OF MUTANT CHARACTERS. 



191 



The best type is that known as "alternated, " where the successive genes 



b + c 



alternate between the two chromosomes 



-, so that the max- 



+ Pr + 



imum of evenness of distribution of characters is attained. It requires 

 double crossing-over to put all the mutant characters in the same 

 individual, and accordingly the alternated experiment gives a minimum 

 number of the combinations that are most inviable. This principle 

 becomes still more important in more complex experiments, as, for 



, S' H- b + c + s p 

 example, - -*. 



+ d + p r + p x + 



The next mutant whose locus was mapped with reference to purple 

 as a base was "streak," a dominant character which shows as a dark 

 streak from the scutellum forward along the dorsal region of the thorax. 



TABLE 44. Streak purple curved back-cross data. 



The triple black-cross streak purple curved, which was made in two 

 of the four possible ways and is therefore partially balanced (table 44), 

 showed that streak is far to the left of purple, that is, beyond black, and 

 in the opposite direction from vestigial and curved. The streak 

 purple cross-over value was 35.0, which showed that streak is so far to 

 the left of purple that only an approximate calculation of its position 

 could be made from the data. In a region of such length the correc- 

 tion to be supplied because of double crossing-over is quite large and 

 correspondingly inexact. On the basis of data that have since become 

 available it appears that there is about 37.3 per cent of separation 

 between streak and purple. Purple has since played an important 

 role in the mapping of several other genes, the details of which will 

 appear in accounts of these mutations. 



A SUMMARY OF THE LINKAGE DATA INVOLVING PURPLE. 



Besides the data reported in the various sections of this paper, there 

 are available data from three other principal papers : Bridges's study 

 of age variation in crossing-over (J. E. Z., 1915), Muller's study of cross- 

 ing-over by means of the progeny test (Am. Nat., 1916), and Plough's 

 study of temperature variations in crossing-over (J. E. Z., 1917). 



