Moidart and Glengarry, and wore the Macdonnell tartan[12] (1745-1760), and so might come to pass as Macdonnells, though they still regard the Marquis of Ailsa, a Kennedy, as their chief, at least in Eilean Shona (Loch Moidart). In the same way the Camerons of Glen Nevis, though called 'Camerons,' were really MacSorlies, a branch of the Macdonnells, and from the sixteenth century to 1754 were always on ill terms with the chief of the clan Cameron, Lochiel. These are very modern instances, but illustrate Sir Henry's theory of incomers.

The members of the Roman tribes had traditions that they were not, really, of the same original blood with each other. Only by a fiction were they of the same blood. They did not all descend by natural increase from one patriarchal ancestor. There really did exist 'a variety of alien groups in a local tribe,' however they might all adopt the same name, and assert descent, in West Scotland from Somerled, let us say. This fact, of heterogeneousness within the 'tribe' among others, was so obvious and so imperfectly explained, by friends of the Patriarchal theory, that it occupied 'writers belonging to the school of so-called prehistoric inquiry,' as Sir Henry styled it.[13] They were not satisfied with the theory that Society arose in the Patriarchal Family, based on direct descent from, and ownership by, a single male ancestor. To be sure a Cameron will 'cross the hill,' and call himself Stewart, and a Chinese immigrant into Australia has discreetly entitled himself Alexander Mac-gillivray. But such accretions, and such legal fictions, do not explain the heterogeneousness of the local tribe, which, by the theory of some historians, is of common descent. 'Prehistoric inquirers' could not but notice that, among ruder 'non-Aryan' races of various degrees of culture, 'the family is radically different from the Patriarchal Family,' and suggests a different origin.

Roughly speaking, the groups of real or fancied kindred among various low races exhibit the peculiarity that the kin-name is often inherited from the mother, not from the father; that the maternal blood is stronger in determining such cases of inheritance as arise; and that marriage is forbidden within the recognised limits of the maternal kinship. It was natural for inquirers to derive this condition of affairs, this reckoning in the female line, from a state of society in which fatherhood (owing to promiscuity, or to polyandry—several husbands to one wife) was notably uncertain. Bachofen, who first examined the problem, attributed the system to a supposed period of the Supremacy of Women: McLennan to dubious fatherhood, and possible early promiscuity. The recovery of supremacy by men, or the gradual advance in civilisation, especially in accumulation of property, would finally cause descent to be reckoned through the male line, as among ourselves.

As to the question of early promiscuity—sexual relations absolutely unregulated—Dr. Westermarck, Mr. Crawley, and others have argued, and Mr. Atkinson argues, that it never existed, at least to any wide extent, and with any potent influence. We hear rumours of savages utterly promiscuous, say the Mincopies of the Andaman Islands, just as we hear of savages utterly without religion. But later and better evidence proves that the Andamanese have both wives and a God.[14]

Again, the lowest savages known are so far not 'promiscuous,' that they recognise certain sets of women as persons with whom (as a general rule, subject to occasional exceptions) certain sets of men must have no marital relations. It was the opinion of Mr. Darwin, as of Mr. Atkinson, that sexual jealousy, from the first, must probably have been a bar to absolute promiscuity, even among the hypothetical anthropoid ancestors of human race. To tell the truth, our evidence on these points, as to existing savages, is, as usual, contradictory.[15]

WHAT IS EXOGAMY? DIFFICULTIES OF TERMINOLOGY

In these inquiries a great source of confusion arises (as all students must be aware) from the absence of exact terminology, of technical terms with a definite and recognised meaning. Thus when my friend, the late Mr. John Fergus McLennan, introduced the word 'Exogamy,' in 'Primitive Marriage' (1865), he probably knew perfectly well what he meant. But he did not then, from lack of practice in an inquiry practically novel, and originated by himself, express his meaning with exactness. He at first spoke of exogamy as the rule 'which prohibited marriage within the tribe.'[16] But the word 'tribe' was later taken by Mr. McLennan to mean, and is now used as meaning, what cannot be a primitive community, a local aggregate of groups amicably occupying a considerable area of country; say the Urabunna tribe of Central Australia. Mr. McLennan did not wish to say that exogamy forbids an Urabunna tribesman to marry an Urabunna tribeswoman; he meant that exogamy prohibited marriage within the recognised kindred—that is, in this case, between members of totem kindreds of the same name, say Emu or Kangaroo. This fact he later made perfectly clear. But meanwhile such terms as 'horde,' 'tribe,' 'sub-tribe,' 'family,' 'gens,' 'section,' 'phratria,' 'clan,' many of them derived from civilised classical or Celtic usage, have been tossed up and down, in company with 'class,' 'division,' 'section,' and so on, in a way most confusing.[17] Odd new terms come from America, such as 'socialry,' 'tutelaries,' 'ocular consanguinity,' 'ethnogamy,' 'conjugal conation,' and so forth.[18] Most perplexing it is to find words like clan, family, tribe, gens, phratry, words peculiar to civilised peoples, Greek, Roman, or Celtic, applied to the society of savages. 'The term "clan" implies descent in the female line,' says the late Mr. Dorsey, following Major Powell; but why take the Celtic term 'clan,' which has no such signification, and confer it on what is really a totem kindred with descent in the female line?[19] Next, 'several of the Siouan tribes are divided into two, and one into three sub-tribes. Other tribes are composed of phratries, and each sub-tribe or, phratry comprises a number of gentes.' Is there a distinction between the 'sub-tribes' of some tribes, and the 'phratries' of others, or not? Apparently there is not, but the method