STEPHENS: RESUME OF THE SYMPOSIUM 503 



independent variables: (a) between organisms; (b) between types 

 of event, "break" or "mutant"; (c) between heterochromatin and 

 euchromatin; (d) between positions in chromosomes; and (e) between 

 alleles at the same locus. When most of these variables are con- 

 founded in a system, the number of possible permutations in effects 

 is quite large, and consequently the chances of obtaining similar 

 effects quite low. The unsophisticated observer may be pardoned 

 for wondering how many of the observed differences are due to 

 "specificity" and how many to the laws of statistical probability. 

 Undoubtedly, there are differences in reaction pattern, but if these 

 shift every time a new variable is added, the concept of specificity 

 does not appear to be helpful. 



It has also been suggested that one of the promising lines of 

 research with mutagens is the development of specific mutagens by 

 which mutation can be directed. As a tool in refined genetic and 

 biochemical analysis this is obviously true. But I think it would 

 be difficult for a practising plant breeder to decide what kind of 

 specific mutagen should be given priority and to what end he would 

 direct it. The level of specificity required by the geneticist, as I see 

 it, would be of limited value to the practical breeder. A mutagen 

 which would selectively substitute a particular base pair at a par- 

 ticular site and result in a changed amino acid sequence in a known 

 protein would be invaluable to the geneticist, as Smith has already 

 pointed out. But between a particular protein and the phenotype 

 recognized by the plant breeder lies a long and completely unknown 

 pathway through enzyme syntheses, differentiation, and development. 

 The characteristics with which the plant breeder must work- 

 earliness, disease resistance, stiffness of straw, yield, and the like — 

 are not easily expressed in chemical terms at the end of synthetic 

 processes, much less at the beginning. Again, to use a directed muta- 

 tion, it is first necessary to know what direction, and presumably, 

 the nature of the change which would produce it. It is perhaps 

 unlikely that a mutagen which would produce exclusively mutants 

 resistant to a particular disease would be useful against different 

 strains of the disease in the same variety, or against similar strains 

 in different varieties. The more specific the mutant, the more lim- 

 ited its use is likely to be. 



On the contrary, I would suggest that the reaction patterns 



