128 MUTATION AND PLANT BREEDING 



year later, admixture of sodium nitrite to the food of Drosophila 

 produced mutations in one experiment, but repetitions gave 

 negative results (76). These tests had been undertaken with the idea 

 that nitrite, like formaldehyde, might produce mutations by acting 

 on the proteins, which then were thought to be carriers of genetical 

 specificity. Quite recently, nitrite has been used again and with spec- 

 tacular success. Starting with very precise chemical predictions 

 about desamination of the bases in nucleic acid, a group of Ger- 

 man workers (63) found a method for the production of mutations 

 in tobacco mosaic virus treated in vitro. In this case, there can be 

 little, if any, doubt that mutations were produced by direct chemical 

 action on the genetical material, for treatment was equally effective 

 when applied to the naked RNA of the virus. Subsequently, sim- 

 ilar results have been obtained with the RNA of polio virus (14) 

 and the DNA of the pneumococcus-transforming principle (56). 

 Nitrite has also been used successfully for the production of muta- 

 tions in bacteriophage (91) and bacteria (48). Treatment of bac- 

 teriophage produced many sectored mutants, as would be expected 

 from chemical reaction with one of the two strands of DNA. When 

 an exceptional phage strain with single-stranded DNA was treated, 

 there were no sectored mutants (89). Although it seems plausible 

 to assume that nitrite also acts directly on DNA in bacteria, proof is 

 still lacking. 



In conclusion of this section I would like to say this. Work with 

 chemical mutagens has not thrown new light on the nature of the 

 gene, except insofar as the effects of certain mutagens can be under- 

 stood best from our present model of gene structure. In my opinion, 

 it would be regrettable if further research were directed wholly 

 towards the study of this particular class of mutagens. The very 

 variety of mutagenic chemicals shows that the stability of the hered- 

 itary mechanism is under complex control. Analysis of many 

 different types of mutagen may help us unravel some of these 

 complexities. It may also teach us how to loosen the control in ways 

 that produce desirable mutations for applied purposes. 



The Relation Between Intergenic and Intragenic Changes 



The question whether there is an essential difference between 

 intergenic and intragenic changes has been often discussed (62), 



