248 



acters could be readily demonstrated. 



The "ClB" technique with Dro- 

 sophila, designed by Muller, was ad- 

 mirably suited to this purpose, and 

 x-ray experiments with this technique 

 (6, 7) demonstrated beyond question 

 a very strong effect of x-rays on the 

 frequency of mutation. The total fre- 

 quency of lethals in the X-chromo- 

 some was increased more than 100- 

 fold. In addition, many visible muta- 

 tions were found, including dominants 

 as well as recessives, and including 

 mutants previously known from their 

 spontaneous occurrence as well as 

 many mutants not previously ob- 

 served. 



These experiments were promptly 

 followed by others designed to test 

 more critically the genie nature of the 

 induced mutations. The mutant lethals 

 might be suspected of being defi- 

 ciencies; even the visibles could con- 

 ceivably be due to short deficiency or 

 gene destruction. But if the treatment 

 could induce mutation to a variant 

 allele and could, in further applica- 

 tions, induce reverse mutation to the 

 parental allele, it was argued, the two 

 mutations could not both be due to 

 gene loss. Induced mutation and in- 

 duced reverse mutation at the same 

 locus were shown to occur in a num- 

 ber of loci of Drosophila in experi- 

 ments by Patterson and Muller {8) 

 and Timof eeff-Ressovskv {9). 



Subsequent experiments with a wide 

 variety of forms among the higher 

 plants and animals and with micro- 

 organisms showed the broad generality 

 of the effects of ionizing radiations 

 upon the frequency of mutation. In 

 later experiments, ultraviolet radiation 

 and various chemical treatments were 

 also shown to affect mutation fre- 

 quency. 



The analysis of the induced muta- 

 tions, however, soon indicated that the 



STADLER 



accepted definitions and criterions re- 

 lated to genes and gene mutations 

 needed reconsideration. 



The purpose of experiments with 

 gene mutation is to study the evolution 

 of new gene forms. The techniques for 

 studying gene mutation are, therefore, 

 designed to measure the frequency of 

 these changes in the genes. But a 

 change in the gene may be recognized 

 only by its effects, and it soon became 

 clear that various extragenic alterations 

 might produce the effects considered 

 characteristic of typical gene muta- 

 tion (/O). 



Thus the working definition of mu- 

 tation necessarily differs from the ideal 

 definition. It is this working definition 

 that must be considered in generalizing 

 from the experimental evidence. The 

 mutations experimentally identified as 

 gene mutations may include not only 

 variations due to alterations within the 

 gene but also variations due to losses 

 of genes, to additions of genes, and to 

 changes in the spatial relationships of 

 genes to one another. To identify these 

 mechanical alterations, certain tests 

 were applicable. But there was no test 

 to identify mutations due to a change 

 within the gene; it was simply inferred 

 that the mutants that could not be 

 identified as the result of specific me- 

 chanical causes were, in fact, due to 

 gene mutation in the ideal sense (//). 



When we conclude from an experi- 

 ment that new genes have been evolved 

 by the action of x-rays, we are not 

 simply stating the results of the ex- 

 periment. We are, in the single state- 

 ment, combining two distinct steps: 

 (i) stating the observed results of the 

 experiment, and (ii) interpreting the 

 mutations as due to a specific mecha- 

 nism. It is essential that these two steps 

 be kept separate, because the first step 

 represents a permanent addition to the 

 known body of fact, whereas the sec- 



