as a magicworking society, the problem is solved,—but where is the evidence for this intentional combination? Perhaps I have overlooked some other solution, and I merely draw attention to the difficulty which the situation presents.
(Vol. xix., p. 288.)
As regards present-day survival of amulets for protection against diseases, I may note that I still find them in use, but not commonly. Red silk round the wrist for rheumatism I have seen in Fife,—(although the wearer was a "gangrel" and perhaps had Irish blood in his veins),—and in Aberdeenshire of old days red worsted would be tied round a child's wrist to keep away the "witches." The chief thing, however, for the latter purpose was the little heart-shaped silver "witch-brooch." It was pinned to the child's underclothing at its first dressing. The shape was probably derived from its being originally the mounting of an "elf shot" or "fairy dart," i.e. flint arrow head. An old man in Kincardineshire some thirty