Crossing European and Japanese gypsy moths, Goldschmidt [1,2,3,4] noticed that the sex types secured were not pure--i.e., that certain crosses produced females which bore a distinctly greater resemblance to the male type than others, and _vice versa_. One of these hybrids of "intersexes," as he calls them, would always possess some female and some male sexual characters. He found that he could separate the males and females, respectively, into seven distinct grades with respect to their modification toward the opposite-sex type, and could produce any one of these grades at will by breeding. For example, the seven grades of females were roughly as follows: (1) Pure females; (2) Females with feathered antennæ like males and producing fewer than the normal number of eggs; (3) Appearance of the brown (male) patches on the white female wings; ripe eggs in abdomen, but only hairs in the egg-sponge laid; instincts still female; (4) Instincts less female; whole sections of wings with male colouration, interspersed with cuneiform female sectors; abdomen smaller, males less attracted; reproduction impossible; (5) Male colouration over almost the entire wing; abdomen almost male, with few ripe eggs; instincts intermediate between male and female; (6) Like males, but with rudimentary ovaries and show female traits in some other organs; (7) Males with a few traces of female origin, notably wing-shape. The males showed the same graded approach to the female type. Their instincts likewise became more and more female as the type was modified in that direction. That is, a moth would be 12% or 35% female, and so on. Goldschmidt watched the crosses which produced seven different grades of maleness in his females. The moth material, like the birds and mammals, suggested a dual basis for sex in each individual. The grades of maleness and femaleness made it seem probable that the factor which determines sex must be stronger in some instances than in others, i.e., that the difference between two of these grades of female is originally quantitative, not qualitative--in amount rather than in kind.