No. 526] 



VARIATIONS IN UROSALPINX 



583 



TABLE III 



Two Lots of Shells of One Thousand Each to Show Extremes of 



























62 



j 



63 



Devil's Foot 

 Prince's Iliiy 



1 













99 

 5 



141 l'X 



9 36 



217 



163 73 

 65 75 



2. 

 13 



> 14 



2| 84 



Devil's Foot 

 Prince'- l',:iy 



3 

 83 



117 



75 



88 



83 





33 



15 







;>*.:;;>* 

 6:1.976 



.849 

 $.407 



.0279 

 .051-1 



Difference 



5.(118 



.558 





4. ^4 Comparison of Atlantic ivith Pacific Shells. — In 

 the summer of 1898, following the method employed by 

 Bumpus with Littorina, 7,006 Urosalpinx shells from 

 various localities near Woods Hole were obtained and 

 measured, with which were compared 1,528 introduced 

 shells from two localities in California. From Table IV, 

 in which these data are brought together, it will be seen 

 that Bumpus 's work on Littorina is apparently confirmed, 

 that is, there is more variability in the introduced than 

 in the endemic form, although the margin of probable 

 error permits an overlapping in some instances. 



Cali- f Belmont Beds i 1,008 59.051 3.023 ±.0454 

 fornia \ San Francisco Bay j 220 | 60.892 3.361 | ±.0703 



Difference .803 [ 



5. Shells from Buzzard's Bay and Vineyard Sound 

 Compared— For the sake of scientific peace of mind the 



