THE ORIGIN OF GYNANDROMORPHS. 27 



SOMATIC MUTATION. 



That mutation may take place in somatic cells comparable to the 

 mutation process in the germ-tract can not be doubted. The bud- 

 sports long familiar to botanists probably furnish in some instances 

 examples of this sort; but the best authenticated cases are the modern 

 ones that have been analyzed by recognized genetic methods. Few 

 examples are known to zoologists; the monsters, freaks, and duplica- 

 tions that are frequently found are generally due to environmental 

 effects on the embryo. 



If somatic mutation occurs in only one chromosome of a pair, as 

 seems to be the case with germinal mutations, the immediate result 

 will not be seen except when the mutation is dominant. In the case 

 of mutation in the germ-tract, a recessive gene in one chromosome 

 of a pair may likewise not have opportunity at first to express itself, 

 but if it is carried to one of the offspring it will there become multiplied 

 and get into daughters and sons (or in hermaphroditic species into 

 pollen and ovules). Chance union of the gametes that contain the 

 mutated chromosomes will later bring even the recessive genes to 

 expression. It is more probable, therefore, that recessive mutations 

 will appear in the sexually reproducing species more readily than 

 in those with vegetative reproduction, except where the latter are 

 already heterozygous. The same comparison may be made between 

 parthenogenetic species and sexual ones. In the former, a recessive 

 mutation appearing in one chromosome of a pair will have no oppor- 

 tunity to show effects, and the line may be lost by chance alone. 

 Preservation will be favored only if the heterozygous state has an ad- 

 vantage over the original form. Sexual reproduction, therefore, has 

 the advantage that every recessive mutation will have a far better 

 chance of showing itself as a character modification and, if beneficial, 

 of being preserved by natural selection. In fact, if it could be shown 

 that a preponderant number of recessive mutations have furnished 

 the material for evolution, it might possibly appear that we had some 

 hint as to how the process has come to be such an almost universal 

 method of propagation. On the other hand, dominant mutations 

 might nourish, as well by the one as by the other method. 



The best authenticated case of somatic mutation in plants is that 

 described by Emerson, who has brought forward convincing evidence 

 that in corn a gene for certain types of variegation (striped seeds) 

 mutates not infrequently to a gene for uniform-colored grain. The 

 gene for medium variegated "mutates much more frequently than 

 that for very light variegation. " By crossing plants from the mutated 

 grains to pure recessive types Emerson has shown that when the 

 mutation occurs it involves only one member (at a time) of the pair 

 of allelomorphs in question. In these cases the mutation takes place 

 in cell lines (subepidermis) that may ultimately contribute both to 



