Not only did Rivers record the islands the various people came from, but their totems as well. By this laborious work a great deal of valuable information will result that could not be obtained in any other way or with anything like the same accuracy. The clan marriages of the population of Mabuiag for several generations will not fail to reveal the rules that regulated marriage and descent.

There was an interesting psychological difference between the Mabuiag folk and those of Murray Island. As has just been pointed out, great secrecy had to be maintained in the latter island when pursuing genealogical inquiries; but quite the opposite condition prevailed in Mabuiag. Here the information was obtained in public, and a doubtful point in genealogy was frankly discussed by men and women. This enabled Rivers to make more rapid progress than in Murray Island, and at the same time he was equally sure of his facts: what in Murray Island required private confabulations with various men at different times could here be settled practically offhand.

Much varied sociological information can be obtained by recording genealogies in this way. For example, one can get definite facts on the number of children in a family, the proportion of the sexes, the number that die before they themselves have children, the number of adopted children, the idea on which relationship is based, the relationship nomenclature, the relation of totems to individuals or communities, the personal or group restrictions as to marriage, the relative fertility of related or unrelated stock, the effect of crossing between different races, and so forth.

Without entering into further detail, I would like to emphasise the fact that by this system Rivers has supplied anthropologists with a new method of research, by means of which important data can be collected with absolute accuracy on subjects concerning which it has hitherto been very difficult to obtain reliable information.

Some white men resident on Mabuiag had crews of mainland (Queensland) blacks, and we took this opportunity to measure, psychologise, and photograph some dozen of these men. This was a fortunate chance for us, as we wanted to make a few comparative observations on the North Queensland aborigines.

There were also a few South Sea men living on Mabuiag, who had married native women, and we studied several of them, and their half-caste children as well. My old friend Billy Tanna was still on Mabuiag with his numerous progeny, and three or four other Tanna men besides, whom we also measured. Tanna is one of the New Hebrides group.

One Lifu man, Sŭni or Charley, had the longest head I have ever measured. It was 215 mm. (8½ inches) in length. It was also narrow and high; the length, breadth (or cephalic) index was 66·9. Rivers found in Sŭni the first example of true colour blindness he had yet come across. It was amusing to see Sŭni’s total inability to discriminate between pink and blue and red and green, and his other attempted matches were very quaint. There were two other Lifu men on the island, and great was Rivers’ delight to find that one of them was also red-green blind. One felt tempted to frame all sorts of wild theories about a colour-blind race. During his return home, both at Thursday Island and at Rockhampton, in Queensland, Rivers investigated four other Lifu men, but only one of these was colour-blind. Still, it is an interesting fact that not a single other case of colour-blindness was found among one hundred and fifty natives of Torres Straits and Kiwai, or among some eighty members of other races, including Australians, Polynesians, Melanesians, Tamils, and half-castes, and yet three out of seven Lifu men were colour-blind. Lifu is one of the Loyalty group in the South Pacific. The inhabitants are Melanesians, with, in some instances, a little Polynesian admixture.

Ray pursued his philological studies in Mabuiag, and found that there was a very marked difference, both structurally and in vocabulary, between it and the Murray Island tongue. From Saibai to Muralug and from Badu to Tut one language is spoken, but there are at least four closely allied dialects corresponding to as many groups of islands. The grammar of this language is decidedly of the Australian type, though there is no marked connection in structure or vocabulary with languages of the neighbouring mainland of Australia.

Later, Ray had an opportunity of verifying this conclusion by a partial investigation of the language of the Yaraikanna tribe of Cape York, which is, however, totally distinct from that of the islanders of either the eastern or western tribes.

A marked peculiarity of the Mabuiag language is the extremely indefinite signification of the verbs, which require to be made definite by prefixes indicating the part of the body concerned, the direction of the action, or the place concerned. For example, palan apparently indicates the putting forth of something; and thus we have poi-palan, to shake off dust; gagai-palan, to fire gun or arrow; minar-palan, to make marks, to write; ibelai-palan, to cover as with a blanket; balbalagi-palan, to make not crooked, to straighten; berai-palan, to make like a rib, to slacken a rope; dan-pali (an intransitive), to open the eyes, to awake; aka-pali, to show fear. Another peculiarity is the partiality of the language for noun constructions; indeed, as all the verb suffixes are the same as those of nouns, it may be doubted whether the verb exists as it is understood in European languages. “I have seen you,” is in Mabuiag ngau ninu imaizinga, literally, “mine your seeing”; the imperative plural “Fear not” is nitamun akagi, “Your not (being) afraid.”