MANNER OF PRODUCTION OF MUTATIONS 507 



a limited application of the mutation-by-breakage hypothesis. One is 

 that in cases in which there are known to be two adjacent loci, each 

 giving an effect somehow distinguishable from the other (such as those of 

 yellow and achaete, achaete and scute, scute and the lethal just to the 

 right of it, or lozenge in positions 1, 2, and 3), so very few of the mutants 

 showing apparently normal salivary-gland chromosomes involve altera- 

 tions of both loci at once, as might have happened on this view when the 

 supposed mutagenic break had passed between them.'' The explanation 

 that the breaks are within instead of between genes is hardly applicable as 

 an alternative here, in view of the objections which have already been 

 raised to this supposition (p. 451). Another difficulty, applying equally 

 to the association of mutations with breaks between genes and within 

 them, is the previously cited lack of cases of structurally changed chromo- 

 somes having mutational effects in organisms in which position effects 

 appear to be rare or absent, that is, in most kinds of animals and higher 

 plants in w^hich the effects of structural changes have been studied at 



all. 



Arguing otherwise, perhaps, are the Drosophila mosaics caused by 

 nitrogen mustard, reported by Auerbach (1950a, 1952), in which one por- 

 tion of the body contained a structural change with a break close to the 

 locus of white and another portion contained a viable mutation to white. 

 Before this evidence could become convincing, however, it would have 

 to be shown that this viable white represented a gene mutation rather 

 than a deficiency of the white locus (which is also known to be viable) 

 or some other structural change. Similar considerations apply to some 

 other reported mosaics. 



On the whole, then, despite the a priori plausibility of the view that 

 individual chromosome breaks and gene mutations are frequently associ- 

 ated, there is a distinct lack of support for it in some of the directions in 

 which such support might be expected. It must therefore at this time be 

 regarded as an unfounded possibility, with the exception of such com- 

 paratively rare cases as are brought about by the clustering of ionizations 

 (see pp. 522-523). 



6. LARGER SCALE EVIDENCE OF SPATIAL LIMITATION IN MUTAGENESIS 



BY IONIZATION RADIATION 



The fact that point mutations (including real gene mutations as well as 

 some of the small structural changes) and single chromosome breaks are 

 produced by individual hits (regardless of whether these are to be 

 regarded as ionizations, activations, or clusters) does not prove that these 



6 However, this observation would also seem to speak for the comparative rarity of 

 ultraminute deficiencies as compared with gene mutations involving these loci. By- 

 contrast, such cases do occur much more frequently close to heterochromatic regions. 



