348 CELL HEREDITY 



regardless of the phase of the receptors. When the unexpressed state 

 of f/2 was transduced, the phase of the progeny was determined by the 

 state of the receptor H2 site. 



The inference seems well taken that the //j locus is able to repress the H j 

 locus. One might say: when the H2 locus turns itself on, it turns //, off. 

 Thus, phase variation is an example of the controlling effect of one locus 

 upon the expression of another. There are evident parallels between this 

 system and some described by McClintock, but further data are required 

 for a more detailed analysis. 



In Paramecium, there are a wide diversity of antigenic types which 

 have been investigated by Sonneborn and his students, and by Beale. 

 Paramecia carry genetic information distributed at many different loci 

 for the synthesis of a variety of surface antigens, only one of which 

 is ever expressed at a time (except in heterozygotes to be discussed 

 below). Presumably, a mechanism must exist not only for selection of 

 the expressed antigen but also for repression of the formation of all the 

 others. A great deal of discussion, speculation, and model building has 

 been evoked by the well-demonstrated interplay of chromosomal and 

 nonchromosomal influences in the determination of antigenic expression 

 in Paramecium. 



The genetic analysis developed by Beale of the antigenic changes in- 

 duced by temperature provide an illustration of what is known at present 

 about the genetic control of antigen formation. He found that the 

 related stocks 60 and 90 behaved similarly in response to temperature 

 changes. At temperatures of 18° C or lower, they developed S antigens. 

 When transferred from low temperature to 25° C after a number of divi- 

 sions, the animals changed their antigenic type from S to G and, simi- 

 larly, at temperatures above 29° C, they switched to a different antigenic 

 type noted as D. Although the antigens of the two stocks were very 

 similar at each temperature, they could be distinguished by careful 

 testing. A convenient feature of the system was the long delay (many 

 divisions) after exposure to a particular temperature before the antigenic 

 change was expressed. Thus it was possible to cross animals which had 

 been grown at different temperatures. 



It was established, first, that the antigens S, D, and G were each deter- 

 mined by a different unlinked gene, and that the small antigenic differ- 

 ences between stocks segregated as allelic differences. For example, in a 

 cross of stocks 60 X 90, kept at 25° C throughout the experiment, the 

 F J hybrids expressed both alleles by containing a mixture of the 60G and 

 90G antigens. At the next autogamy, the two alleles segregated, giving 

 rise to pure 60G and pure 90G animals in a 1 : 1 ratio. 



