PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 65 



The enamel pattern of the lower cheek teeth is markedly similar in 

 the two forms. It is important to note that the anterior triangles in 

 the first lower cheek tooth of the type are confluent as in Pitymys, not 

 as in Cope's illustration. This error in the drawing tends to exagger- 

 ate the appearance of the anterior loop, which in reality is not greatly 

 different from that in some of our specimens. Cope noted the re- 

 semblance between his form and P. pinetorum, but apparently the 

 Pleistocene form is distinct from the living species in exhibiting a 

 slightly simpler enamel pattern. 



The teeth in Microtusi?) speothen (Cope, 1871, pp. 87-88, fig. 13) 

 and M.(?) dideltus (Cope, 1871, p. 89, fig. 15) are distinctly larger 

 than in the Cumberland Cave form. In M.(?) speothen the enamel 

 pattern in the first lower cheek tooth is of a different type than that 

 in Pitymys. This tooth in M.(?) dideltus, while showing confluent 

 anterior triangles as in Pitymys, has an anterior loop that is narrow 

 and elongate, and the portion of the tooth posterior to the loop is 

 relatively wide. 



As noted by G. S. Miller, Jr. (1896, p. 59), the anterior lower cheek 

 tooth of Pitymys is not different from that in Pedomys, or Microtus 

 ochrogaster, living in the Mississippi Valley. Hence, reference of the 

 Cumberland fossil to the genus Pitymys in the absence of satisfactory 

 skull material cannot be made with certainty. 



ONDATEA cf. ANNECTENS (Brown) »' 



Figure 36 



A very small species of Ondatra is represented in the Cumberland 

 Cave material by a broken mandibular ramus (fig. 36), U.S.N.M. 

 no. 12044, containing all the teeth. In size this specimen agrees closely 

 with the type of the Conard Fissure species, 0. annedens, being only 

 slightly smaller and possessing a somewhat narrower incisor. It 

 agrees with 0. annedens in most other important details observed and 

 is therefore provisionally referred to this species. 



It was suggested by Hollister (1911, pp. 33-34) that Brown had 

 wrongly interpreted some of the characters described. Brown con- 

 sidered the Conard Fissure species as "an intermediate type connecting 

 Fiber (=Ondatra) with Microtus through the subgenus Neojiber." 

 The type of 0. annedens as observed by Hollister indicates "a species 

 close to existing forms in everything except size", and he stated that 

 the "anterior loop of first lower molar is more deeply cut by the 

 reentrant angles than in any specimen of an existing species examined." 

 If these anterior reentrants are to be considered within the loop, as 

 implied by Hollister, then the number of reentrants in Mi posterior 

 to the loop on each side appears to be one less than in 0. zibethica, 

 approaching the condition seen in Neojiber alleni. In either case the 



" Brown, 1908, p. 197. 



