6. The relation between chemical substances circulating in the body--either derivatives of food taken up from without or of chemical compounds formed naturally inside the body--and the production of sexual characters is best shown in the polymorphism found among the social ants, bees, and wasps. Here we have, as a rule, in addition to the two sexes a third one, the workers, which are in reality rudimentary and for that reason sterile females. They differ more or less markedly from both the typical male and female in their external form, and, as a rule cannot copulate owing to their deficient structure. This third sex, the sterile neuters, can be transformed at desire into sexual females in certain species, as P. Marchal has demonstrated. He worked with a form of social wasps in which the workers are sterile and smaller than the real females. In such a society of wasps all the males and workers die in the fall and only the fertilized females survive, each one founding a new nest in the following spring. From the first eggs laid, workers arise, small in stature and sterile; these workers are nourished by their mother. Then these workers take care of the feeding of all those larvæ which arise from the eggs which their mother continues to lay. Throughout the spring only workers arise from the eggs. The males appear in the summer, the real females towards the end of the season when the sexes copulate. Marchal isolated a number of the sterile workers, providing them with food but giving them no larvæ to raise. He found that the workers which thus far had been sterile became fertile, producing, however, only males. This latter fact is easily understood from what has been said regarding the bees, namely, that the female produces only one type of eggs, hence the unfertilized egg can give rise only to males. The astonishing or important point is that the ovaries of the workers begin to develop as soon as they no longer have a chance to nourish the larvæ, provided the food which would have been given to the larvæ is now at their disposal. In other words, the development of their ovaries is the outcome of eating the food which under normal conditions they would have given to the larvæ. The food must, therefore, contain a substance which induces the development of eggs. The natural sterility of the neuters or workers is, therefore, to use P. Marchal’s expression, a case of “food castration,” (“castration nutriciale”).[196] The workers originate from fertilized eggs and are therefore females, but for the full development of the ovaries and the other sexual characters something else besides the XX chromosomes is needed and this is supplied in this case by the quantity or quality of the food. May we not conclude that the same thing may happen generally, except that these substances are formed by the body under the normal conditions of nutrition through the influence of constituents of the second X chromosome?