5 24 NEW GENUS AND SPECIES OF THE SORECID.^. 



of the malar process of the maxillary bone to the glenoid fossa. 

 The anteorbital foramen is very large. The rostrum is strongly 

 compressed in front of the molariform teeth, its sides rising almost 

 vertically to form the lateral walls of the anterior portion of the 

 nasal chamber. The width opposite the third unicuspid is 2.20 

 mm., while the distance across the molars is 6 mm. — or 2.72 times 

 greater. The pterygoid hamuli are slender styliform processes 

 extending almost directly backward. 



The mandible presents no marked peculiarities. The horizontal 

 ramus is stout and slightly convex. The coronoids are large and 

 high, diverge outward, and incline a little forward. The styliform 

 angular processes are relatively short for a Shrew ; they are 

 parallel, and are directed backward, and slightly downward. 



Nothing- whatever is known of the habits of Bendire's Shrew. 

 Its structure, however, leads to the belief that it is not amphibious 

 like Akosorex, nor an inhabitant of dry uplands, like the various 

 species of Sorex proper, but that it is a marsh species, dwelling in 

 wet meadows, and occasionally taking the water either in the pur- 

 suit of its prey or as a means of escape from its enemies. 



In view of its isolated position, speculations concerning its 

 genetic affinities are perhaps unwarrantable ; still, a somewhat 

 critical study of its peculiarities shows that, whatever its past 

 history, it cannot be regarded as intermediate between Sorex and 

 Neosorex. Neither can it have been derived from Neosorex. 

 Hence the logical inference is that Atophyrax, in common with 

 Neosorex, was early differentiated from a group of thirty-two- 

 toothed Shrews of which the genus Sorex contains the nearest 

 living allies. Having abandoned a fossorial for, in the one case 

 a natatory, in the other a paludal habit, Akosorex and Atoph- 

 yrax doubtless began to diverge in the same direction, their dis- 

 tinctive features having been developed and intensified as their pe- 

 culiarities of habit becanie fixed — each retaining in different de- 

 grees of modification certain characteristics of the original stock. 



For the very accurate drawings accompanying this paper I am 

 indebted to the kindness of Mr. Ernest E. T. Seton, who executed 

 them, under my supervision, from the alcoholic specimen. 



