116 Nature of the Genetic Material 



small rearrangements than the salivary chromosomes allow. Thus any 

 conclusion on mutation drawn from the number of gene mutations 

 versus rearrangements rests upon a very slender base and is, in the 

 best of cases, weighed in favor of point mutants. Even in the salivary 

 chromosomes the cytological evidence may be deceiving. Kodani 

 (1947) has shown that chromosome tips may contract and a single 

 thicker band be formed out of a series of distinctive bands. This shows 

 that some bands may be compound, and a break within them almost 

 impossible to detect. If we look dispassionately at the cytological 

 facts, we realize that rearrangements might be of any size from almost 

 the whole length of a chromosome to a single band, and even smaller 

 than this in the salivary chromosome containing up to 1,000 bands. 

 Hence the separation of gene mutation from rearrangement by 

 cytological criteria is completely gratuitous and is only an indication 

 of the Hmitations of our methods. The problem of whether there is a 

 difference between a radiation-induced mutant and a chromosome 

 break cannot be solved by the direct cytological method. It has to be 

 solved indirectly, from evidence derived from radiation effects. 



The most important group of facts tending to show, in my 

 opinion, that there is no difference between point mutants and chro- 

 mosomal breaks concerns the action of oxygen upon both phenomena. 

 Stone et al. ( 1947, 1948 ) have shown in their important experiments 

 that mutants are produced after the irradiation of the medium in 

 which bacteria grow, and it was considered most probable that the 

 production of peroxides by irradiation, especially the H2O2, is re- 

 sponsible (see also Wagner et al, 1950). Since that time, Thoday and 

 Read ( 1947 ) found that the amount of breakage also is increased with 

 irradiation in the presence of oxygen. It may be added that there is at 

 least a strong suspicion that chemical mutagenesis also acts via the 

 production of a peroxide (Auerbach, 1951). Such facts lead to the 

 question whether the dualism inherent in the Treffer theory in inter- 

 preting point mutants and chromosome breaks is still tenable. Ac- 

 cording to Timofeeff et al. (1947), Lea and Catcheside (1945), and 

 Lea (see Lea, 1947), a point mutation is the result of a single hit, as 

 discussed in Delbriick's theory, producing within the target, that is, 

 the gene molecule, a pair of ions. A break, however, is induced by 

 the production of a series of ion pairs during the passage through the 

 chromosome of primary or secondary charged particles. Chemical 

 changes resulting from this cluster of ionizations within the molecules 

 of the chromosome result in the chromosome breaks. Different radi- 



