Genie Control of Development 351 



independently according to its dose, as was assumed thus far in our 

 discussion also. This relation is, however, supposed not to work in the 

 hemizygous condition (opposite a deficiency). For example, Stern's 

 + V+^ is normal; +^/0, almost so; but ci/+^ has abnormal veins (+^ 

 is what we called ci^ because it acts at low temperature like a low ci 

 allele with a little penetrance, just as, e.g., vg"' in the vestigial series ) . 

 Stern calls this an interference between ci and +^. The dosage studies 

 had shown that ci acts in the same direction as + or +^, namely, 

 toward production of venation. Yet, when ci and one of the two + 

 alleles are brought together in the same nuclei, their joint effect is 

 smaller than that of the more eflfective allele alone. This suggests to 

 Stern some mechanism of competition. If + and ci use the same sub- 

 strate, present in a limited amount, a competition for it might deprive 

 + of its full share in favor of ci. If ci turned over the substrate less 

 eflBciently than + into the product of action P, the joint action may 

 be less than the better one alone. Since it is not known what the 

 reactions are, Stern prefers now the term "interference" to "com- 

 petition." 



We might look at the facts in a somewhat different way, assum- 

 ing that Stern's -f^ is not a + but a weak ci allele with a little 

 penetrance, ci^. If ci produces only little vein substance, ci/ci more, 

 and ci ci ci almost the full quantity, ci^ also would produce almost 

 the full quantity, much more than ci and ci ci. Therefore, on a simple 

 dosage basis, ci^/ci should produce an amount between ci^/ci^ and 

 ci/ci and nearer to the former, which is the case for nos. 1-4 in the 

 enumeration. What is now the expectation for ciyO? It should be less 

 normal than ci/ci^ on a pure dosage basis, but it is more normal. 

 Therefore, some feature additional to dosage should be involved and, 

 in our way of looking at the facts, it is not an antagonism of ci and 

 +^ (ci^) which has to be explained in addition to the dosage phe- 

 nomenon (which is clearly good also for the relation of nos. 5-7 in the 

 enumeration), but rather the action of the deficiency. Thus far we 

 assumed with Stern that ci/0 and so on has only a simplex action. But 

 we know now that deficiencies may have a position eflFect similar to 

 the mutant effect, as the action of many homozygous deficiencies 

 proves (see 1 S C c dd aaa). Thus, if in ciVO the deficiency has a 

 position effect not too far below that of ci, normality or almost 

 normality is expected. The simple dosage relation holds thus far, in 

 which the diflFerent potencies of different alleles act in the same 

 direction as real dosages. Therefore, neither competition for a sub- 

 strate nor interference is needed for this group of effects. (A check 



