26 



THE ORIGIN OF GYNANDROMORPHS. 



SOMATIC MOSAICS. 



Somatic mosaics can be accounted for by autosomal elimination 

 in the same way that gynandromorphs are accounted for by X-chromo- 

 somal elimination. Somatic mosaics might also be expected to arise 

 from binucleated eggs and to be as often found as are gynandromorphs 

 with the same origin. As a matter of fact, we have found only one 

 certain case, which is less than expected on the latter view. The case 

 is as follows : -^ 



The grandmother was spineless (third-chromosome recessive) and 

 the grandfather was spread (another third-chromosome recessive). 

 The daughters and sons were wild-type. A pair of these gave a 2 : 1 : 1 : 

 ratio, as expected, because of no crossing over in the male. 



One of the granddaughters (No. 561, Oct. 3, 1914, text-fig. 7) was 

 a mosaic of spineless and not-spineless. The left side of the thorax 

 and abdomen and the left wing and the middle 

 and last left leg were spineless. The rest of the 

 female (including all of the head and left foreleg) 

 had long bristles and hairs of the wild type. 



Simple eUmination of the third chromosome from 

 the spread parent would explain this case were it 

 not that the existence of an individual lacking an 

 autosome is doubtful, because none have as yet 

 appeared through autosomal non-disjunction. On 

 the alternative view of a binucleated egg, one 

 nucleus contained the spineless third chromosome, 

 the other a spread-bearing chromosome ; both nuclei 

 were fertilized by X sperm bearing the spineless 

 X chromosome, and gave the female spineless on the 

 left side and wild-type on the right side. 



The fact that the overwhelming number of hy- 

 brid mosaics are gynandromorphs, involving there- 

 fore the sex chromosome, can not be explained as due to failure to 

 discover autosomal mosaics if they occurred. In most of our cases 

 these would be just as striking as in the cases where the sex chromo- 

 somes are involved. Evidently some peculiarity in the separation of 

 the halves of the sex chromosomes makes the elimination of one of 

 the daughter halves more probable than in the case of other chro- 

 mosomes. Such a supposition is, of course, in harmony with the pecu- 

 liar behavior of the sex chromosome at the reduction division of the 

 male, at least when it lags on the spindle. On the other hand, when 

 it does divide, as in the female, no such peculiarity is recorded, and 

 it is this reduction, rather than the former one, that we need for com- . 

 parison. 



Text-figure 7. 



