260 



THE SECOND-CHROMOSOME GROUP 



examining the character of the FI flies, and it was not until the 

 F 2 began to hatch that it was realized that the other alternative 

 was correct that "star," as the character was called, was an auto- 

 somal dominant. Two of the F! pairs gave in F 2 no star whatever (1627, 

 1629), while a third pair (1628, table 108) gave stars among both males 

 and females to the extent of half the flies (52 per cent). The fact that 

 half the flies were stars showed that this culture 

 came from a heterozygous dominant and a wild- 

 type FI pair. That star was an autosomal dom- 

 inant was proved by the sister cultures which 

 gave no stars; had star been sex-linked all the 

 FI females would have been star and hence every 

 F 2 pair should have given results like those of 

 culture 1628. 



These facts were confirmed by the results of 

 further tests of star males; for star males out- 

 crossed to wild females gave in FI stars to the 

 extent of half the flies (table 109, 337 stars in a 

 total of 683, or 49.3 per cent), and the stars were 

 evenly distributed among the males and females. Had star been sex- 

 linked, none of the males but all of the females should have been star. 



TABLE 108. PI, star cf X wild 9; FI pair (Fifties chosen at random). 



TEXT-FIGURE 83. Star eye, 

 showing the arrangement of 

 the facets and hairs. Com- 

 pare with the normal con- 

 dition shown in plate 10, 

 figure 3c. 



TABLE 109. Pi, star cf X wild 9 . 



LETHAL NATURE OF THE HOMOZYGOUS STAR. 



At the same time that the male out-crossed tests were made, a few 

 pairs of star female by star male were mated. In the next generation, 

 which corresponded to an F 2 , the flies in one culture (1739, table 110) 

 were exactly two-thirds star and one-third wild-type, which is the 

 typical yellow-mouse ratio that had already been met with in Dro- 

 sophila in the case of streak. The other culture (1740, table 110) gave 

 nearer to a 3 to 1 ratio. Further matings were necessary to be sure 



