Genie Control of Development 343 



Miiller (1933) proposed names for mutant loci which act in the 

 way just considered, caHing them, in the order of the foregoing dis- 

 cussion, amorphic, antimorphic (partly neomoi-phic ) , and hypomor- 

 phic mutants; he derived the terms from the idea that the mutant 

 loci act in the same way or in an opposite way from the normal gene. 

 I am opposed to the use of these terms because they are linked to the 

 classic theory of the gene, and especially because they might foster 

 the idea that we are dealing with dijfferent kinds of genes or mutant 

 genes, while actually the differences lie in the type and kind of 

 developmental reactions which are affected by the mutant loci and 

 their different doses. Moreover, the terminology does not characterize 

 the real differences between reactions of the all-or-none effect type 

 (coincident with Muller's amorphs), and shifts in the kinetics of 

 reactions either by inhibition of determining processes of definite 

 speed (Muller's antimorphs) or by affecting the quantity of reaction 

 products (Muller's hypomorphs). 



In the dosage experiments the majority of genie actions were of 

 the last type. Mohr (1923fl) found that a great many mutants opposite 

 a deficiency showed an exaggerated phenotype, more extreme than 

 the homozygous mutation effect. I proposed (1927) the explanation 

 by dosage: one dose of the mutant locus produces less of the active 

 substance controlling the phenot\'pe than two doses do, and therefore 

 makes the mutant phenotype more extreme. A corollary of this was 

 the expectation that additional doses would make the phenotype more 

 normal. This turned out to be true of most recessive mutants. The 

 proof was supplied in Stern's classic work on the bobbed alleles in 

 Drosophila (1929fl). Bobbed shortens the bristles. By ingenious 

 experimentation Stern could build up flies with 1, 2, 3, 4 doses of bb, 

 and the length of the bristles was strictly proportional to dosage, the 

 wild-type length being the maximum possible (saturation point, 

 threshold effect). Other examples studied since have shown the same 

 result. We may safely assume that wherever the deficiency-exag- 

 geration effect is found, the same type of action of the mutant locus is 

 present. When the exaggeration effect is absent, it may mean either 

 that the mutant action is of the all-or-none type (e.g., one synthetic 

 step) or that the deficiency has a position effect which makes a/ — 

 identical in action with a/a. (See above under position effect.) Thus 

 it is demonstrated that the majority of recessive loci in Drosophila 

 affect such processes of the kinetics of development which, in a 

 general way, may be conceived as controlling quantities of end 

 products, whatever the direct interference with the kinetics may be. 



