Genie Control of Development 355 



arrangement either to be without an efiFect different from ci or to 

 increase the ci effect. The actual results parallel closely those tabulated 

 for R( + ) combinations, and are therefore regarded by Stern as an- 

 other illustration of interference. There is a final fact: when different 

 R(ci) alleles are put in an orderly series according to the vein- 

 destructive effect in the heterozygote R(ci)/+, and this series is 

 compared with a similar one for R(ci)/ci, the two series do not show 

 the same order. 



These are the facts, which are anything but simple. Stern asks 

 now: What distinguishes the position alleles from the ordinary alleles, 

 and what differentiates the different R ( + ) and R ( ci ) alleles from 

 each other? It is obvious that the answer involves the theory of 

 position effect as well as of standard mutation, and I have given my 

 own answer previously when discussing the Dubinin effect. Originally, 

 Stern made the assumption that the specific substrate for each gene 

 is present in a specific concentration in the neighborhood of the gene, 

 but in a different concentration in the other regions to which the 

 rearrangement had shifted the gene. Thus the different effects of 

 different "position alleles" would correspond to these different sub- 

 strates. (This is one form of theory of position effects in terms of gene 

 neighborhoods, which I could not accept in my discussion of position 

 effect. ) The details recorded now, however, have convinced Stern that 

 this explanation does not work, partly because the results would be 

 contradictory, partly because they would require close spatial ap- 

 proximation of the two "interfering" alleles within the nucleus, which 

 is not true of some rearrangements. Therefore, interference must be 

 mediated by primary or derived gene products or takes place at a still 

 later stage of the reaction chains inside or outside the nucleus. Thus, 

 on the whole, not much remains of the competition for substrate 

 theory or of the theory of interference, and, in spite of an extraordinary 

 amount of material and a brilliant analysis, the explanation of dosage 

 effects does not seem to have progressed much beyond the general 

 description with which I started the discussion, when presenting my 

 own views. With reference to the Bar case and the conclusions I drew 

 from it {I S C c aa), I might say that the general features of the ci 

 dosage work agree with a rather simple explanation in terms of re- 

 action kinetics, proportional to dosage, but that also secondary effects 

 have to be considered, the nature of which is not known. In addition, 

 the interpretation of position effect is decisive for that part of the facts 

 which involves the Dubinin effect. All these facts and discussions may 



