[196] This account of Marchal’s beautiful experiments is taken from Caullery, M., _Les Problèmes de la Sexualité_. Paris, 1913. It is known that the future queens among the bees receive also a special type of food which the workers do not receive. Again the idea of “food castration” of the latter is suggested. In rotifers Whitney[197] has shown that the cycle in the production of males and females can be regulated by the food. In some species a scanty supply of green flagellates produced purely female offspring, while a copious diet of the same green flagellates produced a predominance of male grandchildren, sometimes as high as ninety-five per cent. This was confirmed by Shull and Ladoff.[198] [197] Whitney, D. D., _Science_, 1916, xliii., 176. [198] Shull, A. F., and Ladoff, S., _Science_, 1916, xliii., 177. 7. The effects of the removal of the ovaries or testes upon the development of secondary sexual characters differ for different species. In insects the secondary sexual characters are not altered by an operative removal of the sexual glands as in the caterpillar, _e. g._, _Ocneria dispar_, according to Oudemans. This result has been invariably confirmed by all subsequent workers, especially by Meisenheimer. Crampton grafted the heads of pupæ of butterflies upon the bodies of other specimens of the opposite sex, but the sexual characters of the head remained unaltered. In vertebrates, however, there exists a distinct influence of a secretion from the sexual glands upon the development of certain of the secondary sexual characters, which do not develop until sexual maturity. In a way the observations on arrhenoidy and thelyidy referred to above are indications of this influence.