THE ORIGIN OF GYNANDROMORPHS. Ill 



SUMMARY. 



(1) (a) The main outcome of this work on gj'nandromorphs of 

 Drosophila is an experimental demonstration of the principal cause of 

 the regional differences that gives rise to the coml)inations of male and 

 female in the same individual. The demonstration was made possible 

 by taking advantage of the genetic situation in this material. 



(b) Many of the gynandromorphs were hybrids of known sex-linked 

 characters, i. e., characters whose genes are carried by the sex 

 chromosomes. 



(c) By adding to such crosses additional characters whose genes lie 

 in other than the sex chromosomes it has been possible to prove that 

 the male and female parts of the gynandromorph differ by the sex 

 chromosomes alone, i. e., both male and female parts contain the same 

 autosomal group. 



(d) It was possible, in consequence, to show that these gynandro- 

 morphs are not due to partial fertilization (Boveri), or to polyspermy 

 (Morgan), but to chromosoma,l elimination (Morgan). Chromosomal 

 ehmination means that at an early stage in embryonic development 

 one of the daughter chromosomes of one of the X's fails to pass over 

 to one of the daughter plates, and accordingly gets left out of that 

 nucleus. In consequence, one of the two cells will contain only one 

 X chromosome and produce male parts, while the sister cell with two 

 daughter X chromosomes will produce female parts. The evidence 

 that elimination of this kind takes place rests on cases in which the 

 X chromosome derived from the father contains different sex-linked 

 genes from the X chromosome derived from the mother. 



(e) A census of the available gynandromorphs shows that a paternal 

 X chromosome is eliminated as often as a maternal X chromosome. 



(2) A logical consequence of the proof that the gynandromorphs 

 arise through elimination is that they should all start as females, 

 i. e., as XX individuals. If the elimination always takes place at the 

 first division the expectation would be for the male and female parts 

 to be equal; but if at the second, third, or any later division of the 

 nuclei, we should expect to find, on the whole, a preponderance of 

 female parts over male parts. Such is strikingly the case. 



(3) A second logical consequence of chromosomal elimination is 

 that starting as an XX individual; the male parts will be XO, and not > 

 XY as in the normal male. Now, it has been shown by Bridges ' 

 (191G) that XO males arising from primary non-disjunction are sterile : 

 (although in structure, etc., they are exactly like XY or normal ' 

 males). The great majority of gynandromorph individuals with male 

 abdomen and testes are infertile, while if the corresponding parts are 

 female the individual is fertile. The few gynandromorphs, fertile as 

 males, are known from other genetic evidence to have come from XXY 



