well trained observers among these tribes of the North Pacific border have shown with great clearness,"—something not at present to the point[31] But we regret the absence of clearness. Can we rely on Holmberg who described the state of affairs as it was fifty years ago, and who knew nothing, I presume, of Australian phratries and totem kins? In his time the Tlingit, like a dozen South-Eastern tribes of Australia, had animal-named kins in animal-named exogamous intermarrying phratries with female descent. Or was Holmberg (and was Mr. F. Boas in his list of animal-named Tlingit clans) led astray by the "crests"? Did each of these inquirers mistake "crests" for totems of clans?

One thing is clear, the Tlingit and the other tribes being possessed of wealth, and of gentry, and of heraldry, cause almost inextricable confusion by their use of heraldic badges, named "crests" by some; and "totems" (or both crests and totems at once) by other well trained observers. I am inclined to believe that most of these crests were, originally, representations of the totems of distinct totem kins. My reason is this: Mr. Swanton tells us that "the crests and names which among the Tlingit are on the Wolf side" are "on the Raven side" among the Haida. Among these people, animal names and crests are divided between the two phratries, the same name or crest not occurring in both phratries. This is merely the universal arrangement of totems in phratries.

Even now, among the Tlingit, says Mr. Swanton, "theoretically the emblems" (crests) "used on the Raven side were different from those on the Wolf or Eagle side," (precisely as, in Australia, the totems in Eagle Hawk phratry are different from those in Crow phratry), "and although a man of high caste might borrow an emblem from his brother-in-law temporarily, he was not permitted to retain it" (His brother-in-law, of course, was of the phratry not his own.) All this means no more than that occasionally a man of high caste may now impale the arms of his wife.[32] With castes and heraldry, born of wealth and rank, we have stepped out of totemism at this point It has been modified by social conditions. "Some families were too poor to have an emblem," did they also cease to have a totem? Some of the rich "could," it was said, "use anything." Is this because they pile up sixteen quarterings? "The same crest may be, and is, used by different clans, and any one clan may have several crests...."[33] Many "clans" now use the same crest, and there are quarrels about rights to this or that "crest." Some members of the Wolf phratry assert a right to the Eagle crest. Mr. Frazer thinks that "such claims are perhaps to be explained by marriages of the members of the clan with members of other clans who had these animals for their crests."[34]

That is precisely my own opinion. If "crests" were originally mere representations of each person's totem animal they have now become involved, through rank and social degrees, with heraldry, and with badges not totemic, such as a certain mountain. Meanwhile all the Tlingit "clans," if we follow Mr. Swanton's evidence, or almost all the "clans" are now mere local settlements, at least they bear local and other descriptive names. I nearly despair of arriving at Mr. Swanton's theory of what a Tlingit "clan" really is! But he gives a list of "the geographical groups," the "clans," and the phratry to which each of the clans belonged....

Thus we have (1)

PHRATRY RAVEN.

Then (2)

TONGAS (I take Tongas to be "a geographical group").

Then under TONGAS GānAXA'di, People of Gā'NAX.

PHRATRY WOLF.

TONGAS (Geographical group, apparently).
Te'goedî, People of the island Teq°.

GānAXÁdî and Te'goedî seem to be "clans," but then clan Te'goedî, "People of the isle Teq°," looks like "a geographical group"!

There are fourteen "geographical divisions" of this kind, and sixty-eight "clans" of this kind, with descriptive or local names. The clans "were in a way local groups," says Mr. Swanton. They were also "clans or consanguineal bands," each "usually named from some town or camp it had once occupied." They "differed from the geographical groups ... being social divisions instead of comprising the accidental occupants of one locality."[35]

Be it observed that Mr. Swanton speaks of "these geographical divisions or tribes"; which increases the trouble, for, if the Tlingit be a "tribe," and the geographical divisions of the Tlingit be also "tribes," things are perplexing.

Once more, the Tlingit reckon descent in the female line. Now how can "a consanguineal band," which reckons descent in the female line, look like "a geographical group"? A totem kin, with male descent, in Australia and elsewhere, like a Highland clan, say the MacIans, necessarily becomes "a geographical group," say in Glencoe. But how, with female descent (unless the women go to the men's homes), a Tlingit "consanguineal band" can also have a local habitation is to me a difficult question. The names of the phratries descend in the female line. Do the local and descriptive names of "the clans or consanguineal bands," also descend in the female line? I cannot presume to say. Mr. Frazer throws no light on this point believing, as he does, that the "clans" within the Tlingit phratries, are the familiar totem kins, of animal names. If so, the children must inherit the maternal totem "clan" name.

Only one thing is clear to me, a Tlingit of the Wolf phratry can only marry a bride of the Raven phratry; a Tlingit of the Raven phratry can only woo a maiden of the Wolf phratry. If totem kins there be in the phratries, these totem kins are exogamous. If there be no totem kins in the phratry, are Mr. Swanton's clans of local names locally exogamous? May persons marry within the region where they are settled? I know not, but I rather incline to suppose that members of both phratries may be found in Mr. Swanton's clans of local name; indeed it must be so, and therefore a pair of lovers may perhaps wed within their "clan or consanguineal band," and within their local group, which, thus, is not exogamous. If so, the Tlingit clan is not exogamous. But all this is purely conjectural.

While, in Mr. Swanton's version, the Tsimshians, with female descent, have two exogamous "clans" with animal names, and two with other names; while in Mr. Frazer's book they have four animal-named exogamous clans, there is a third story resting on the authority of Mr. William Duncan, a missionary among the Tsimshian from 1857 onwards.[36] Mr. Duncan's information Commander Mayne incorporated in his book.