SOMATIC MOSAICS IN LEPIDOPTERA. 



Ill 



ton. This was the same tint as these half bhie specimens and may be 

 a completely blue intersex genetically quite different from syujpapha. 



Knnowoti hybrid uinni, Harrison {tuihsir/uaria 3' + quercinaria $ ). 

 A male showed on the left side the usual mixed characters of both 

 parents, but on the right those of the male parent suhmitiaria. Even 

 in the external genitalia the division was apparent. (Harrison, Knto- 

 nii)loi,iist, 1916, xlix., pp. 58 and 59, fig. 8.) 



The specimens of Sipitomis pheijea and Acidalia viit/ularia, included 

 in my former list, were gynandromorphs, and ought to have been in- 

 cluded in the list of gynandromorphs showing a division into typical 

 and varietal or aberrational colour, as well as into male and female 

 characters. 



Morgan discusses the origin of these heterochroic specimens, which 

 he calls somatic mosaics. This name is a better one, because it includes 

 mosaics of different structures as well as those of different colours and 

 patterns. He states that they may arise in three different ways. 

 Firstly by the elimination of an autosomal chromosome, as opposed 

 to a sex chromosome, the elimination of which produces a gynandro- 

 morph. This elimination must take place at the first cell division of 

 the fertilised ovum, in cases in which the halving is perfect, at a later 

 one, where the distribution of the mosaic is unequal. Secondly, by the 

 fertilisation of a binucleate ovum, an event which Doncaster has proved 

 to occur. Thirdly, by a recessive somatic mutation in a sex chromo- 

 some. 



If elimination of an autosomal chromosome were as common as 

 that of a sex chromosome, these mosaics of colour alone would be as 

 common as mosaics of colour combined with sex, mosaic or hetero- 

 chroic gynandromorphs. As a matter of fact they are much rarer, but 

 this may be explained by the way in which the sex chromosome lags 

 behind during cell division. Loss of an autosomal chromosome is 

 thought to be incompatible with life, but there is no evidence that loss 

 of one such chromosome in half the cells of the body would be fatal, 

 even if loss in all the cells would be. In the insects under discussion 

 half the cells would have their full complement of chromosomes. 



In any case this theory will not explain cases in which the somatic 

 factor in question is carried by the sex chromosome. It excludes 

 those in which a mosaic of the same somatic characters is known both 

 alone and associated with gynandromorphism, such as the ones 

 occurring in CoUan, Dnjas, Pajiilin, Psilura, A(jlia and Zycfaena, unless 

 these gynandromorphs are produced in the unusual way from binu- 

 cleate eggs. 



The great majority of gynandromorphs in Drosophila have been 

 proved to be due to the elimination of a sex chromosome. If all somatic 

 mosaics arise from binucleate eggs their number should not exceed that 

 of gynandromorphs wiih the same origin, but the number is consider- 

 ably greater. Some may be due to this cause, but it is unlikely that 

 all are produced from binucleate eggs. 



Morgan considers this to be the origin of one example of a mosaic 

 of somatic without sexual characters, which appeared in a female 

 Droso/ihila, but admits that loss of an autosomal chromosome explains 

 it equally well. 



The third theory is that of somatic mutation. If a somatic muta- 

 tion occurs in only one chromosome of a pair, as it seems to do in germ 



