Phenocopy and Norm of Reaction 257 



temperature, with overlap into normal at the extreme end of the 

 series. There are mutants of all these types, which, in addition, are 

 rather insensitive in their action to genetic modification, and there are 

 others which respond very easily to genetic modifiers. The actions of 

 some mutants work with a threshold system that permits overlap of the 

 effect with normal in different degrees. This condition is described as 

 different degrees of penetrance. At one end are mutants which succeed 

 only rarely in passing the threshold and thus have such a low pene- 

 trance that they may be overlooked as genically controlled actions. At 

 the other end are the isoalleles which produce effects that are always 

 above the threshold of normalcy, producing the same phenotype as 

 the normal allele, and therefore can be found out only by roundabout 

 means. In between are all imaginable penetrance grades. These types 

 may be more or less subject to shifting by external or genie environ- 

 ment. All these variants tend to show that the mutant action has not 

 only a quahtative but also a quantitative element. 



From this conclusion it is only one step to the phenomenon of 

 phenocopy. If the quantitative element of mutation is so sensitive to 

 external and internal environment, it might follow that all or most 

 mutational effects take place within a spectrum of possible develop- 

 mental changes the range of which alone determines the possibility of 

 appearance of both mutational or environmental effects. The old, long- 

 forgotten, and erroneous, but nevertheless at that time justified dis- 

 cussion of so-called parallel induction was the outcome of this situa- 

 tion, which at that time was not analyzable. This old idea, derived 

 from certain experiments with butterflies (see discussion in Gold- 

 schmidt, 1927, 1938a), was that the same agent which has a purely 

 environmental effect upon the phenotype may sometimes simul- 

 taneously affect the genotype so as to produce a mutant with the same 

 phenotypic effect. 



Some of the classic experiments of Dorfmeister, Weismann, 

 Standfuss, Merrifield, and Fischer in changing the pattern of the 

 butterfly wing by appropriate temperature shocks seemed to require 

 such an interpretation, the facts themselves being thoroughly estab- 

 lished. The solution of this problem comes from the discovery of 

 phenocopy ( Goldschmidt, 1929Z?, 19S5a,b), which was made when 

 I was trying to check the possibility of parallel induction with a 

 genetically well-known organism, Drosophila. Even in the first experi- 

 ments it turned out that the phenotype of a large number of standard 

 mutants of Drosophila could be exactly copied as a non-hereditary 

 variation if temperature shocks were applied at a critical stage of 



