254 BUMP US. [VOL. I. 



any British locality; but do the extremes of variation of the 

 American shells from all localities equal or exceed the extremes 

 of variation of the British shells from all localities ? With the 

 data at hand it is possible for us to arrive at a reasonably cer- 

 tain conclusion, though to answer the question with absolute 

 certainty one should have a very large and representative col- 

 lection of shells from many localities in both countries. 



The most elongated shell among the British series was col- 

 lected at Kincardineshire. Its index is 81. The most elon- 

 gated American shell was collected at Seaconnet, and its index 

 is 80. While the most ventricose British shell has an index of 

 98 (vide Chart I), the most ventricose American shell has an 

 index of 104 (vide Chart XIII). The most extreme cases of 

 variation in stature are, then, presented by the American shells. 

 This is quite a different thing from extreme amplitude of vari- 

 ation in particular localities, and, while it may result from the 

 fact that a larger number of American localities have been 

 examined, it seems hardly probable, from the data at hand, 

 that an equal number of British localities would yield an equal 

 number of such extreme variations. 



The British shells from three localities taken at random 

 gather in 17 grades, from Si f /> to 98/0. The American shells 

 from ten localities gather in 24 grades, from 8o r /o to 104/0. 

 Moreover, if we can show that the extremes of variation of 

 3000 American shells from the three localities which offer the 

 least variation are further removed than the extremes of the 

 3000 British shells taken at random, then our position is further 

 strengthened. This we can do, for the combined amplitude of 

 variation of 3000 British shells is 17, while the combined 

 amplitude of variation of the three least variable American 

 series is 19. 



A Third Test of Variability. 



Every one has doubtless observed the difference in the 

 general proportions of the body of the child and the adult. In 

 the former the head is relatively larger, the trunk longer, and 

 the legs shorter. To use a conchological term, the child is 

 more ventricose. Are the smaller shells, from the several 



