GENE CONTROLLED PROCESSES I9I 



reminded of the "alternative reaction systems" invoked to explain how 

 a single gene can, in some cases, decide between two complicated types 

 of sexual differentiation. But if the genotype really provides alternative 

 methods of reacting, or alternative valleys along which the develop- 

 mental processes may go, it should be possible to reveal these poten- 

 tialities without using the "realizer" genes which normally bring them 

 into action. This requirement can be fulfilled. Goldschmidt,^ in par- 

 ticular, has shown that if normal flies are subjected to severe environ- 

 mental influences (sub-lethal temperatures, etc.) a small percentage of 

 modifications appear which exactly parallel the changes produced by 

 genetic factors. These so-called phenocopies are evidence that the 

 alternative valleys are given even in the normal genotype. They can be 

 said to provide evidence of positive values for the expressivity of genes 

 which are not actually present! They are not, of course, inherited, 

 though other hereditable changes may be produced by the same 

 agencies (p. 279). Their importance is that they reveal the underlying 

 "landscape" of the normal genotype; unfortunately, the methods 

 known for producing them are as yet fairly imspecific and do not 

 reveal much about the nature of the "push" required to carry the 

 developmental reactions into the side-valleys. 



(c) Specificity. — The last variable to be discussed in relation to gene- 

 effects is the specificit}', which is the name proposed for the varia- 

 tions in the actual qualitative nature of the effects. That is to say, it is 

 the variation in the course of the side- valleys, giving a variation not in 

 the amount of the final effect but in its kind. This again is controlled 

 largely by the rest of the genotype, partly by the environment. The 

 variations which have been worked on are mainly variations in pattern, 

 and are discussed on p. 200 in connection with other pattern effects. 

 ^ Goldschmidt 1935, 1938, Friesen 1936. 



