OF MUTANT CHARACTERS. 



205 



The independence shown in the arc pink and arc maroon crosses 

 was interpreted to mean that arc was in a separate chromosome from 

 these two and therefore probably in the second chromosome. That 

 this was the case was proved by the result of the cross of arc by black, 

 which produced in F 2 no double-recessive bLack arc (table 51). 



TABLE 51. PI, arc 9 X black <?; FI wild-type 9 9 X 

 F\ wild-type cf cf . 



While this result was accepted as providing that arc was linked to 

 black and was therefore second-chromosome, it was not regarded as 

 proving absolute linkage of black and arc, but merely linkage so close 

 that two cross-over gametes had not chanced to meet. The correct 

 interpretation that the 2:1:1:0 was the result of no crossing-over 

 in the male was not suspected; however, it was considered remarkable 

 that all the autosomal linkages thus far encountered had been so 

 extreme. 



LOCUS OF ARC. 



In order to conduct a back-cross test of the amount of crossing-over 

 between black and arc it was necessary to obtain the double recessive 

 black arc. By accident the most advantageous method was used, 

 namely, F 2 blacks and F 2 arcs were mated together. Several mass- 

 cultures of this sort were started from B 42 and fortunately black 

 arc flies appeared in F 3 . From these a pure stock was made. 



The actual back-cross test was not carried out for some months; 

 meanwhile the black vestigial back-crosses had demonstrated that in 

 that case at least there was no crossing-over in the male. It was 

 necessary to test this fact by other cases, since the purple vestigial 

 back-cross had previously given a few apparent cross-overs in the male. 

 " Coupling" back-cross tests of the female and of the male were there- 

 fore started at the same time from the mating of black arc male to 

 wild female. The test of the male furnished 306 flies, every one of 

 which was a non-crossover (table 52). 



The result of the male test was particularly striking in view of the very 

 free crossing-over shown by the parallel tests of the female (table 53), 



In the female tests there was a total of 798 flies, of which 286 or 

 35.9 per cent were cross-overs. 



