were the ancient Egyptians, who, if they belonged to the district of
the sheep, might not eat mutton, which their neighbours, the
Lycopolitae, devoured at pleasure. These restrictions appear to be
connected with the almost universal dislike of cannibals to eat
persons of their own kindred except as a pious duty. This law of
the game in cannibalism has not yet been thoroughly examined, though
we often hear of wars waged expressly for the purpose of securing
food (human meat), while some South American tribes actually bred
from captive women by way of securing constant supplies of permitted
flesh.[3] When we find stocks, then, which derive their names from
animals and decline to eat these animals, we may at least SUSPECT
that they once claimed kinship with the name-giving beasts. The
refusal to eat them raises a presumption of such faith. Old
Bosman[4] had noticed the same practices. "One eats no mutton,
another no goat's flesh, another no beef, swine's flesh, wild fowl,
cocks with white feathers, and they say their ancestors did so from
the beginning of the world."


[1] The evidence of native interpreters may be viewed with
suspicion. It is improbable, however, that in 1817 the
interpreters were acquainted with the totemistic theory of
mythologists, and deliberately mistranslated the names of the
stocks, so as to make them harmonise with Indian, Australian, and
Red Indian totem kindreds. This, indeed, is an example where the
criterion of "recurrence" or "coincidence" seems to be valuable.
Bowditch's Mission to Ashantee (1873), p. 181.

[2] This view, however, does not prevail among the totemistic
tribes of British Columbia, for example.