Havelock Ellis[18] and G. Stanley Hall[19] have applied the idea of a "race-type" female with peculiar insistence to the human race. Goodale has finally killed the bird evidence upon which earlier workers so largely founded this doctrine, by showing that the "race type" toward which birds tend unless inhibited by the female ovarian secretion is the male type, not the female. There is a great difference in the way the internal secretions act in birds and in man, as will be pointed out later. It is so important that such a major point as general variability must be supported and corroborated by mammalian evidence to prove anything positively for man. As already noted, the statistical studies of Pearson and Mrs. Hollingworth _et al._ have yielded uniformly negative results. In the utilization of data gathered from non-human species, certain differences in the systems of internal secretion must be taken into account. Birds differ from the human species as to internal secretory action in two vital particulars: (a) In the higher mammals, sex depends upon a "complex" of all the glands interacting, instead of upon the sex glands alone as in birds; (b) The male bird instead of the female is homogametic for sex--i.e., the sperm instead of the eggs is uniform as to the sex chromosome. Insects are (in some cases at least) like birds as to the odd chromosome--the opposite of man. But as to secondary sex-characters they differ from both. These characters do not depend upon any condition of the sex organs, but are determined directly by the chemical factors which determine sex itself.[20] In crustacea, the male is an inhibited female (the exact opposite of birds), as shown by the experiments of Giard and Geoffrey Smith on crabs. A parasite, _Sacculina neglecta_, sometimes drives root-like growths into the spider crab, causing slow castration. The females thus desexed do not assume the male type of body, but castrated males vary so far toward the female type that some lay eggs[3, p.143; 20]. It is the discovery of such distinctions which makes it necessary to re-examine all the older biological evidence on the sex problem, and to discard most of it as insufficiently exact. The work of Steinach[12, pp.225f.] on rats is another well-known example of changing sex characters by surgery. Steinach found that an ovary transplanted into a male body changed its characteristics and instincts into the female type. The growth of the male sex organs he found to be definitely inhibited by the ovaries. He went so far as to transplant the whole uterus and tube into the male body, where it developed normally. One of the most interesting of his results is the observation of how the instincts were changed along with the type of body. The feminized males behaved like normal females toward the other males and toward females. Likewise they were treated as normal females by the males.