414 GENETICS OF SOMATIC CELLS 



since heterozygous and homozygous B cells have not yet been proved to be distinguish- 

 able by their reaction with anti-// reagents. For this reason, the origin of the excep- 

 tional cells by mutation, somatic crossing over, or phenotypic change must be left 

 open at the present time. In fact, it cannot be excluded that the transitory absence 

 of an agglutinogen might be a normal feature of the erythrocytic life cycle. Since 

 erythrocytes are differentiated end products, and cannot be subjected to progeny 

 tests, it cannot be decided whether the exceptional behavior of the variant cells would 

 breed true or not. This difficulty can be overcome by using cells capable of division. 

 An analogous study has, in fact, been carried out with populations of neoplastic cells, 

 induced in F x hybrids of coisogenic resistant mouse strains heterozygous for the iso- 

 antigens determined by the H-2 locus. The results of this study are quite compatible 

 with the possible occurrence of somatic crossing over, although they do not conclusively 

 prove it. They will be described in more detail in the chapter on neoplastic cells. 



Other types of genetic variation. — There is an extensive literature dealing with pheno- 

 typic variation in the somatic tissues of the same plant or animal. Many of these are 

 certainly nongenetic; in plants, such may arise, for example, by viral infections or 

 by cytoplasmic changes affecting the chloroplasts. The various genetic mechanisms 

 that may play a role can be classified in different ways. Swanson 1305 lists them as 

 "chimeras, in which tissues of different genetic or chromosomal constitution lie adjacent 

 to, or overlap, each other; endomitosis and somatic reduction which alters the chromo- 

 somal complement of the cell; abnormal fertilizations to givegynandromorphsor mosaics; 

 somatic crossing over which reveals hidden heterozygosity; chromosomal elimination 

 or fragmentation; and gene mutations." It is obviously impossible to review all these 

 mechanisms within the limits of this article; discussion will be restricted to a few types 

 that have been well documented and appear to be of general significance for the analysis 

 of somatic variation already at the present stage. Chromosomal mechanisms as such 

 will not be considered unless relevant to the cases discussed, since they will be dealt 

 with in another chapter. 



Chimeras have been well known in plants for several decades, but the knowledge 

 of their existence in animals and, particularly, their experimental production are of 

 comparatively recent date. Most available evidence concerns the hematopoietic 

 system, and more particularly, erythrocytic antigens. The various possible mechanisms 

 that can give rise to erythrocytic mosaicism have been reviewed by Cotterman 228 who 

 classified them as natural chimeras (usually made possible through chorionic vascular 

 anastomosis between twins during embryonic life), artificial chimeras (produced by 

 parabiosis or by transplantation) and mutational, nonchimerical mosaics that may 

 arise by gene mutation, somatic segregation, or crossing over, or by chromosomal 

 aberration or variegation mechanisms of several kinds. The first mosaics were dis- 

 covered by Owen 984 who found that cattle twins were almost invariably identical in 

 blood types, in spite of the fact that identical twins in cattle are relatively rare. A 

 closer examination revealed that the peripheral blood was often a mixture of two 

 populations of cells, but never more than two. On genetic analysis, it was found that 



