SOMATIC MUTATIONS OF BRUCHUS. 1 7 



the fact that the normal female in Drosophila has two (XX) sex- 

 chromosomes in her body, but if one is eliminated on one side of 

 her body, male characters will then develop on that side. There is 

 in Drosophila no essential difference between gynandromorphs 

 (sex-mosaics) and somatic mosaics, except that the former con- 

 cern the elimination of one sex-chromosome, while the latter in- 

 volve the elimination of an autosome. The mechanism of elimina- 

 tion, too, is the same. 



It is true of Bmchus that all visible mosaics were females, but 

 the author ('21) has shown that these color factors in this insect 

 are not sex-linked, but are sex-limited, autosomal traits. It is im- 

 probable that any change in the X-chromosome could produce these 

 mosaics. Morgan and Bridges ('19) have shown that gynandro- 

 morphs are neither caused by partial fertilization, as first suggested 

 by Boveri, nor due to polyspermy, as first interpreted by Morgan, 

 but are caused by chromosome elimination. Chromosome elimina- 

 tion, as described by the above writers, means that during some 

 early stage of development of the embryo one of the daughter sex- 

 chromosomes fails to pass over to the daughter plate, and thereby 

 becomes eliminated from the nucleus. Relative to the mosaics of 

 B ruchus, it is impossible to account for their appearance through 

 the elimination of the X-chromosome, because these characters in- 

 volve the autosome in which the allelomorph series for the R (red) 

 gene is located, and not the X-chromosome. 



Morgan and Bridges ('19) account for mosaics of Drosophila 

 by chromosomal elimination ; in one case they tried to account for 

 the origin of a mosaic, as produced from binucleated eggs, because 

 each paternal nucleus was known to have had a different ancestry. 

 It is unnecessary to make this latter assumption for these mosaics 

 of B ruckus, because both nuclei of a binucleated egg (if such could 

 occur) would of necessity be alike, and when fertilized by a normal 

 sperm, all of the offspring would be homozygous, unless it is as- 

 sumed that a mutation occurred in one nucleus before fertilization, 

 and in that event the effect would not be localized in a single wing, 

 but would involve half the entire body. 



In Drosophila nearly all mosaics can be interpreted as caused by 

 chromosome elimination, because the homologous chromosome 

 from one parent carries factors that are different from its mate. 



