Wayne Dennis has recently returned to his discussion of alleged cases of children partially reared in isolation or in association with animals (2). His conclusions confirm his earlier analysis (3) and he insists utmost caution be used in accepting evidence on feral children. As he points out, it is difficult to see how such evidence can ever be anything but inconclusive. The more important aspect of discussions on wild children concerns applicability of such data as do exist to general question of socialization of human offspring. There is a tendency to regard these data as substantiation for hypothesis early conditions of nurture are important in future life of child. For purposes of analysis Dennis first accepts evidence and then proceeds to review six cases from Singh and Zingg (5) on which fullest data are available. Quite rightly, he refuses to accept his own analysis of these cases: that socialization which ordinarily takes place during early years is readily lost upon close contact with animals; animal influences are relatively permanent. As I understand it, Dennis is not in disagreement with hypothesis concerning importance of early nurture but wishes to be most exact in outlining its conditions. In present paper. I should like to carry analysis of Dennis' cases a step further and to discuss a principle which seems to overcome dangers of an over-simplified argument concerning importance or lack of importance of early years of rearing. I will also indicate nature of field investigations in India' which led to my utilization of this principle, tentatively called here the principle of reinforcement. It should be noted principle of reinforcement differs in important respects from reinforcement theory as employed in psycho-sociological studies of learning. Reinforcement as used in this paper means more than reward which reinforces response in a simple stimulus-response situation, a point which will be referred to in greater detail below.