igoi.J Gidiey, The North American Species of Equus. I05. 



show all the intermediate stages as well as both extremes of this 

 character. A fact worthy of mention here is that in the Ameri- 

 can Museum material, the absence of a cup, or a very much in- 

 terrupted inner wall, of the external incisor, is the rule in the 

 species with a short muzzle, while in Equiis scotti^ which is 

 a long-nosed type, the entire absence of a cup is much less- 

 frequent. Figure 5 shows a series of lower external incisors (^, 

 B, C,) of E. caballits and another (Z), E, E) of E. scotii, illustrat- 

 ing the variability of this character. The only other character of 

 the lower teeth which Cope ' seemed to consider important is the 

 form of the groove or channel separating the two lobes of the 

 metaconid-metastylid column, Which he designates as (r) a 

 "sharp entrantangular groove," (2) an "open gutter," (3) a 

 " flat channel." An examination of a large number of lower 

 teeth of Equus and of the Loup Fork genera of horses soon con- 

 vinced the writer of the great variability and extreme unreliability 

 of this character in specific determination. In the Loup Fork 

 genera it seems to be the rule that when the separation of the 

 lobes of the metaconid-metastylid column is well marked, the 

 bottom of the groove is nearly always sharp, but when not well 

 marked then the groove may be said to be a flat channel. It 

 seems to be only a matter of individual variation as to which 

 character is presented. In all the species of the genus Equus 

 this separation is usually an open gutter. 



Summary. 



An application of the foregoing principles of the variability of 

 tooth characters to the specific definitions used by Owen, Leidy, 

 and Cope shows the unreliability of most of the characters they 

 have employed. Thus the degree of complexity of the enamel 

 folding is seen to be greatly affected by both age and individual 

 variability — a fact which must be correctly understood before 

 this character can be of even subordinate value as a specific 

 determinant; the other characters given by these authors have 

 been shown to be so affected by either age or individual variabil- 

 ity, or both, that no dependence whatever can be placed on any 

 of them; and it has been also shown that size, especially the 

 transverse diameter of the molars and premolars, although least 



^ Report Geol. Surv. Tex., i8g3, p. 66. 



