Chromosomes and Genes 129 



effect might be more easily found because heterochromatin breaks 

 more easily; it has been much studied because the mottling appeared 

 more intriguing than the simple effect. Furthermore, simple position 

 effects will, as a rule, be overlooked and discarded as ordinary mutants, 

 if for some reason a cytological check is not made. (See my yellow- 

 white case, I 3 B Z?). There can be no doubt, on the basis of our 

 former discussion (l 2 C d bb), that the difference is purely quanti- 

 tative. In the simple position effect, Lewis' S effect, the action of the 

 rearrangement (i.e., pattern change) is so strong that it is com- 

 pletely above the threshold for the process in question. In the white- 

 eye case, this means that in all ommatidial pigment-forming cells, 

 the synthesis of pigment precursors is inhibited (not going into the 

 well-known chemical details) just as in the white mutant. In the 

 presence of the heterochromatic break the same inhibition occurs, 

 but it acts so near the threshold that random variation among the 

 different pigment cells pushes it in some cells below the threshold and 

 pigment is formed; that is, mottled eyes appear. This statistical condi- 

 tion can be pushed in either direction by appropriate modifier action 

 (e.g., extra heterochromatin), even to the point of disappearance of 

 mottling (I 2 C d bb). This shows clearly that the two types are 

 one and the same thing with varying degrees of action around a 

 threshold condition. I realize, of course, how difficult it is for Lewis 

 to accept this solution, as he adheres strictly to the one gene — one 

 action concept, which, if applied also to position effect, requires a 

 unitary result. But we shall see later that the concept does not work 

 for ordinary mutants either. 



There are, of course, loci like white, yellow, scute, for which 

 both types of position effect are known, as expected. Different facts 

 and deliberations have led E. Sutton Gersh (1952) and Baker (1953) 

 to deny a basic difference between mottled and non-mottled position 

 effect. Though Lewis' S and V position effects are clearly variants 

 of the same phenomenon, there is a possibility that a type of position 

 exists which does not fit the definition of a break acting as if a 

 nearby intact locus had mutated. I refer to the repeatedly mentioned 

 possibility that a dominant "inseparable from a rearrangement" is 

 actually a position effect per se; this would constitute a direct effect 

 of breakage and change of pattern without relation to a special 

 locus. In the present section I have already emphasized the fact 

 that the existence of this type of position effect is very difficult, if 

 not impossible, to prove. 



Still another point must be mentioned: the relation of return 



