246 



STABLER 



visible shadow, so to speak, in the 

 chromosome. Some alterations of the 

 gene-strings are readily detectable by 

 visible alteration of the chromosomes. 

 The cytogenetic analysis of individual 

 mutations provides a wholesome check 

 on hypotheses derived from the statis- 

 tics of mutation frequencies. 



An illuminating example of this is 

 afforded by certain interpretations of 

 the evidence on mutation rate as 

 affected by x-ray treatment and by 

 temperature. At an early stage in the 

 study of x-ray-induced mutations, 

 Delbrueck (2) constructed a tentative 

 "atomic physics model" of the gene, 

 as inferred from the frequency of point 

 mutations observed under varying 

 physical conditions. This has become 

 widely known through its application 

 and discussion in the engaging little 

 book What Is Lije? {3), published 

 several years later by the eminent the- 

 oretical physicist, Erwin Schrodinger. 



In this view, the gene is considered 

 a molecule, and the observed muta- 

 tions are considered to represent its 

 transitions from one stable state to an- 

 other, as a result of thermal agitation 

 or the absorption of radiant energy. 

 The linear-dosage curve and the con- 

 stancy of mutation yield, regardless of 

 variation in the time factor, show that 

 the x-ray-induced mutations result 

 from single "hits"; the constant pro- 

 portionality of mutation yields to ion- 

 ization, regardless of variation in 

 wavelength, shows that the unit "hit" 

 is an ionization. Calculation of the vol- 

 ume within which these hits must 

 occur to account for the mutations 

 observed provides a basis for estimat- 

 ing the average size of the gene-mole- 

 cules postulated. This turns out to be 

 of the order of 1000 atoms. The rela- 

 tive frequency of spontaneous muta- 

 tions at different temperatures permits 

 the calculation of the activation energy 

 required for the occurrence of a muta- 



tion, which turns out to be about 1.5 

 ev. Unstable genes are assumed to have 

 correspondingly lower activation en- 

 ergies, and the fact that temperature 

 affects their mutation rate less than 

 that of normally stable genes is in 

 agreement with expectation on this 

 basis. The energy spent in one ioniza- 

 tion is about 30 ev, and it is therefore 

 to be expected that irradiation will 

 cause the mutation of any of the genes, 

 regardless of their relative stability un- 

 der normal conditions. The propor- 

 tional increase in mutation rate will, 

 therefore, be much less for genes 

 distinctly unstable at ordinary tem^ 

 peratures than for genes of normal 

 stability. These expectations also are 

 realized. 



This is an impressive picture, but it 

 has been evident for many years that 

 it has no valid relationship to the ex- 

 perimental data from which it was 

 derived. The detailed analysis of in- 

 dividual cases among the x-ray-in- 

 duced mutations has shown clearly 

 that many of these result not from a 

 structural change in a gene but from 

 some alteration external to the gene, 

 such as physical loss or rearrangement 

 of a segment of the gene-string. We 

 have no basis for estimating the pro- 

 portion of such extragenic mutations 

 among the total of mutations observed 

 and no ground for assuming that this 

 proportion is the same among the mu- 

 tations observed under the various ex- 

 perimental treatments. 



The basis of the model is the as- 

 sumption that the statistics of observed 

 mutation are in fact the statistics of 

 structural alteration of the molecules 

 that constitute the gene-string. The in- 

 vestigations of specific mutations con- 

 tradict this assumption and show that 

 the model has no basis in reality. 



It is interesting to reflect that if the 

 determiners of heredity had chanced 

 to be of a lower order of magnitude, 



