NATURE OF THE GENETIC EFFECTS 399 



abnormad mutations (those to a less normal phenotype). Yet the fact 

 that they can occur at all shows that not all mutations are of the nature 

 of losses. Moreover, the finding that even amorphs can give reverse 

 mutations to or toward normal (sometimes, however, requiring two steps 

 for the actual attainment of normality) shows that, despite their having 

 lost the ability of producing a given biochemical reaction, the genes 

 themselves are in these cases still there, and retain much of their original 

 structure. 



Of course the mere fact that a mutation has caused a change in a given 

 mutant character to or toward normal is not in itself evidence that a true 

 reverse mutation of the given mutant gene has occurred. For a mutation 

 in a second, quite different gene sometimes has a so-called "suppressor" 

 effect, that is, an effect antagonistic to that of the mutant gene primarily 

 under consideration, so as to cause the phenotype of an individual having 

 both the original mutant gene and also the second or suppressor gene to 

 be more nearly normal than that of an individual with just the first 

 mutant gene. Therefore all suspected reverse mutations must be sub- 

 jected to genetic analysis before they can be known definitely to be true 

 reverses. Definite reverse gene mutations, proved to be such by genetic 

 analysis, have been obtained by ionizing radiation for a considerable 

 number of genes of Drosophila (Muller, 1928d; Patterson and Muller, 

 1930; Timofeeff-Ressovsky, 1929, 1931b, 1932, 1933a, b). They have also 

 been obtained in the mold Neurospora (Giles, 1952) by application of 

 both ionizing radiation and ultraviolet, and nutritional deficiencies have 

 thereby been restored. It is still a question, however, whether those 

 produced by ionizing radiation in Neurospora involve actual gene muta- 

 tions or structural changes; it seems not unlikely that some of them may 

 involve one of these phenomena and some the other. Moreover, it 

 cannot justifiably be assumed that even a "true" reverse gene mutation 

 necessarily restores the precise chemical configuration of the original 

 normal gene. 



Probably the most interesting class disclosed by the dosage studies is 

 that of neomorphs, already mentioned in Sect. 9. This class, like the 

 others, is found among both spontaneous and radiation mutations. An 

 increase in dosage of the mutant gene in this case increases the departure 

 of the phenotype from normal. Yet such a mutant gene does not cause 

 a reaction opposite or antagonistic to that of the normal allele, or one 

 that competes with the latter, since in these cases a change in the dosage 

 of the normal gene itself exerts no influence on the given effect. Thus the 

 neomorph is the cause of some reaction of a different nature from that 

 mediated by the normal gene, a reaction which in this sense is "new" to 

 the organism. So far, all the neomorphs studied appear to have been 

 cases in which the functioning of the normal gene had been altered by 

 the position effect of a structural chromosome change, rather than by a 



