No.2.] COPE ON MIOCENE RODENTIA. SG9 



from the Castoridcc on various grounds. I do not think any of his char- 

 acters are tenable, excepting that drawn from the form of the mandible, 

 which is expressed thus in Mr. Alston's diagnosis: "angular jDortion of 

 mandible much twisted." This character will be better described as 

 follows : Angle of mandible with a transverse edge due to inflection on 

 the one hand, and production into an apex externally ; the inflection 

 bounding a large interno-posterior fossa. 



Mr. Alston enumerates four genera of Castoridce — Castor, DiobroticuSj 

 Steneofiher, and Castoroides. J. A. Allen has shown that the last-named 

 genus cannot be referred to this family. The characters of JDiobroticus, 

 as given by Alston, are as follows : " Skull much as in Castor. Third 

 upper molar and lower premolar elongate, with four enamel folds, the 

 rest with only two ; all the folds soon isolated." This diagnosis appears 

 to separate the genus satisfactorily. The definition of Steneojiber is as 

 follows: " Parietal s not parallelogrammic ; interparietal subhexagonal ; 

 basioccipital not concave ; grinding teeth as in Castor, the subsidiary 

 folds sooner isolated." The distinction from Castor here rests exclu- 

 sively on the forms of the parietal, interparietal, and basioccipital bones. 

 This kind of definition is always of questionable validity, as the terms 

 "parallelogrammic," "hexagonal," etc., are not intended to be exactly 

 used and cannot be exactly applied. The Castor [Steneojiber) penimu- 

 latus illustrates this fact, for there is no striking ditference in the 

 forms of the two bones to which these terms are applied, as compared 

 with the Castor fiber. The basioccipital bone differs from that of the" 

 beaver, but not so as to conform to J. W. Alston's diagnosis of the genus 

 Steneofiber. Its inferior surface is concave, but doubly so, as a keel 

 occupies the median line. In the S. viciacensis, according to Filhol, 

 this region is shallowly concave, without median keel. Although im- 

 portant as specific characters, these variations do not appear to me to 

 require the recognition of as many genera. The possession of the epi- 

 trochlear foramen in the 8. viciacensis is at first sight an important 

 character. Mr. Filhol, however, informs us that of thirty-four humeri 

 which he has studied, sixteen possess the foramen, and in eighteen it 

 is wanting. 



The Castor tortus was described by Leidy from the Loup Fork forma- 

 tion. He coined the subgeneric name Eucastor for it without corres- 

 sponding definition. In his monograph of the Castoridcc, J. A. Allen 

 referred this species* to a genus distinct from Castor, and defined it, 

 using lor it Leidy's name Eucastor. This genus appears to me to be 

 valid. The three genera of Castoridce will then be defined as follows ^ 



Molars and premolars with ouc inner and two or three outer folds Castor. 



" Inferior premolar and third superior molar elongate, with four enamel folds; the rest 



with only two. " Diobroticm. 



Superior premolar enlarged, with one inner fold; inferior molars small, with two 



lakes Eucastor. 



* Monographs of North American Rodeutia, Couesaud Alleu, U. S. GeoL Siirv..Tt*n-s., 

 Ib77, xi, p. 450. 

 24 G B 



