﻿28 



NORTH AMERICAN FAUNA. 



[No. 12. 



the level of tlie pterygoids. Tbe median ridge just here widens 

 abruptly and sends out on each side a short i^rocess, which is met by a 

 similar one arising from the palatine on the opposite side of the groove 

 (fig. 7 A, /. h)\). These processes nsnally meet and fuse, thus com- 

 pletely obliterating the groove, though they are frequently separated 

 by a narrow space. In Fvotomys (fig. 7 B) the sloi)ing part of the 

 median ridge has disappeared, together with the lateral pits, but 

 traces of the median ridge (fig. 7 B, m. r.), the lateral grooves (fig. 

 7 B, J. gr.), and the bridges (fig. 7 B, ?. hr.) may still be recognized. 



At different times subgeneric weight has been given to the form of 

 the external ear, and to the proportional length of the tail to the head 

 and body. Neither one, however, is of auy value, except in special, 

 isolated cases. The form of the ear is essentially the same in all the 

 subgenera, though there are slight modifications in length and in the 

 development of the valvular fold by which the meatus is closed. 



The relative length of the tail is far too variable to serve as a useful 

 diagnostic character. 



KEYS. 



The following keys to the genera and subgenera of Microtinw are 

 wholly artificial and do not bring the groups together according to 

 natural affinities. Since analytical keys are of no value except as aids 

 in identifying specimens, it is necessary that they should be based on 

 characters that may be studied without difficulty in ordinary museum 

 material. Such material, however, is usually so imperfect that a single 

 key made with reference to one set of characters (as, for instance, the 

 form of the bony palate or the number of mamm?e) might be of little 

 use. Hence several keys are here introduced, each based primarily on 

 a special set of structures. Of the three keys to the genera, ISTo. 1 is 

 made, so far as xjossible, with reference to the skull alone; l^o. 2, with 

 reference to the teeth, and l^o. 3, with reference to external characters. 

 Of the keys to the subgenera of Microtus^ No. 5 is based primarily on 

 characters derived from the structure of the bony i^alate, and is thus 

 useless for the identification of si^ecimens the skulls of which are not 

 available for study. Key No. 6 is based on the pattern of enamel fold- 

 lug and may be used with specimens having broken skulls. The lines 

 in italics inserted in parentheses in this key are for the identification of 

 individuals with abnormal enamel patterns. These usually occur in the 

 proportion of about one to four (see p. 25). Hence, one fourth of any 

 given lot of specimens will agree with the characters given in jDaren- 

 theses', the great majority, however, with those in heavy type. Key 

 No. 7, based i^rimarily on the mammie and footpads, is made almost 

 exclusively with reference to external characters. It is necessarily 

 incomplete, since the number of mamma? and footpads is in several 

 instances unknown. It is, of course, impossible to use this key except 

 with alcoholic specimens or freshly killed animals. Key No. 8 — if it 



