372 CELL HEREDITY 



tradk'ted l)\ findings in somatic cells themselves. As a matter of fact, 

 experimentation on such cells mimics the approaches undertaken in 

 microbial genetics. 



A priori, we could expect that somatic cells must comprise a hetero- 

 geneous population because of mutation. If the human body consists 

 of some 10 '^ cells, and the mutation rate for a given gene is as low 

 as 10"**, distributed throughout one individual, there would be con- 

 siderablv more than a million cells carrying a chromosome which was 

 mutant at that locus alone. Aside from technical difficulties inherent 

 in studies of somatic cells, there is the major obstacle of diploidv to 

 overcome. Rt^cessi\e mutations are not immediately observable in homo- 

 zygous diploids, but mutations in one of a pair of recessive genes can 

 be detected directly. Many domesticated flowering plants have been 

 selected for their mutability. The bud variegations which modify flower 

 color mav include, at least in part, the reproductive cells. Their 

 behavior in crosses shows that the \ariegation involves a somatic muta- 

 tion. When variants appear in individuals produced asexually from cut- 

 tings or runners, it is difficult to be sure that the change results from 

 mutation at a specific locus because the test of crossing is not available. 

 SimilarK , in animals where the soma is usually irreversibly differentiated, 

 somatic \ariants might arise bv other means. But some mav occur 

 through mutation. 



The inheritance of antigens usuallv docs not inxoKc dominance; each 

 member of a pair of alleles independently causes the production of the 

 one antigen characteristic of it. This is also true of the antigens deter- 

 mining tissue incompatibilitv. If the antigenic constitution of trans- 

 planted tissue is different from that of the host, the transplant, after 

 a sojourn of a variable length of time, v\ill be rejected. The foreign 

 antigen elicits a reaction in the host which exentuallv destrovs cells 

 containing that antigen. This is why successful grafting can usually be 

 achieved only within the same organism or between identical twins. In 

 the adult, tho.se antigens which were naturallv present in the cells of the 

 embryo for some reason do not elicit this reaction. Even the injection of 

 foreign antigens into an embrvo will protect it as an adult against the 

 same antigen. With experimental animals it is possible bv inl)reeding to 

 produce essentially isogenic indi\ ichuds which behave as identical t\\ ins 

 in grafting. It is also possible to have lines which ar(> isogenic except 

 that they differ in the alleles at one of the iiianx loci determining 

 histocompatibilitv. 



The Kleins in Sweden have made lixbrids between such lines of mice 

 differing onlv at the H-2 locus, one of the 14 known to deternnne his- 

 tocompatibilitv. Tumors arising in such mice after treatment with 



