THE ORIGIN OF GYNANDROMORPHS. 21 



under discussion. The critical evidence that shows that they were 

 not due to separation of whole maternal and paternal nuclei was first 

 obtained and published by Morgan (in 1914). Prior to that time 

 Bridges (1913) had published an account of two hybrid gynandro- 

 morphs, and had suggested that they were due to somatic non-dis- 

 junction. By this term it was meant at the time that at an early 

 embryonic division of a female the two daughter halves of one of the 

 X chromosomes did not disjoin from each other to pass, as normally, 

 into sister cells, but were included in the same cell, the other cell 

 not receiving its half. The non-disjoining X was assumed to divide 

 normally and the result was an X cell developing into male parts and 

 an XXX cell developing into female parts. This hypothesis served 

 to explain all the facts known at that time. Soon, however, it was 

 established (Bridges, 1916) that XXX individuals are unable to sur- 

 vive, and this brought into question the conclusion that the female 

 parts of gynandromorphs were XXX. This difficulty was later 

 avoided by the assumption of ''elimination" (earlier called "mitotic 

 dislocation," Morgan, 1914). As already stated, this meant that one 

 of the daughter X's was caught by the mid-plate and prevented from 

 taking its place in either nucleus. 



There is another class of gynandromorphs (including here four cases) 

 in which another procedure may account for the results. Primary 

 equational non-disjunction occurred, as evidenced by the presence 

 in each of the four gynandromorphs of two X chromosomes from the 

 mother, one of these being a non-cross-over and the other a cross-over 

 X, as is usual for XX eggs produced in this fashion. This XX egg 

 was then fertilized by an X sperm, giving an XXX individual. This 

 XXX zygote is prevented from dying and at the same time converted 

 into a gynandromorph by the occurrence of somatic reduction at the 

 first or a very early embryonic division. In each of the four cases 

 the male parts of the gynandromorph were derived from one of the 

 two maternal X's, which suggests that the essential feature of this 

 somatic reduction is the active separation of the two X's that came 

 from the mother and the passive inclusion of the X from the father 

 with one or the other of them. There have been other cases which 

 may support this view, cases in which XX eggs equationally produced 

 have been fertilized by Y sperm, and then the two X's have likewise 

 reduced, with the result that each cell gets one X, and the entire 

 individual is converted into a male which is a mosaic of different 

 parts clearly marked by the character corresponding to the two dif- 

 ferent X's. The difficulty with this view is that it assumes that 

 reduction can take place between two X's at a cell division without 

 the X's themselves splitting, although all of the other chromosomes 

 do so at this time — a situation for which no support is given by 

 cytology. It is to be noted in this connection tliat all cases that appear 



