346 ACTION OF NATURAL SELECTION 



of the condition of the young shell, from which this 

 adult was formed by the subsequent deposition of new 

 material. By measuring the upper whorls of the adult 

 shells, one is accordingly able to determine the char- 

 acters possessed, not by all young shells, but by the 

 young shells which were successful in attaining the 

 adult condition. How would measurements on such 

 adult shells compare with those on young and growing 

 shells, some of which would almost certainly undergo 

 destruction before attaining their full development? 

 To answer this question, Professor Weldon measured 

 with great exactness the radius of the spiral at various 

 (angular) distances from the apex of the shell in 100 

 adult individuals, and also in 100 young individuals of 

 less than half their length. The means of the values 

 so obtained were practically identical in the two classes 

 of shells, so it seems to follow that the mean spiral of 

 young shells is not altered during growth by any process 

 of selective destruction. On the other hand, the varia- 

 bility of the radial spiral measurements was consider- 

 ably greater in the young shells than in the adult ones 

 (on an average, in the proportion of 120 to 100 for the 

 first whorl and a half.) Hence we may conclude that 

 during the growth of this mollusc some processes are at 

 work which effectually eliminate the abnormal shells 

 more rapidly than the others, and so diminish the 

 variability of the survivors. As the average character 

 of the race does not undergo any change, it follows that 

 the abnormalities eliminated are evenly distributed 

 about the mean. Such a process of selection has been 

 termed by Professor Pearson * periodic. 

 * " Grammar of Science," p. 413. 



