212 THE VARIATION OF ANIMALS IN NATURE 



(19) di Cesnola (1907) : comparison of earlier and later whorls 

 of the shell of Helix ( = Arianta) arbustorum. 



The procedure was identical with that of the preceding 

 studies (17), (18). The characters of the young shells were 

 similar to those of the adult. ' The mean character does not 

 sensibly alter during growth, but is the same in young and 

 adult.' The same difference in the variability of young and 

 older shells was found as in Clausilia, and was held to prove the 

 occurrence of periodic selection. 



The same criticism may be applied to this study as to (17). 



We give in tabular form what we hope is a fair assessment 

 of the value of these studies. 



(1) 



Selection- 

 probable 



Lutz (1915) ? 

 Crampton (1904) 

 Thompson, Bell and 

 Pearson (191 1) 



(2) 



Analogy with natural 

 process doubtful 



di Cesnola (1904) 

 Poulton and 



Saunders (1899) 

 Boettger (1932) 

 Bumpus (1899) 

 Beljajeff (1927) 



(3) 



Other explanations 

 possible 



Weldon (1899) 

 Kane (1896) 

 Lutz (1915) 

 Weldon (1901) 

 di Cesnola (1907) 



(4) 

 Procedure defective : or 

 numbers too low 



Harrison (1920) 

 Trueman ( 1916) 

 Jameson (1898) 

 Kane (1896) 

 Boettger (1931) 

 Davenport (1908) 

 Kellogg and Bell 



(1904) 

 Bumpus (1899) 



(5) 



No selection 



found 



Haviland and Pitt 



(1919) 

 Pearl (191 1) ? 



Weldon (1904) 



(6) 



Selective agency unknown 

 or doubtful 



Harrison (1920) 

 Jameson (1898) 

 Kane (1896) 

 Crampton (1904) 

 Thompson, Bell and 



Pearson (191 1) 

 Bumpus (1899) 



It will be seen that on this analysis (which should be 

 checked by reference to the actual accounts) there is a little 

 evidence suggesting a significant difference between survivors 

 and eliminated. It must be admitted that any amount of 

 positive evidence, however slight, is of value. On the other 



