MANNER OF PRODUCTION OF MUTATIONS 527 



effects on the same locus. Moreover, several different considerations 

 render it more probable that the gene is fiberlike or sheetUke in shape 

 (even though with coils or foldings) than that it forms a compact more or 

 less spherical mass, and if this is true of its sensitive volume, there would, 

 for any given volume, be far less difference in the efficiency with which 

 the different types of radiation produced mutations in an individual gene 

 than is indicated in the tables used. 



Lea and Catcheside have regarded this calculated sensitive volume as 

 representing approximately the size of the gene itself, on the ground that 

 in the case of enzymes and viruses the sensitive volumes, calculated on 

 the same assumptions from the frecjuency of their inactivation by radia- 

 tion, had been found to correspond fairly closely with the known sizes of 

 these particles. In such experiments the particles had been in a medium 

 containing a small amount of other protein, to combine with the activated 

 radicals produced in the exterior water which otherwise would have 

 reacted with the particles and caused additional inactivations in them 

 (as shown for viruses by Luria and Exner, 1941). Lea and Catcheside 

 assumed that the nuclear medium must contain sufficient protein to 

 protect the genes in this way. This is not so self-evident, however, and 

 it is even possible that, in specialized substances surrounding the genes, 

 reaction chains might be started which in turn affect the genes. More- 

 over, it does not seem evident that a process (ionization) which always 

 destroyed or completely inactivated a virus or enzyme particle would 

 always cause mutation in a gene, especially since it is known that, when 

 the gene is caused to mutate, it is usually not destroyed. There might 

 very well be stabilizing mechanisms incorporated in the chromosome 

 material of- an advanced organism (including a more decided "cage 

 effect" of its structure) which a more primitive particle did not possess. 

 In that case most ionizations which arose within it might fail to produce a 

 permanent change. 



A volume 6 m/i in diameter is about a thousandth or less of the "maxi- 

 mum" gene size which had been calculated by Muller (1929b, 1935) by 

 several different methods. The term "maximum" is used here to mean 

 the size which a gene would have on the admittedly only limiting assump- 

 tion that a chromosome, when in the condensed state in which it occurs 

 at metaphase or in a spermatozoon, contained no material other than its 

 genes. It seems likely that even when in this form the chromosome does 

 contain some other material. However, it seems very unlikely, especially 

 in view of the survival value of having the spermatozoon trimmed down 

 to the smallest possible dimensions, that the gene material should occupy 

 only a thousandth part of the chromosome's bulk. Moreover, if this 

 were the case, the relative constancy in the amount of desoxyribonucleic 

 acid found per chromosome-set in the nuclei of virtually all cells of a 

 given species, at all stages (a principle recently confirmed by Patau, per- 



