502 MUTATION AND PLANT BREEDINC 



superior screening techniques available in the fruit-fly. A simpler 



explanation might be that the Drosophila chromosome is composed 



of relatively longer or more easily broken cistrons. 



Today one cannot easily be critical of Goldschmidt's 2 (then 



controversial) statement of 1955: 



"The important point is that there is no individual locus and 

 no gene which mutates, but a segment of a chromosome which 

 is rather large in chemical terms, which has an orderly, internal 

 structure of a definite sequence, i.e., polarized, and which may 

 even overlap the next one; and that any happening within such 

 a segment which changes this sequence visibly or invisibly 

 appears as a mutant " 



Specificity 



In this Symposium we have heard the word "specificity" used 

 often, in various connections and at several different levels. Dr. Li 

 has told us that a term may be used in any sense we wish, provid- 

 ing that we first define it. I am not so sure that we have a common 

 definition in the present instance; in fact, we have achieved an 

 "unspecific specificity". 



Specificity presumably implies a unique property or class of 

 properties. At the chemical level the substitution of a pyrimidine 

 base normally present in DNA by a base analog is highly specific 

 in a chemical sense. But since the natural pyrimidine base occurs 

 over and over again in the nucleotide sequence, the substitution 

 per se can hardly be considered genetically specific. Genetic spec- 

 ificity would require that substitution would only take place when 

 the natural base were located in a certain position in the sequence. 

 It does not follow, then, that a base analog needs possess any genetic 

 specificity as a mutagen, though it may be "site-specific" within a 

 given locus. 



At quite another level of organization, different spectra of 

 reactions resulting from applications of different mutagens have 

 been presented as evidence of "specificity". One feels confident 

 that there is an underlying truth in this viewpoint, but the term 

 has to be employed so loosely that it is not particularly useful. Few 

 of the comparisons seem clear cut. One has a series of more or less 



2 (;oldschmi(lt, R. li. L955. Theoretical Genetics. Berkeley: University of California 

 Press. Page 161. 



