2l8 NEW GENUS AND SPECIES OF THE SORECID.-E. 



relationships of Bendire's Shrew, it must be compared with typical 

 representatives of the genera to which it is most closely allied, and 

 these are found to be Sorcx and Neosorex~^\\.'~, affinities with 

 Blarina being very remote. But, as preliminary to this inquiry, it 

 becomes expedient to indicate the differences existing between the 

 two first-mentioned genera. 



The only tangible diagnostic characters that have been assigned 

 to the genus Ncosorcx are the long fimbriated feet, the-great length 

 of the tail, and the circumstance that the known species are " sharply 

 bicolor, blackish above and whitish beneath." No distinctive 

 cranial or dental characters having been pointed out,* I have in- 

 stituted a comparison between the type specimen of Neoso7'ex 

 navigator f and a specimen of Soi^ex Coopcri. The differences 

 noted are as follows : In Neosorex the rostrum is longer, and its 

 sides meet the cranium at a decided angle ; while in Sorex the sides 

 of the rostrum are but little out of line with the cranium. In 

 Neosorex the rostrum is more sharply compressed just in front of 

 the molariform series, so that its anterior portion is more attenuate, 

 and the unicuspids more nearly parallel than in Sorex. The 

 greater development of the facial portion of the skull is best 

 shown in the lengthening of the upper jaw and the shortening of the 

 floor of the cranium. In Sorex, the ratio to the entire length of the 

 skull of the distance from the front incisor to the hinder margin of 

 the palate is 43.2 ; in Neosorex, 46.3. On the other hand, the ratio 

 to the entire length of skull of the distance from the hinder margin 



*I am aware that Coues, in his "Precursory Notes on American Insectivorous Mammals" 

 (Bull. U. S. Geol. and Geog. Survey, Vol. Ill, No. 3, 1877, pp. 631-653), says that in A'eosorex 

 the posterior hook of the upper incisor is "as large as the succeeding tooth " (p. 641). But Baird, 

 in his original description of the genus, expressly states that the basal hook only equals the fourth 

 unicuspid, which is considerably smaller than the first. Having the type specimen of N. navigator 

 before me, and examining it with special reference to this point, I find that the basal hook is but 

 little more than half — certainly not two-thirds — as large as the first unicuspid. 



f I am indebted to the courtesy of Professor S. F. Baird for the privilege of examining the type of 

 Neosorex tiavigator (No. '^^^^), together with several other representatives of the Neosorex group, 

 belonging to the United States National Museum. My thanks are also due Captain Chas. E. 

 Bendire and Dr. Elliott Coues for their kindness in selecting and transmitting the specimens. 



