CONTINUED SELECTION IN HYDRA 



23 



TABLE 2. 



Distribution of variations in the number of tentacles of buds produced by the last 



selected generation 



The difference in the average number of tentacles (0.040) is 

 not significant but the little that appears is in a direction the 

 reverse of that to be expected if variations within the clone are 

 inherited. 



In a consideration of diverse races the range and extent of vari- 

 ation may be of equal importance with the mean. The standard 

 deviations (a), measures of the amount of variability, are not 

 significantly different for the two groups. The range of varia- 

 tion must be considered in two ways; first, with respect to indi- 

 vididual polyps, second, with respect to variation in the mean 

 number of tentacles of different lines. As appears in table 2, the 

 polyp with the fewest tentacles (4) appeared in the group se- 

 lected for a large number, that with the most (9) in the other 

 group. The range of variation of individual polyps, like the 

 difference between the averages, is the reverse of that which 

 should appear if selection were effective. The mean number of 

 tentacles of all buds produced by each parent in the last selected 

 generation is given in table 3, where the constants are arranged 

 in the order of magnitude. Here again the line with the highest 

 average belongs to the group selected for few tentacles and that 

 with the lowest to the group selected for many, although the 

 differences are not significant. 



Most theories of evolution by continuous variation assume that 

 the range of variation increases as the mean of the selected group 

 diverges from the racial mean, so that the average of the selected 

 group may be brought beyond the original range of the race by 

 the continued selection of extremes. In this experiment no wid- 

 ening of the range was found : the extent of selection, the differ- 



