GENE CONVERSION 26-22 



50 ME spores, indicating nearly complete transformation to the 

 dominant form. 



Two of the important matings in the complete pedigree (Table 

 26-11) are 608 x 587 and 571 x 587. In both these matings an a x a 

 hybrid produces only aaaa asci, proving that recessive genes do 

 not in themselves have the ability to transform nonfermenters into 

 fermenters, but can only acquire fermentative capacity from a fer- 

 menter. This does not exclude the possibility that nonfermenter 

 clones may carry a mechanism which endows the fermenter gene 

 with transforming capacity. 



PROOF THAT THE MULTIPLE FACTOR HYPOTHESIS 

 CANNOT EXPLAIN THE IRREGULAR RATIOS 



Table 26-12 shows data obtained in the course of tetrad analy- 

 sis of heterozygous hybrids which have yielded irregular ratios. 

 The asci were obtained from hybrids after mating two phenotypi- 

 cally different clones which yielded some 2:2 segregations and in 

 addition other asci containing 4:0, 3:1, or other types of tetrads. 

 Only asci in which all four spores are viable can be used in such 

 an analysis. Table 26-13 shows the frequencies of phenotypes ex- 

 pected on tetrad analysis of hybrids when from one to four pairs of 

 genes are involved and a single epistatic gene is sufficient to deter- 

 mine the phenotype. The data in Table 26-11 differ significantly 

 from all these expected types; in fact, the Pppp and pppp tetrads do 

 not fit into any scheme of conventional Mendelism and, therefore, 

 make it impossible to apply any statistical tests. 



These data prove that the multiple factor hypothesis does not 

 provide an adequate explanation for the phenomenon. The large ex- 

 cess of PPpp asci strongly support the view that only a single pair 

 of alleles are involved in determining the phenotype. Many matings 

 of homozygous P x P and p x p hybrids support the view that these 

 irregularities are confined almost exclusively to the heterozygotes. 

 In the P X P hybrids there may be sufficient material present so 

 that losses do not occur frequently, and in the p x p hybrids, the 

 locus may be so degraded that it cannot be repaired in the absence 

 of a P allele. The Pppp asci suggest that on occasion the p locus 

 may detach enough material from the P locus to degrade it to the 

 p phenotype, but that the amount removed either does not reach the 

 p locus or is insufficient to repair it. 



THE SICa^FICANCE OF CONVERSION IN THE LONG-TERM 

 ADAPTATION TO GALACTOSE BY SACCHAROMYCES 



It has long been known that Saccharomyces cerevisiae ferments 

 galactose by an adaptive enzyme produced in the cell during the 



