326 



SCIENCE 



[N. S. Vol. LI. No. 1317 



Sex combs on fore 

 tarsi 



Number of dorsal ab- 

 dominal tergites. , 



0^^posito^ 



Spermathecae 



Penis 



First genital tergite. 



Anal plates 



Claspers 



Tip of abdomen. . . 

 Gonads 



Present 



5 



Absent 



None 



Present 



Present 



Lateral 



Present 



Black 



Testes 



Absent 



7 



Present 



2 



Absent 

 Absent 

 Dorsal and 



ventral 

 Absent 

 Banded 

 Ovaries 



Absent 



7 

 Present 



2 

 Absent 

 Present 

 Lateral 



Present 

 Black 



Very minute 

 if present 



Crosses of normals from the interses stock 

 have made it possible to study tlie character. 

 The intersexes are modified females — i. e., 

 they have two X-chromosomes. This is shown 

 by the fact that in cultures in which half of 

 the males show sex-linked recessive characters 

 but all the females are wild-type, the inter- 

 sexes n^ver show these sex-linked characters. 

 This relation has been found to hold true for 

 three sex-linked characters that are not closely 

 linked to each other; and the intersex gene 

 itself has been found not to be sex-linked (see 

 below). Therefore the relation just noted 

 can not be due to linkage between the intersex 

 gene and the sex-linked genes in question. 



Numerous crosses of the intersex stock to 

 unrelated stocks have never given intersexes 

 in Fj, but have frequently produced them in 

 Fg- The intersex character is therefore re- 

 cessive. 



Pair matings that have produced intersexes 

 have given a total of 5105 : 165 intersex : 754c?. 

 There is an excess of males, but this is evi- 

 dently a 3 :1 ratio of females to intersexes, 

 indicating not only that the gene is recessive 

 but also that it is not sex-linked. The final 

 proof of the latter ]X)int has been obtained 

 through the discovery that the intersex gene 

 is linked to the autosomal recessive gene for 

 " plum " eye-color. Three Fj pairs from a 

 mating between the intersex stock and the 

 plum stock have given in F, : 



^^ij5g absence of the intersex plum class 

 shows that the two genes are linked; and plum 

 is known to be an autosomal recessive. 



It has been shown by Morgan and Bridges^ 

 that individuals of D. melanogaster^ that are 

 partly male and partly female are produced, 

 though only rarely, by most stocks. These 

 " gynandromorphs " have been shown, by gen- 

 etic evidence, to have two X-chromosomes in 

 their female parts and only one X in their 

 male parts. They are sex mosaics, and each 

 part develops as it would in a whole animal 

 of the same genetic constitution. There is 

 strong evidence that the intersexes described 

 here are not of this nature. The male and 

 female parts in them probably both possess 

 two X-chromosomes. This has been shown as 

 follows. A total of 104 intersexes have been 

 produced by females heterozygous for the sex- 

 linked gene for " yellow " hairs and bristles. 

 Half of these intersexes — about 50 — must then 

 themselves have been heterozygous for yellow. 

 If the intersexes are really gynandromorphs, 

 the male parts at the posterior end of the 

 abdomen should have contained a single X- 

 chromosome, and in about half of the speci- 

 mens that were heterozygous for yellow (i. e., 

 in about 25 individuals) this should have 

 been the yellow-bearing X. As Morgan and 

 Bridges have shown, these parts should then 

 have borne yellow hairs and bristles. The 

 104 intersexes were all carefully examined for 

 this point, and none of them had yellow male 

 parts. 



We may conclude that the intersexes are 

 females, modified by a recessive autosomal 

 mutant gene that causes them to show male 

 parts, though these parts themselves still have 

 two X-chromosomes. The normal sex-deter- 

 mining mechanism is not affected at all, but 

 the end result is modified by a gene that is 

 not even in the sex 6hromosomes. It has 



2 Carnegie Inst. "Washington (1919), publ. 278, 

 pp. 3-122. 



3 I have unpublished data on exactly similar cases 

 in D. simulans itself. 



