418 GENETICS OF SOMATIC CELLS 



frequent on the genetic background of strain 201, while the gene appears to be stable 

 on a C3H background. In most cases, the spots occupied only about 0.1 per cent of 

 the coat and the germinal tissue was not involved. In five animals where the pro- 

 portion of full-colored fur was much larger (100 per cent in two), the germinal tissue 

 was also involved, however, and the reverted character of full color was transmitted 

 to a certain proportion of the young. Since even the two completely full-colored 

 animals produced less than 40 per cent full-colored young, it was concluded that 

 somatic rather than germinal mutation was responsible even in these individuals. 

 These events seem to be capable of occurring early or late in development, and, when 

 occurring early, they may involve the gonads. Appropriate crosses have confirmed 

 that the reversions were due to events at the pe locus or within 0.4 per cent crossover 

 units from it. Russell and Major suggested that somatic mutations of this type may be 

 valuable for following cellular lineage in development of the mouse. 



Other interesting cases where somatic mutations have included the gonads in 

 mice have been described by Dunn 295 and Bhat. 86 In both Dunn's and Bhat's case 

 the mosaics (which, incidentally, were mutants for allelomorphs of the albino and the 

 agouti series, respectively) carried the gene s (piebald), in the heterozygous condition. 

 Dunn has pointed out that the frequent association between piebaldness and mosaicism 

 can hardly be accidental, whatever the causal relation may be. The majority of color 

 mosaics in other rodents have occurred in animals with some white spotting. 



Another case where both soma and gonads were involved but which turned out 

 not to be due to simple somatic mutation upon breeding analysis has been described 

 by Carter. 156 A mosaic female mouse developed from a zygote of the constitution 

 W v \ + , W° being a semidominant color and macrocytic anemia mutant. Parts of this 

 animal showed a somatic deficiency for W v , but she bred as though her germinal tissues 

 were partly deficient of the wild-type allele + w . Somatic mutation could hardly 

 have been the mechanism of formation, since two mutational steps would have to be 

 postulated, W v to + in the soma and + to W° in the gonad. Deletion was also im- 

 probable, since two deletions in homologous chromosomes would have to be postulated. 

 On the other hand, nondisjunction during one of the early cleavage divisions, somatic 

 reduction, and somatic crossing over have been regarded as possible explanations, 

 quite compatible with the facts. Carter 156 quotes also some other cases of mosaic 

 animals from the literature which fit somatic reduction or somatic crossing over much 

 better than point mutation. 



The cases where both soma and gonads are involved must be regarded as highly 

 exceptional, and most animal mosaics are purely somatic. In plants, such somatic 

 changes may often be subjected to further study; a "bud mutation" on a tree, for 

 instance, can be propagated asexually and produce a new line of distinct individuals. 

 In fact, most old plants vegetatively propagated, such as pelargoniums and potatoes, 

 have become chimeras owing to somatic mutations at some time during their history. 240 

 In animals, somatic mutations have a different effect: development and life being 

 limited, the mixtures do not survive beyond the life of the individual in which they 



