SEX DETERMINATION 221 



the autosomes, carrying a tendency towards maleness, and the X 

 chromosomes carrying a tendency towards femaleness. We have the 

 ratios X/A = i for females, X/A = 2/3 for intersexes, and X/A =1/2 

 for males. Other ratios can be obtained in (i) diploids with an extra X 

 chromosome and (2) triploids with only i X. The first gives a ratio of 

 X/A = 3/2; the flies only survive in a few cases and are sterile and 

 malformed females known as superfemales. The second type gives a 

 ratio of X/A = 1/3 and again give sterile and more or less inviable 

 flies, this time of male type known as supermales. If the balance theory 

 of sex, which we have just been discussing, is really the whole of the 

 story, these animals should be, as their names suggest, more extreme 

 sexual types than normal males and females. But it is not at all clear 

 that this is the case. The intersexes of Drosophila, Hke those of Lyman- 

 tria, are complex mixtures of male and female parts, which seems to 

 show that during development there are only two alternative ways in 

 which sexual diff'erentiation can occur. There is no evidence for an 

 intermediate type of development, and thus it is perhaps unUkely that 

 there are other more extreme types. It is possible that the supermales 

 and superfemales have merely followed the normal male and female 

 types of differentiation, modified by their extra or deficient X chromo- 

 somes exactly as they might have been modified by any other duplica- 

 tion or deficiency which was not cormected with sex. 



The intersexes themselves are very variable in grade. They range 

 from nearly pure males to nearly pure females. It can be shown^ that 

 their anatomical peculiarities can be explained in the way suggested by 

 Goldschmidt for Lymantria;^ they start development as males and, 

 after a switch-over, continue it as females, and their grade of inter- 

 sexuality depends on the relative lengths of the two periods. Kerkis,^ in 

 fact, has given evidence that the same phenomenon occurs in the 

 normal sexes. The gonads start growing at the same rate in both sexes 

 and in the male the growth rate changes slowly and continuously 

 throughout larval Hfe, while in the female there is a sudden change of 

 rate at an early period which probably corresponds to the switch-over. 



Many genes are known which affect the anatomical development of 

 the sex organs in Drosophila. Thus there is a gene which prevents the 

 formation of the penis and claspers in the male, another which removes 

 the sex-combs, etc. Similarly the anatomical development of the inter- 



^ Dobzhansky 1930. 



^ But cf. Bridges 1938a, where he adopts a point of view more hke that of 

 Bahzer 1937. 3 Kerkis 1934. 



