Vol. V] TAYLOR— NEW SUBGENUS OF PHENACOMYS 125 



IV, opposite page 44, the first inner triangle is shown opening 

 broadly into the outer loop of the anterior trefoil as in longi- 

 caudiis. In this specimen also the first inner triangle is shown 

 to open broadly into the first outer triangle. On plate III, 

 facing page 42, the first inner triangle in P. ungava is shown 

 opening narrowly into the first outer triangle. Miller (1896, 

 p. 41) has figured the enamel pattern of celatiis. According 

 to his illustration the first inner triangle opens narrowly both 

 forward and back. These are the only instances I have noted 

 in the literature in which conditions are found similar to those 

 in longicaudus. 



It seems safe to conclude that longicaudus is characterized, 

 with qualifications as above noted, by a more open condition of 

 the loops and triangles of this tooth than in the subgenus 

 Phenacomys. 



(4) Second lower molar 



Similar in pattern to Phenacomys intermedins, the type of 

 the genus Phenacomys. Small antero-external loop, elongated 

 antero-internal triangular digitation, one short outer triangle, 

 one long inner triangle, and a posterior transverse loop. 



The small antero-external loop is never closed in longi- 

 caudus. In one specimen of intermedins (No. 174425) the 

 loop is larger than in any specimen of longicaudus, with the 

 possible exception of No. 21143, opening broadly into the 

 opposing triangle; in another specimen of intermedins (No. 

 174431) the antero-external loop is only narrowly open; while 

 in a third (No. 174432) the loop is large and opens broadly. 

 All specimens of orophilus at hand, with the exception of Nos. 

 205916 and 67327, have this loop tightly closed off from the 

 opposing triangle. In Merriam's figure 7, illustrative of the 

 genus Phenacomys (1889, p. 31) the loop is tightly closed. 

 In his plate III, facing page 42, the antero-external loop both 

 in celatus and ungava is broadly open as in longicaudus. Plate 

 IV, opposite page 44, shows the antero-external loop opening 

 broadly in latimanus as well as in celatus and narrowly in 

 intermedius. In Miller's figure of celatus (1896, p. 41) the 

 same relations are shown. 



