It is not a little remarkable that the two extremes of vitality should occur in the two classes in which in-breeding prevails. The larger class of the concubitants (in which class also is found the highest fecundity) shows the highest vitality of the four groups. The smaller class, the relations other than concubitants, second in point of fecundity, discloses the lowest vitality, and yet the proportion of these marriages which would be regarded as incestuous by our system is small. It is not to be forgotten, however, that in marriages which are regarded by the people as socially right and proper, more care may be bestowed upon the offspring both by the relations of the parents who nurse the mother and child and by the parents themselves. By the same reasoning it is probable that the offspring of marriages regarded as incestuous are neglected by the relations of the parents, and, as a consequence, that less pride is taken in them by the parents themselves. It has not been found that concubitants marry either earlier or later in life than the members of the other classes, and it is to be remembered that concubitants are very often natives of different villages, which may tend to prevent the relations attending upon the mother in her confinement. One of our native witnesses assured us, moreover, that the union of concubitants was seldom a happy one. Quarrels between husband and wife would certainly outweigh any advantages in favour of child-bearing which the social propriety or fitness might be held to create. But even supposing that the influences at work to make concubitancy so satisfactory a procreative element in the population are of a moral nature, the difference is so marked that there is a balance over to be accounted for by some other explanation. That they rear a larger proportion of their children may be partly or wholly due to the fact that their relationship to each other gives them a higher sense of responsibility, but that they bear more children capable of being reared argues a superior physical fitness for procreation. I am aware that the figures are far too small to allow of any generalization from them, but at the same time it is to be remembered that the inhabitants of these twelve villages represent a fair sample of the population, and also that we found the relative positions of the married classes to be generally the same in each village taken individually.