Allen — TJie Water Shreio of Nova Scotia. 17 



of whicli is from Profile Lake, Franconia Mountains, New Hampshire, is 

 very sharply distinguished from palustris and acadicus by its much 

 blacker dorsal surface which is hardly at all contrasted with the dark, 

 smoky brown belly, save at the chin, which is white. It is further dis- 

 tinguished by its tail, which is typically black both above and below. 

 Nevertheless intergradation takes place in southern Canada, so that it must 

 rank as a subspecies of palustris. Thus a specimen taken by A. E. and 

 O. Bangs at Lake Edward, Quebec, is as silvery on the belly as in acadicus, 

 but the chest and throat are dark brown ; moreover the tail is entirely 

 dark except at the base ventrally, where for about a centimeter it is pure 

 white. A second specimen from North Bay, Ontario, is quite like albiharbis 

 in having a dark belly and chest, and white chin, but the basal half of 

 the tail is white ventrally. These specimens are therefore intermediate 

 between palustris, acadicus and albiharbis. The continental forms of the 

 typical subgenus, so far as now recognized, should stand as follows: 



Neosorex palustris palustris (Richardson), from Hudson Bay west on 

 the plains to the foothills of the Rocky Mountains. 



Neosorex palustris navigator Baird, the Rocky Mountains from British 

 Columbia to Colorado, and the Sierra Nevada of California, a slightly 

 smaller race. 



Neosorex palustris alaskanus (Merriam), known from southern Alaska 

 (Point Gustavus, Glacier Bay); yet smaller with a shorter and more 

 massive skull. 



Neosorex palustris acadicus subsp. nov. , from Nova Scotia westward, 

 intergrading in southern Canada with typical palustris and the following. 



Neosorex palustris albibarbis Cope, from the Adirondacks of New York 

 and the White Mountains of New Hampshire southward along the Alle- 

 ghanies; a darker race, with dark belly and tail, and white chin. 



The problematical Neosorex hydrodromus (Dobson) described from 

 Unalaska Island, Alaska, remains unknown except from the original 

 diagnosis. 



