I venture to prefer, as already indicated, the system (3) that totem groups not only existed, but were already exogamous, before the great 'bisection' producing the 'phratries' came into existence, though I argue that 'bisection' is a misleading term, and that the apparent division was really the result of an amalgamation of two separate and independent local totem groups.
This theory (presently to be more fully set forth) is original on my part, at least as far as my supraliminal consciousness is concerned. I mean that I conceived myself to have hit on the idea in July 1902. But something very like my notion (I later discovered) had been printed by Dr. Durkheim, and something not unlike it was propounded by Herr Cunow (1894). Mr. Daniel McLennan had also suggested it: and I find that the Rev. John Mathew had stated a form of it in his Eagle-Hawk and Crow (1899), (pp. 1922, 93-112). Mr. Mathew's hypothesis, however, involves a theory of contending and alien races in Australia. This theory does not seem well based, but, however that may be, I recognise that Mr. Mathew's hypothesis of the origin of exogamy (p. 98), and of the origin of the 'phratries' or 'primary classes,' in many respects anticipates my own. He opposes Mr. Howitt's conclusions, and I may be allowed to say that I would prefer Mr. Howitt, owing to his unrivalled knowledge, as an ally. On the other hand, the undesigned coincidence of Dr. Durkheim's, Mr. Daniel McLennan's, Mr. Mathew's, and Herr Cunow's ideas with my own, raises a presumption that mine may not be untenable.
THE CLASS SYSTEM IN AUSTRALIA
Though the existence of what are called exogamous 'phratries' (two to each tribe) was made known, as regards the North American tribes, by Mr. Lewis Morgan (to whose work we return) in the middle of the nineteenth century, almost our earliest hint of its existence in Australia came from the Rev. W. Ridley, a learned missionary, in 1853-55. In Mr. McLennan's Studies in Ancient History[2] will be found an account of Mr. Ridley's facts, as they gradually swelled in volume, altered in character, and were added to, and critically constructed, by the Rev. Mr. Fison, and Mr. A. W. Howitt. These gentlemen were regarded by Mr. McLennan as the allies of Mr. Morgan, in a controversy then being waged with some acerbity. He, therefore, criticised the evidence from Australia rather keenly. It is probable that Mr. Morgan and Mr. McLennan both had some right on their parts—seeing each a different side of the shield—though a few points in the discussion are still undecided. But it seems certain that the continued researches of Messrs. Fison and Howitt, reinforced by the studies of Messrs. Spencer and Gillen in Central Australia, have invalidated some of Mr. McLennan's opinions as to matters of fact.
Much trouble and confusion will be saved if we remember that, as has been said, under the 'classificatory system,' three sets of rules applying to marriage exist. The totem rule exists, rules as to marriage in relation to generations and so-called degrees of kindred (real or 'tribal') exist ('classes'), and, thirdly, there are the rules relative to 'phratries,' the phratries, being, I think, in origin themselves totemic. We shall mainly consider here the so-called 'bisection' of a tribe into two exogamous and intermarrying 'phratries,' while remembering Herr Cunow's opinion that a 'class' is one thing, a 'phratry' quite another.[3]
THE VARIETIES OF MARRIAGE DIVISIONS IN AUSTRALIA
Perhaps the most recent, lucid, and well-informed writer on the various divisions which regulate the marriages of the Australian tribes is Mr. R. H. Mathews.[4] In some regions, the system of two intermarrying phratries exists, without further subdivision (except in regard to totem kins). Sometimes each phratry is divided into two 'sections' (or 'classes'), making four for the tribe. Again, each phratry may have four 'subsections' or 'classes,' making eight for the tribe. Each phratry, like each 'class,' 'has an independent name by which its members are easily recognised.'
Obviously we need, of all tilings, to know the actual meanings of these names, but we do not usually know them. As we shall see, where a tribe has two 'phratries' and no subordinate 'classes,' the names of these 'phratries,' when they can be translated, are usually names of animals. In a few cases, as will later appear, when there are 'classes' under and in the 'phratries' their names seem to indicate distinctions of 'old' and 'young.' But Mr. Mathews nowhere, as far as I have studied him, gives the meanings of the 'class' names, some of which are of recent adoption. Mr. Mathews usually gives only 'Phratry A' and 'Phratry B.' We now cite his tables of the simple 'phratry' system, of the 'phratry' plus two classes system, and of the 'phratry' plus four classes system; making four, or eight, such divisions for the tribe.