how could this idea occur to members of an undivided commune,' who had never known anything better?

That is the difficulty; and we get rid of it by disbelieving in a primeval undivided commune; and by supposing a long past of forbidden unions, the prohibition then resting on no moral ideas, but on the interest of the strongest, the jealousy of the adult sire. These prohibitions later evolved into conscious morality; and were at last susceptible of improvement by deliberate design. I shall now examine more in detail the ideas which do not win my assent.

MR. FISON ON THE GREAT BISECTION

In 1880, in Kamilaroi and Kurnai,1 Mr. Fison, a learned missionary and anthropologist, gave his account of the organisation of certain Australian tribes. He speaks of (1) The division of a tribe, or community, into two exogamous intermarrying classes. (2) The subdivision' (mark the phrase) of these classes into four.' (3) Their subdivision into gentes, distinguished by totems, which are generally, though not invariably, the names of animals.'

2

Now totems we know, and we have cited Mr. Mathews for the other divisions. Take (1) the two exogamous intermarrying classes.' Examples are

Male, Kumite; female, Kumitegor (one class,' which I call 'phratry'). Male, Kroki; female, Krokigor (the other class,' 'phratry ').

Again.

Male, Yungaru (opossum); female, Yungaruan.
Male, Wutaru (kangaroo); female, Wutaruan.

What are these two 'primary' exogamous divisions? And why call them 'primary'?

1 P. 27 et seq.

2 There is a tradition of an aboriginal Adam, who had two wives, Kilpara and Mukwara, these being the names of two phratries. On this showing brothers married paternal half-sisters (Kamilaroi and Kurnai, p. 33).

PRIMARY CLASSES?'

My object, as has been said, is now, contrary to general opinion, to repeat that the great dichotomous 'division' of a tribe into two exogamous, intermarrying, 'classes' or ' phratries,' is not 'primary' at all, but is secondary to groups at once totemic and exogamous, and is not, in origin, a bisection, but a combination. If I am right, the consequences will be of some curiosity. First, it will appear that the 'primary divisions' are themselves totemic in origin, thus implying the pre-existence of Totemism. Next it will be made to appear probable that the pre-existing totems were already exogamous before the phratries arose, and that exogamy does not date, as the best authorities hold, from the making of the great dichotomous divisions or 'phratries.' For no such dichotomous division, I suggest, was ever made.

THE PRIMARY DIVISIONS' ARE THEMSELVES

TOTEMIC AND EXOGAMOUS

We see that, of the two 'phratries' Yungaru and Wutaru, Yungaru is opossum (according to Mr. Chatfield) or 'alligator' (according to Mr. Bridgman); while Wutaru is 'kangaroo.' These two primary phratries,' therefore, have totemic names, and (in my opinion) were originally two local totem groups, each containing members of various totems derived from alien mothers. The same thing may be true when the meanings of the primary class names' ('phratries') can no longer be discovered. If so, the 'primary divisions' are, in origin, mere totem distinctions, involving, I think, the pre-existence of the rule of exogamy, which is also involved in the rules of the 'primary divisions.' Mr. Fison writes (what is obvious) 'in some places the primary divisions are distinguished by totem names at the present day.'1

“

1 Kamilaroi and Kurnai, p. 40.

6

Probably they were so distinguished everywhere, in ancient times,' he adds, and this is certainly the case in North America, as we shall see later. Mr. Fison's opinion is my own so far, and, if it is right, if the 'primary class divisions' ('phratries'), within which marriage is now forbidden, were originally two totem divisions, then Totemism is earlier than the primary divisions.' On this point Messrs. Fison and Howitt say that the divisions on which marriage regulations are based are denoted by class names or by totemsfrequently by both class names and totems.' In a note they add, Class names, so called by us solely for the sake of convenience, and because they cannot always be positively asserted to be totems, though the strong probability is that they are always totems.'1

6

6

By class names' the authors, I think, here mean the names of the primary exogamous divisions' or 'phratries.' These are often, if not always, known by totem names. But the classes,' as distinguished from the 'phratries,' are not known by totemic names, as far as I am aware. Herr Cunow, we shall see, asserts that in some cases they denote mere seniority, 'big' and 'little,' 'young' and 'old.' Unless they can be proved to be totemic, we must, I repeat, carefully avoid confusing the classes,' four or eight, with the phratries,' in which they are included. The confusion is general and very misleading.

