1910.] NATURAL SCIENCES OF PHILADELPHIA. 23 



shows the various counties represented in the collection, and the 

 larger figures the total number of specimens from each county. While 

 an inspection of this chart shows a lamentably large number of coun- 

 ties wholly unrepresented, it is to be borne in mind that the physical 

 conditions of Michigan are much more uniform than in many other 

 States where there is a greater diversity of surface, and that, therefore, 

 as far as they go the specimens from the region south of the Saginaw- 

 Grand Valley may be fairly considered as representative of all the 

 counties in that district; the shells from the Grand Traverse region 

 as of those characteristic of the sandy plains of the northern part of 

 the Lower Peninsula, and those from the Upper Peninsula as repre- 

 sentative of that region. 



The small number of specimens from so large an area is, of course, 

 unfortunate, but that could not be helped. 



In considering the results of a study of this series, it is necessary 

 to bear in mind the manner in which it has been accumulated. It 

 is not an entirely "unprejudiced" series of specimens, i.e., it does 

 not include in every instance all the specimens that were collected at 

 the locality represented. The collection in this particular may be 

 divided into three classes: 



1. A very considerable number of sets, mostly small in number of 

 individuals, which are all the specimens that were collected at the 

 particular place represented. 



2. Sets received from correspondents, which were, no doubt, selected 

 from a more or less extensive series of duplicates. 



3. Sets collected by myself and selected from a larger number of 

 specimens. In such cases it has always been my intention to preserve 

 a representative series, consisting of the extreme forms and a fair 

 representation of what appeared at the time to be average specimens. 

 But, of course, the proportion of extreme specimens would be largely 

 in excess of what it would be had the whole series been retained. 



In view, therefore, of the very large proportion of selected speci- 

 mens included in the series, it would seem likely that the series as a 

 whole would give results, especially as to size, that would be in excess 

 of that obtained from an entirely unprejudiced series. 



II. One hundred and twenty-four specimens from Isle Royale, 

 Lake Superior, Mich., collected by the University of Michigan Expe- 

 dition in 1906. This series includes all the specimens collected on 

 the island. The number of specimens was so much larger than the 

 series from any other Michigan locality that it was deemed best not 

 to incorporate it with the general Michigan series, on account of the 



