Genie Control of Development 369 



as well be described as ordinary multiple alleles of lowest action 

 (which we did above when calling +^ = ci^), it is expected that 

 examples of "quantified" potencies are frequent among multiple alleles. 

 This is actually the case, and formerly led me to the assumption that 

 multiple alleles are diflPerent gene quantities because they act like 

 dosage differences. If the theory of the corpuscular gene were correct, 

 this would still be a good hypothesis for cases in which (as in ci) 

 dosage of multiple alleles parallels actual dosage actions and fits in 

 with them. In the discussions of the subject a generation ago, it was 

 emphasized that in the many cases of multiple alleles in which there 

 is a pleiotropic effect upon different traits, the order of the alleles, if 

 grouped according to individual pleiotropic effects, frequently does 

 not coincide with the order of alleles, if classified according to the 

 main action of the gene locus. We shall discuss the present aspect of 

 this problem in the chapter on pleiotropy. 



In the same category of facts belongs seriation of environmental 

 effects upon a multiple allelic series, or of modifier effects, when 

 checked for the different pleiotropic actions. However, such facts in 

 themselves do not argue against the existence of potencies paralleHng, 

 in their orderly effect, dosage effects. The development of different 

 traits may or may not be determined by the same primary effects of 

 genie action in proportion to potency. In some cases it may safely be 

 assumed that a single primary effect trails automatically behind it a 

 series of consequences in different organs, a syndrome of effects. 

 (See the vast quantity of work in developmental genetics in rodents 

 by Bonnevie, Gluecksohn-Schonheimer, Griineberg, much of it re- 

 viewed in my book, 1938, in Gluecksohn-Waelsch, 1953, and in 

 Griineberg's book, 1952. ) But in other cases, probably more frequently 

 with mosaic development, as in Drosophila, the pleiotropic effect 

 may be the result of interference of one reaction chain and (or) its 

 products with other simultaneous ones, an interference which might 

 be orderly, that is, parallel for all of them, or completely irregular, 

 depending upon timing, threshold, and substrate conditions at the 

 point of interference. Thus, on the basis of the facts discussed so far, 

 there is no reason to doubt that multiple alleles act in either case 

 according to their potencies. 



But this conclusion is evident only within the classic theory of 

 the corpuscular gene. If more modern ideas of the type discussed 

 above are accepted, mutant loci being changes of the normal order of 

 a pattern, we should have to make special assumptions for orderly po- 

 tency effects (e.g., that they parallel the degree of scrambling of the 



