THE ORIGIN OF GYNANDROMORPHS. 113 



already acquired by the male. Future work must show whether or 

 not such autosomal mosaics are viable. 



(9) Courtship has been watched in a number of flies that were partly 

 male and partly female. Many of them are indifferent; some react as 

 males, some as females. 



(10) In several cases flies that had one white eye and one red eye 

 have been observed to show circus movements. Since the white- 

 eyed fly is less responsive to light than the red-eye fly, the circus move- 

 ments of the gynandromorph with one white and one red eye is what 

 is to be expected. Of course, such cases must be selected so that the 

 legs are not male on one side and female on the other. 



(11) The general evidence from mutation in Drosophila makes it 

 highly probable that when a mutation occurs it takes place in only 

 one chromosome of the pair. Hence any mutation in somatic tissue, if 

 recessive, would be concealed by the presence of the normal allelomorph 

 in the homologous chromosome. If, however, a mutation should appear 

 in the sex chromosome of the male, even though recessive, its effects 

 might be apparent. It is probably significant that the ten cases here 

 described and supposed to be somatic mutations are all males. 



(12) Theoretically, at least, there is the possibility that an indi- 

 vidual starting as a male might produce female parts. If at some 

 embryonic division both daughter X's of an XY cell should pass 

 into the same cell, it would be expected to produce female parts. There 

 is, however, a difficulty with the other cell containing a Y chromosome 

 and no X. It would probably die. 



(13) In addition to the two earlier theories of Boveri and Morgan 

 mentioned above, other theories are critically considered from the point 

 of view of the gynandromorphs of Drosophila. The only other theory 

 besides elimination that we have found necessary to employ in ac- 

 counting for the gynandromorphs of Drosophila, where the genetic evi- 

 dence makes the analysis possible, is the theory of binucleated eggs. 



(14) In the light of the evidence from Drosophila, both the Eugster- 

 bee gynandromorph and von Engelhardt's gynandromorph can be 

 accounted for on the hypothesis of chromosomal elimination, especially 

 since the work of Newell and Quinn shows that the racial characters 

 involved differ in one Mendelian gene (though not necessarily <3ne in 

 the sex chromosome). However, in both cases, if paternal and 

 maternal elimination are equally likely in both combinations, as many 

 gynandromorphs showing the racial character of only one type are 

 expected as those mosaic for racial characters as well as for sex. Such 

 have not been reported. 



(15) (a) In moths several gynandromorphs have been reported that 

 were mosaics for paternal and maternal characters well as as for sex. 

 Some of these, starting as males, can be explained by chromosomal 

 elimination. 



