112 THE ORIGIN OF GYNANDROMORPHS. 



mothers or to be themselves XXY zygotes. In such cases, after elimi- 

 nation, the male parts are expected to be XY, and hence an individual 

 of this origin with a male abdomen and testes is expected to be fertile. 



(4) A striking fact in regard to these gynandromorphs is that the 

 male and female parts and their sex-linked characters are strictly 

 self -determining, each developing according to its own constitution. 

 No matter how large or how small a region may be, it is not interfered 

 with by the aspirations of its neighbors, nor is it overruled by the 

 action of the gonad. 



(5) Four experiments were made in which suitable material was 

 carefully scrutinized for gynandromorphs. In the 88,000 flies, there 

 were found 40 gynandromorphs, or 1 to 2,200. Since only those that 

 start as females give this kind of gynandromorph, chromosomal 

 elimination may have occurred once in 1,100 individuals. 



(6) (a) If chromosomal elimination took place at the first division 

 of the segmentation nucleus, a half-and-half gynandromorph is 

 expected (right-and-left or anterio-posterior) . Whether dorso-ventral 

 separation is expected for such a division depends on whence comes the 

 material that ultimately reaches the dorsal surface of the fly. 



(6) If the chromosomal elimination took place at the second-division 

 period in one of the nuclei only a quadrant is expected to be male, etc. 



(c) The fact that most of our mosaics include large regions of the 

 body may mean that elimination takes place more often during the 

 first or second division, but it may also mean that when smaller regions 

 are involved the gynandromorph would be more often overlooked. 



(7) (a) Both gonads of the same individual are always alike, i. e., 

 both are testes or both are ovaries, even when the external markings 

 of the abdomen are male on one side, female on the other. This 

 result finds its explanation in the assumption that the germ-plasm 

 of Drosophila, as in some other flies, arises from a single cell. This 

 cell, arising after elimination, must be either a spermatogonium or 

 oogonium. If the cell be the former the sex-linked factor of the 

 germ-plasm must be that of the male-determining X chromosome 

 alone and not show any of the factors contained in the other X of the 

 female parts. Such is the case. 



(h) Conversely, the ovary of a gynandromorph containing both X 

 chromosomes should produce eggs containing the original X chromo- 

 somal combinations as well as their cross-over combinations. This, 

 too, is the case. 



(8) It is a striking fact that we have found so few cases of autosomal 

 elimination. The lack of such mosaics may be due to the failure of 

 the ordinary chromosome to lag in division as the X is assumed to do, 

 or it may be that a fly or part of a fly can not exist if one autosome 

 is absent from its complex. That a part may exist with one X chromo- 

 some lost might be explained as due to that condition having been 



