STABLER 



tention that the properties commonly 

 ascribed to "the classical, corpuscular 

 gene" go far beyond the evidence is, I 

 think, fully justified. 



Sturtevant is correct if, by the 

 gene, we mean the gene of the opera- 

 tional definition, since this implies no 

 unproved properties. If it were true 

 that there are no discrete units in the 

 gene-string, Sturtevant points out, the 

 most direct way of establishing the 

 fact experimentally would still be by 

 studying the properties and interrela- 

 tionships of these distinguishable re- 

 ^•ions. These are the genes of the op- 

 erational definition. 



What is the operational definition 

 of gene mutation? We have recog- 

 nized that our studies of gene mutation 

 have significance for the major prob- 

 lem only to the extent that they iden- 

 tify and analyze the mutations that 

 represent the evolution of new hered- 

 itary units. But it is obvious that no 

 operational definition of gene muta- 

 tion in this sense can now be formu- 

 lated—for these hereditary units are 

 not the genes of the operational defini- 

 tion; they are the hypothetical genes 

 postulated in our interpretation of the 

 experimental evidence. To say that no 

 operational definition is now possible 

 is only to repeat in difl^erent words the 

 foregoing statement that "we have no 

 positive criterion to identify mutations 

 caused by a change within the gene, 

 and that the alterations interpreted as 

 gene mutations in experiments are 

 merely the unclassified residue that 

 cannot be proved to be due to other 

 causes. The major objective in further 

 investigations must be to develop such 

 criterions. 



STUDY OF THE MUTATION 

 OF SPECIFIC GENES 



The main purpose of this paper {11) 

 is to emphasize the unpleasant fact that 

 significant progress in our understand- 

 ing of gene mutation requires the 



251 



investigation of the mutation of spe- 

 cific genes. The fact is unpleasant be- 

 cause the various technical difficulties 

 that arise from the very low frequency 

 characteristic of mutation are at their 

 worst when the study must be made 

 on single genes, particularly on the 

 spontaneous mutation of single genes. 

 The unpleasant statement is a fact 

 because, as we have seen, it is hopeless 

 to identify and exclude the spurious 

 or extragenic mutations in experiments 

 on mutation rates at miscellaneous un- 

 specified loci. 



The chief advantage in focusing the 

 study on the single gene is that this 

 makes it possible to substitute the di- 

 rect experimental analysis of specific 

 mutants for the application of general- 

 izations assumed to apply to mutations 

 at all loci. Each mutant studied may 

 add to the background of detailed in- 

 formation available for the diagnosis of 

 other mutants of the same gene. 



An important further advantage is 

 that the specific loci selected for study 

 may be loci with unusual technical 

 advantages for the recognition and 

 analysis of their mutants. For example, 

 the genes R' and A^ in maize, like 

 other known genes in various species, 

 yield spontaneous mutants that are 

 clearly distinct from the forms pro- 

 duced by recognizable short defi- 

 ciencies at these loci. This does not 

 prove that the spontaneous mutants are 

 not due to still smaller deficiencies, but 

 it supplies a convenient screen for 

 identifying a large class of deficiencies 

 without further investigation. Another 

 very useful aid in discriminating be- 

 tween gene loss and gene alteration is 

 available for the recessive allele a. This 

 allele, although phenotypically dis- 

 tinguishable only by the loss of A 

 action, may be distinguished from gene 

 deficiency by its response to the mu- 

 tagenic gene Dotted (Df), in the pres- 

 ence of which it reverts sporadically to 

 the dominant allele A. The retention 



