596 RADIATION BIOLOGY 



of cases by treatments with given chemicals, such as acriflavin (Ephrussi, 

 1951). It is not to be expected that these seeming mutations would 

 respond in the same way to different mutagenic agents or to changes in 

 their dosage, as would point mutations in the nuclear chromosomes. 



In Neurospora, however, as Tatum (1946) and Giles (1952) have both 

 pointed out, all the self-perpetuating changes so far studied with reference 

 to their mode of transmission have proved to have their basis in the 

 nuclear chromosomes. Hence the cited evidence from this organism, at 

 least, does show clearly the existence of significant differences in muta- 

 tional spectra, in other words, that under different conditions of muta- 

 genesis the mutabilities of different loci will vary, relatively to one 

 another. It is natural to refer these evidences of selectivity to differences 

 in the composition of the genes. It should not be forgotten, however, 

 that the materials surrounding different regions of the chromosome also 

 differ to some extent, usually because of products locally formed under 

 the influence of genes of the same regions (as illustrated by nucleoli, 

 centromeres, secondary constrictions, heterochromatic regions, chromo- 

 meres, etc.). It is possible that these differences, external to the genes 

 themselves, may also play a part in conditioning the relative mutabili- 

 ties which they exhibit under different circumstances, in cases in which 

 genes at different loci are being compared. 



Nothing has been said thus far about the various studies dealing with 

 the relative frequencies and types of induced mutations of genes at 

 different loci, or of different alleles of the same locus, or of "direct" as 

 compared with reverse mutation, under fixed conditions of mutagenesis. 

 This large subject is primarily of interest to the geneticist, as showing 

 the differing mutational potentialities of genes rather than the means of 

 changing them. It is only, when the relative frequencies of the different 

 types, i.e., the mutational spectra, have been studied under different 

 conditions of mutagenesis, or in response to mutagenic agents and "spon- 

 taneously," that light may by present means be thrown on the mecha- 

 nisms involved; for only then has the selectivity been shown to involve 

 the mutagenic agent as well as the genetic material itself. These latter, 

 therefore, are the cases which have been reviewed in the preceding 

 account. 



As for the studies in which spectra, either for varied loci or for the 

 varied mutations of a given locus, have been made only under one pri- 

 mary condition of mutagenesis (e.g., by ionizing radiation), it is sufficient 

 to say that they have already allowed the emergence of a number of 

 principles. Most of these principles were put forward long ago, on the 

 basis of observations on spontaneous mutations (MuUer, 1921, 1923), 

 and have since been confirmed on mutations induced by radiation. 

 Among these principles is the fact that the expressions of different alleles 

 arising by one step from the same gene can differ qualitatively as well as 



