THE ORIGIN OF GYNANDROMORPHS. 91 



found in his series of hooded rats. So far as we know, the conclusion 

 that Castle's series of characters are mainly due to multiple allelo- 

 morphs is far from being established; on the contrary, we are inclined 

 to think that his evidence indicates that he is dealing mainl}'' with a 

 case of multiple factors. Some of the evidence that Goldschmidt 

 himself furnishes for the gipsy moths is perhaps also capable of inter- 

 pretation in the same way. 



Goldschmidt has shown in some detail that the characters or organs 

 of the intersexes, such as the wings or external genitalia, are mosaics — 

 i. e., relatively large segments or pieces are entirely male or female. 

 In the case of the wings there is no obvious regularity in the mosaic 

 pattern, for the right hind wing may be entirely different from the 

 left hind wing, and the male parts of the right wing do not by any 

 means correspond to the male parts of the left wing, nor does either 

 conform strictly to any underlying structure, such as the veins. In 

 so far, then, as each part is strictly male or female and not a blend of 

 both, the gipsy-moth intersex is like the Drosophila gynandromorph. 

 The results are, however, unlike the Drosophila gynandromorphs in 

 that in the gipsy-moth hybrids the phenomenon must occur very 

 frequently. Baltzer has shown for certain sea-urchin hybrids that 

 when the cross is made one way there is always an irregular (?) elimina- 

 tion of chromosomes, and this result invites at least a comparison with 

 the gipsy hybrids. A solution of the case of intersexes in the gipsy 

 moth could probably be reached by the discovery and study of sex- 

 linked characters. 



Several gynandromorphs of Colias have been described (see Cock- 

 ayne), but of unknown parentage. In the moth Algia tau also several 

 gynandromorphs have been recorded, but the published evidence 

 known to us does not give any clue as to their origin. 



As has been stated, the great majority of gynandromorph Lepidop- 

 tera are not hybrids, but show the secondary sexual characters of the 

 male on one side and the secondary sexual characters of the female 

 on the other. There are, however, a few gj^nandromorphs in this 

 ■ group that show racial or specific differences along with the male and 

 female characters. Amongst these only a few have a known ancestry, 

 and amongst these again it is seldom known whether the characters 

 exhibited are sex-linked or not. Even if they are sex-linked the 

 evidence fails to discriminate between a result that depends only on 

 sex-chromosomal differences and a result that depends on a full chro- 

 mosome group. A search through most of the available literature has 

 brought to light only a few cases that bear on the theories that have 

 been already discussed. Nevertheless, it is probable that a more 

 thorough search through the voluminous literature might furnish more 

 of the critical evidence desired. It is not improbable that entomolo- 

 gists who have made varietal crosses may be able to supply some of the 

 needed data. 



