122 " Gynandromorphisni " and Kindred Problems 



gynandromorphisin. Such dift'erences of potency must be very un- 

 usual for no other crosses between different local races have produced 

 gynandromorphs in the same way as in the Lymantrias. 



Even in a large majority of allied but distinct species these factors 

 must have the same potency, for primary hybrid gynandromorphs are 

 only produced with certainty in two instances: A. hyhr. hyhridus and 

 A. hyhr. uperosa. For these and for the secondary hybrid gynatidro- 

 morphs Goldschmidt's explanation is possible ; but for ordinai-y halved 

 gynandromorphs and the halved examples which are sometimes met 

 with in A. hyhr. hyhridus and other primary hybrids, I think another 

 explanation must be sought for. Even for the regular occurrence of 

 primary and secondary hybrid gynandromorphs a more gross physical 

 cause is probable. In difl'erent species the chromosomes often differ 

 in number and in size, and it is known that fusion between corre- 

 sponding chromosomes does not always occur in their hybrids. Such 

 chromosomes may take no jaart in mitosis and may be unrepresented 

 in the nuclei of the cells of the resultant individual. This probably 

 causes the gi-eat differences in reciprocal crosses in somatic characters, 

 and may also account for the very different sex proportions. It may 

 also explain the regular occurrence of gynandromorphs in these hybrids. 



This will remain in doubt owing to the infertility of these true 

 hybrids, though, perhaps, some light may be shed by a study of the 

 behaviour of the chromosomes in the mitoses of the hybrids of the 

 genus Amorpha. 



The great objection to Goldschmidt and Poppelbaum's theoiy is that 

 it only explains the finer mosaic or coarser mosaic gynandromorphs. It 

 does not explain the occurrence of halved examples or of lateral genetic 

 hermaphrodites. In all the vast number of Lymantria gynandromorphs 

 bred there was not a single one completely halved in external characters, 

 nor was there one nearly completely halved. Whereas amongst the 

 32 wild Lymantria. gynandromorphs recorded, although there are many 

 resembling Goldschmidt's and possibly due to the same cause, there are 

 no less than 16 completely halved examples, including at least one 

 lateral genetic hermaphrodite. In other species halved or nearly halved 

 examples are far commoner than the coarse mosaic and these again are 

 commoner than the fine mosaic examples. 



Now it is quite possible to produce a mosaic appearance or a blended 

 appearance of somatic characters ; for instance certain strains of black 

 fowls crossed with white produce fowls with a mosaic of black and white, 

 other strains produce a blended blue F^ generation. 



