Genie Control of Development 353 



will be utilized less efficiently by the cooperation of +^ and the two 

 ci alleles than by the combination of +^ and one ci allele. Thus the 

 results can be explained formally. I cannot help saying that I do not 

 feel too happy about this theory, which appears to me as a restate- 

 ment of the fact of non-consistent effects in terms of different qualities 

 of the causative agent. If only nos. 1 and 3 of the enumeration on page 

 352 were available, we could think of a specific action of the Minute 

 deficiencies, since Minutes are known to interfere strongly with genie 

 actions (see Brehme, 1939, 1941; Goldschmidt, 1949). But no. 2 

 without a Minute shows the same discrepancy, and therefore the 

 combination ci ci + must be responsible. As we have noted. Stern 

 has recently given up the idea of competition for substrate. Thus, for 

 the time being, the riddle cannot be solved, and there must be some 

 feature of genie action which cannot be described simply in terms of 

 velocities of reaction and quantity of products proportional to dosage. 

 The second group of experiments by Stern involves the dominant 

 ci^. It is a strange fact that many multiple allehc series in Drosophila, 

 and in maize also, contain a dominant mutant. Frequently, but not 

 always, this is a visible rearrangement hke dominant eyeless^. The 

 dominant ci^ has the opposite action from ci, namely, Hke Bar, an 

 additive action toward abnormality in different doses. We question 

 at once whether or not dominance in these cases is based upon the 

 same causative condition ( see below ) . Thus ci^ > ci""" ci^ > ci^ ci^ 

 ci^, where > means more normal. The difficulties come in again in 

 the compounds: 



1. ci^/0 = almost normal 



2. ci^/-|- = greatly deficient 



3. ci/ci^ = greatly deficient, more than ci"/-)-. 



The great difference between 1 and 3 cannot be explained by a 

 position effect of the deficiency, which would make both alike. Stern 

 concludes that here a mutual interference of the alleles becomes 

 visible as the heterozygote cV^Z-h is less normal than ci^/0 and +/0. 

 If we express the action of ci^ in the same terms used for the Bar 

 series, as an inhibition of the speed of a reaction, while the ci and bb 

 effects control a smaller amount of reaction product than normal, we 

 would not expect that these two effects, involved as they are in dff- 

 ferent processes of genie action, could be additive. The inhibiting 

 action of ci^ upon a reaction velocity would decide the effect whether 

 an additional ci is present or not. In other words, we are not justffied 

 in treating ci and ci^ as if the same action in different directions was 

 involved and neither should we so consider the action of + in the 



