Payne: Variations and Selection 



17 



analyze this effect or its cause. A few observations have been 

 made, however, during the course of the experiment. Mac- 

 Dowell beheved that the quantity of food influenced the num- 

 ber of bristles on the thorax. The first flies which hatched 

 from any given bottle had more bristles than those which 

 hatched later. He believed temperature to have but little 

 effect. In my work food seems to play only a minor role. It 

 is possible by starvation to produce small weak flies. Such 

 flies have fewer extra bristles, but where the food is kept in 

 good condition, the last flies to appear in a bottle have as many 

 bristles as the first. My experiment has been conducted at 

 room temperature, and this means considerable fluctuation. 

 While I have not made definitely controlled experiments, my 

 observations indicate that low temperature is more favorable 

 for the production of extra bristles than a higher tempera- 

 ture. How it produces its effect, I do not know. 



The Analysis of the Results of Selection 



By a comparison of the males and females in Table I it 

 is noticeable at a glance that the females have a higher num- 

 ber of extra bristles than the males. As previously stated 

 MacDowell ('15 and '17) obtained similar results in his 

 experiments on Drosophila, and thought this difference due 

 to a difference in size. The females are larger than the males. 

 In my own experiments this difference in bristle number in 

 the two sexes is not due merely to a difference in size. Dur- 

 ing the course of my observations it was often noticed that 

 the smaller flies had more bristles than the larger ones. This 

 led me to doubt the applicability of MacDowell's conclusion to 

 my own work. The mean bristle number of the males and 

 females from the eighth to the thirty-eighth generations has 

 been plotted in Figure 4. The two curves show a marked de- 

 gree of parallelism and remain about one unit apart. Since the 

 difference is a sexual one it seemed to me that it might be due 

 to a sex-linked factor which when homozygous produced a 

 more marked effect than when heterozygous. MacDowell tested 

 this possibility in his experiment by crossing extra males and 

 females to the wild. In both cases there was a dominance of 

 the normal in and in Fo the ratios of the normals to extras 

 were the same and the distribution of the extras was similar. 

 I have used a different method in making my test. In fact, 



3—12487 



