One cause is that some factor in self-interest or otherwise allowed such relationship to remain unrecognised, although existent, and another is that, as cousinship did not exist at all, there could be no recognition, or, as Mr. Morgan puts it, ‘his sister’s son is also his son because his sister is his wife.’ To determine which is correct certainly seems difficult, and the whole thing has evidently been considered a most stubborn fact for the opponents of promiscuity. That Mr. Morgan should have seized it in support of his theory, and that the theory should be so largely accepted, is not astonishing. Happily the great value of his ensuing argument as regards tribal development is in no way impaired if it can be shown, as we hope to do, that there is no necessity for an hypothesis of promiscuity to explain the terms in the Malayan table, which apparent need seems primarily to have led Mr. Morgan to evolve the idea of his primitive group. In fact, it becomes evident that, if we can furnish a clue as to how a sister’s son came also to be a brother’s son, without having recourse to the theory of an incestuous union of brothers and sisters, we at least discount the need of Mr. Morgan’s ‘consanguine family,’ in which such incest is supposed to be a most characteristic and essential feature. We hope to prove that the terms which misled him are more apparent than real as proofs of any real affinity in blood, and that the original conception in causal connection was something quite apart. Sir John Lubbock (Lord Avebury) has observed that the lower the _milieu_ of a social status the less we see of the individual and the more of the group. In the case before us the individual as such does not exist at all, and there is only question of the group in its relation to its component classes. To confound one with the other led to Mr. Morgan’s error. There was much, in fact, in Mr. McLennan’s shrewd remark in criticism of Mr. Morgan’s theory that he did not seek the origin of the system of nomenclature in the origin of the classification of the connected persons, and that he courted failure in attempting to solve the problem by explaining the relationships comprised in the system in detail.’[8] But it seems to me that Mr. McLennan fell into the same error when he contented himself with the misleading analogies which a comparison with the Nair family system presented. These, however striking, are, as we shall find, simply the result of the fact that class or communal marriage was the common trait of the polyandrous and the Cyclopean family, nor can I see that Mr. McLennan followed his own excellent advice as regards the possible identity in origin of nomenclature and classification; if he had so done, his acute mind could not have failed in a resolution of the whole problem, whereas his final resume of the argument is in terms which I profess to be quite unable to grasp.