As regards the father-in-law, however, the custom, as far as capture is concerned, would not occur with female descent, for the reasons we have already given of clan kinship in such a case. It might, however, be found as a factor to a certain extent with male kinship, for here it is the father-in-law who is of the same clan as his daughter, and thus interested in her negotiable value. Thus it is possible to imagine enmity between him and her possible captor, who is also of a different clan from himself. As regards avoidance of mother-in-law it has again, perhaps, been accentuated by forcible capture. The effect, however, in relation to descent would be exactly the reverse of that with the father. With early male descent in the primitive tribe as composed of only two clan groups [phratries], it is she who would be of the same stock as her daughter’s husband, and the habit would not arise, the captor is, in fact, a member of the tribe from which she herself has been stolen; although later, when more than two clans were conjoined,[10] it might happen that her son-in-law belonged to another, and here there might arise feelings of animosity. With uterine descent the case is certainly altered. As a mother, and as member of a clan different from that of the male suitor, the figure of the son-in-law might be dreaded as a possible captor of her daughter and other young female members. But here again a difficulty arises, for when the capture becomes an accomplished fact, the mother-in-law and son-in-law would probably not meet again, at least in primitive times: he belonging to the group having patriarchal descent, with capture as the rule; she, to the matriarchal, where the female is normally immobile, between which two forms of group no friendly intercourse could occur. The fact of avoidance in any form presumes contiguity or the habitual presence of the individuals concerned, and this in such a case could not arise. So as in Tylor’s figures we find that in W to H, as the latter is completely cut off from his family, there is not one single case of avoidance between the wife and the husband’s relatives. It is evident that the same rarity of contiguity must have arisen also with the father in male descent; there is here certainly cause of disagreement in the rape, but if the parties see each other no more there would be no necessity of evolution of avoidance to mark the fact. Indeed, I cannot help thinking that the importance of association with hostile capture has been much exaggerated as a factor in the evolution of Avoidance. The question of ‘residence’ and ‘descent’ has not been held sufficiently in account by those who insist on the capture as the sole cause of avoidance. Despite the eminence of the authors favouring this view, I would venture to submit that the balance of proof would much favour sexual jealousy, which we have heretofore found the sole motive power in all changes.