Totemism, according to Mr. McLennan, preceded exogamy, and made exogamy possible. Thus totem distinctions, with exogamy, may be older than the two primary class exogamous divisions,' in which, according to most authorities, exogamy began. Mr. Tylor is cautious: the dual form of exogamy' (the 'phratries,' or two primary divisions') 'may be the original form,' or at least that view is tenable. The origin of exogamy is, however, unknown, in Mr. Tylor's opinion, which commits him to nothing.

Mr. Howitt, if I do not misinterpret him, also regards the

1 J. A. I. xiv. 142.

Journal of the Anthropological Institute, xviii. 264.

two divisions, 'phratries,' as primary, but at the same time agrees with me, and Mr. Fison, that the two 'phratry' divisions were themselves in origin totemic.

THE TOTEM DIFFICULTY

At this point I lose Messrs. Fison and Howitt. I do not know what they mean, and, unless I misconstrue them, they unconsciously hold different opinions at different moments. They start with an undivided commune.' Mr. Fison, however, is not certain on this point. To prevent near marriages (previously universal), the commune is split into two exogamous intermarrying phratries. The names of these phratries are totemic, and each phratry has its totem. Such is their theory. How and why?

6

Or

Did totemic divisions already exist in the undivided commune'? If so, the commune was not undivided! were totem names given, nobody knows why, to the two phratries at the time when the 'bisection' of the commune was made? Did the legislator send half the horde to the right, crying, 'You are sheep,' and half to the left, saying, 'You are goats,'-or rather, say, Emus and Kangaroos? This is not easily thinkable. But, if this was done, whence came the other totem kins, often numerous, within each phratry?

1

Mr. Fison says that the totem kins (or 'gentes') ' arose out of two primary divisions by an orderly process of evolution, such as might be expected from the forces at work,' and 'we have seen how' the phratries subdivided into other subdivisions, distinguished by totems.' But, alas, I have seen nothing of the sort! Mr. Fison has merely asserted the fact. The totems affect the intersexual regulations . .. by narrowing the range of matrimonial selection.' Here would be a reason for the evolution of these totem kins. But this added restriction is exactly what (given phratries) the totems do not effect. There are so many totems in each phratry, but 2 Op. cit. p. 41.

1 Kamilaroi and Kurnai, p. 107.

2

as the same totem (except among the Arunta and similarly disorganised tribes) never occurs in both phratries, the range of sexual selection is thus not more restricted by the totem than by the phratry. The members of each phratry may not intermarry, and all persons of their totem are in their phratry and so are not marriageable to them. They would all be exactly as exogamous as they are, if there were no totem rules, nothing but phratry rules. Thus the totems cannot be later deliberate segmentations of the phratry, for additional exogamous purposes, because they serve no such purpose, except where, among the Kamilaroi, a man may marry in his phratry, if he marries out of his totem. But that is a peculiarity.

Mr. Mathews writes, Under the group' (phratry) 'laws it is impossible for a Dilbi or Kupathin' (phratry names of the Kamilaroi) 'to marry a woman bearing the same totem name as himself, for the reason that such a totem does not exist in the division' (phratry) from which he is bound to select his wife. But when persons of the same group' (phratry) were permitted to marry each other, it became necessary to promulgate a law prohibiting marriage between persons of the same totem.'1 But there were totems before that novelty of marriage within the phratry, and why were they there? Moreover, under phratry laws it was already the rule that no man could marry a woman of his own totem. Obviously we are not toid how the totem kins arose out of the phratries, by an orderly process of evolution such as might be expected from the forces at work.' One sees no reason at all for the rise of totem kins within the phratry, itself, by Mr. Fison's theory, originally totemic.

Totem kins are called subdivisions' by Mr. Howitt, but why were the phratries subdivided into totem kins, and why were there totem groups in the undivided commune' before the bisection, the phratries (the result of the bisection) being themselves, in Mr. Howitt's hypothesis, totem groups? I quote a statement of the case by Mr. Howitt (1889): 'The 1 Proc. Roy. Soc. N.S. W. xxxi. 162.