122 Nature of the Genetic Material 



An alternative would be to assume that no Bar gene is involved, 

 but only a disturbance of the genie balance by the duplications and 

 triplications which throws the eye facet development out of gear. 



However, both of these explanations were ruled out by the work 

 of Dobzhansky quoted previously, and much subsequent work in 

 which the salivary chromosome technique could be used. Elaborate 

 work by Offerman (1935) and Dubinin and Volotov (1936) proved 

 that, in the presence of a very large number of rearrangement breaks 

 in the neighborhood of the Bar locus, in each case the Bar effect 

 was produced without any change at the Bar locus or region itself. 

 The conclusion is inevitable that the Bar effect itself is a position 

 effect of the duplication break. As no individual Bar locus is known, 

 that is, no simple dominant or recessive point mutant (without 

 visible rearrangement) in the Bar region, it must be assumed that 

 such a locus exists, nevertheless, and shows all these "baroid" po- 

 sition effects. (According to Sutton, 1943a, the position effects are 

 due mainly to breaks next to two of the six bands which make up 

 the duplication.) Another possibility would be that a breakage effect 

 of a definite type, in this case Bar eyes, could be produced inde- 

 pendently of an assumed mutable locus, in the same region; this was 

 mentioned above as a general possibility. (We shall return to this 

 problem.) Thus we cannot avoid the conclusion that the Bar effect 

 is in itself not due to a double dose of a Bar locus but to a position 

 effect caused by the duplication break, the same cause as all other 

 baroid position effects of the Bar phenotype. There are a few com- 

 plications to this simple picture. According to Griffen (1941), the 

 Stone translocation, a baroid, acts at one end of the section which 

 is duplicated in Bar, but other breaks involving the left end of the 

 section also have baroid effects. This must mean that there is either 

 no (unknown) mutant locus producing a Bar effect, or that the 

 baroid position effect works over the entire section and not only for 

 certain breaks. This, of course, has nothing to do with the Sturtevant 

 position effect. 



Now the question arises as to what Sturtevant's position effect 

 is. It is clearly not the same thing as the Bar effect, which turned out 

 to be a position effect of a duplication break. This is of course an 

 unending cause of error, owing to the use of the same terminology 

 for two completely different things. The Bar effect is, as we showed, 

 a position effect of a duplication break (in the general meaning of 

 the term, breakage effect) and otherwise has nothing to do with 



