The Origin of Totem Names and Beliefs. 351
a covenant with, and cleave to, a grub, or frog, or lizard, as a supernatural ally. But, for reasons already indicated, in an earlier part of the work whereof this is a chapter, I con- ceive that, originally, totems descended in the female line only. One reason for this opinion is that, as soon as descent of the totem comes by the male line, a distinct step in the upward movement towards civilisation and a settled life is made. It is not very probable that the backward step from reckoning by male lineage to descent in the female line, has often been taken. On the other hand, tribes which now inherit the totem in the male line, exhibit in their institutions many survivals of female descent. An instance is that of the Mandans, as recorded by Mr. Dorsey.' Among the Melanesians, where female descent still exists, there is at work the most obvious tendency towards descent through males, as Dr. Codrington proves in an excellent work on that people. Dr. Durkheim, too, has pointed out the traces of uterine descent among the Arunta, who now reckon in the male line.^ On the other hand, where we find descent in the male line, I am not aware that we discover signs of movement in the opposite direction. In this opinion that, as a general rule, descent was reckoned in the female, not the male line, originally, I have the support of Mr. E. B. Tylor.^ For these reasons the hypothesis of the selection of and covenant with a " supernatural ally," plant or animal, by the deliberate joint action of an early group, at a given moment, involving staunch adherence to the original reso- lution, rather strains belief ; and a suggestion perhaps more plausible will be offered later. But, given the general savage attitude towards the lower and inanimate creation, this theory of the original choice and retention of a plant or animal guardian by a primitive group seems to be one which may be held wdthout disrespect to logic. At all events, it
' Bureau of Ethnology, 1893-94, p. 241.
- UAnn^e Sociologique, v., pp. 93, 99, 100. As far as the proof rests on Arunta traditions, I lay no stress upon it.
- _/. A. /., vol. xviii., No. 3, p. 254.
