286 



CALVIN B. BRIDGES 



some. The test of the amount of crossing over in the female 

 between the loci purple and vestigial (table 3) had given a 

 crossover value of 9.1 units. The next crossover value to be 

 worked out was that of black vestigial as about 20 units (Mor- 

 gan, '12). 



A THREE-POINT MAP 



With these two values alone it was not possible to determine 

 the relative order within the chromosome of the three loci in- 

 volved; it was apparent that black was farther away from ves- 

 tigial than from purple, but it could not be told whether it 

 lay on the same or on the other side of vestigial from purple. 

 This value was expected to be one of two values depending on 

 the order of the genes; it should be an approximation to either 

 the sum (20 + 9 - 29) or the difference (20 - 9 = 11) between 

 the black vestigial and the purple vestigial values. To carry 

 out a back-cross experiment for black and purple it was first 

 necessary to make up the double recessive. No easy task was 

 anticipated in this, for it has just become known that on ac- 

 count of no crossing over in the male no double recessive could 

 be obtained in F2, and in fact none was obtained (table 10). 

 As expected, the r2 ratio approximated 2 : 1 : 1:0. Three sorts 

 of F3 mass culture matings were made: black X black, purple X 

 purple, and black X purple. Of these matings the last type is 

 by far the most valuable, since in case one of the flies happened 

 to come from a black purple crossover egg X a black sperm it 

 would give some purple offspring when crossed tojj purple; and 

 these inbred, would give the required black purples_;^as a quarter 



TABLE 10 



Pi mating, purple cf X hlack 9 ; ^1 mating, wild-type ? 9 and & cf 



