292 PHYSIOLOGICAL GENETICS 



well as the place of it stated. Thus, a very complicated system 

 isbuill up, bu1 it would be diflicull to state whether or not this is 

 necessary and whel her or not one could derive from it conclusions 

 regarding the nature of the gene. A remarkable but not yet 

 completely analyzed fact has been contributed by Kostoff (1935) 

 who finds gene instability in form of variegation after species 

 crosses in Nicotiana. 



If we think of the discussions at the end of the last chapter 

 (Goldschmidt, Haldane) and the facts presented now, it appears 

 more reasonable to assume that the phenomenon of unstable 

 genes has to do with the method of "copying" the mother gene, 

 which may be visualized in quantitative terms of molecules 

 (Goldschmidt) as well as of side chains (Correns). In this case, 

 either no conclusion upon the nature of gene mutation may be 

 drawn, or only one of a quantitative type. 



The first alternative would be realized if Stern's (1935) 

 hypothesis should turn out to be true, viz., that the results 

 ascribed to unstable genes are actually consequences of somatic 

 crossovers after certain chromosome rearrangements, involving 

 also a position effect upon dominance. 



C. Mutations as Changes of a Stereoisomers Type 



If the gene is a molecule, or a group of identical molecules, a 

 mutation might possibly occur through the formation of stereo- 

 isomers . As a matter of fact, some writers actually speak of 

 mutant genes as stereoisomeres. I do not know, however, of any 

 geneticist who has tried to derive such a hypothesis from special 

 facts, and thus far no facts seem to be known that could be better 

 coordinated by this idea than by any other. A chemical fact 

 might be mentioned, however, in this connection. According 

 to Ruzicka, the stereoisomeres of the male sex hormone have no 

 hormonic effect at all, whereas many of the isomeres of the 

 female hormone act alike. This one example demonstrates that 

 it will be difficult to find rules regarding stereoisomeres which 

 will enable one to describe the genetic facts in simpler terms, i.e., 

 to explain them (see also the theory of Karczag, 1928). 



D. Mutations as Polymerizations 



The idea that mutations might be polymerizations (or depoly- 

 merizations) of large organic molecules has been used by Baur 



