Genie Control of Development 387 



acts upon the facet-producing reaction, whatever it is. But it is just as 

 possible that the action of the mutant v, which ultimately results in 

 the prevention of a synthetic step between two amino acids, is 

 primarily of a completely quantitative type, for example, one action 

 affecting the rate of some process, which can interfere with the facet- 

 producing reaction. In the wing mutants affecting facet number (cross- 

 veinless and cut), the probability of an interaction of two kinetic 

 conditions is more probable. This example shows the same interactions 

 at work which we met before, and also the same difficulties of identi- 

 fying exactly the type of interaction. This applies also to some in- 

 stances in which it is recorded that a mutant locus in control of a 

 definite biochemical situation simultaneously affects a morphological 

 character. Griineberg {1952a,b) recounts a series of mutants for coat 

 color with simultaneous morphological effects of a pathological type, 

 for example, skeletal abnormalities caused by disfunction of osteo- 

 clasts, absence of otoliths, or spina bifida. It is quite possible that an 

 abnormal biochemical condition of a pigment precursor directly 

 affects definite embryological processes. But just as well, the indirect 

 type of interference through kinetics may be at work. 



Only one more recent example of the type under discussior^ may 

 be scrutinized. Waddington (1953Z?) combined homozygotes of mu- 

 tants, all of which affect growth of one or more imaginal buds: 

 aristopedia, which turns antennae into tarsi and also affects legs; 

 vestigial, which affects wings; bithorax, which acts on dorsal and 

 ventral metathoracic buds; and others, all having strong morphological 

 effects. Many types of interactions were found. As I have already 

 pointed out (see paper by my student Csik, 1934, 1936, on such 

 interactions in the Drosophila wing and bristles; Goldschmidt, 1938a), 

 in such combinations, different results may be expected. The genie 

 actions may be concerned with processes so different that they will 

 not interfere with each other; thus the phenotype will exhibit a com- 

 bination of the two traits. (Examples are found in Csik's paper.) We 

 could assume that this is so when qualitatively different processes are 

 involved (e.g., processes of the one gene — one synthetic step type), 

 or actions at very different times. It is expected that this is rather rare, 

 since most simultaneous or consecutive reactions should interfere with 

 each other. The other extreme would be that different genie actions 

 concern the same process directly, or, more frequently, indirectly by 

 means of different approaches or points in common. Here the actions 

 may be additive or mutually inhibitive; or sometimes, when the effects 

 are quahtatively different, a compromise between the two effects may 



