SCIENCE. 



[Vol. XX. No 491 



with the anterior crest already described. As in all the 

 early Eocene-Tertiary Mammalia, the pumolars of both the 

 lower and upper series are much simpler than the true mo- 

 lars. The structure of the lower molars of Coryphodon is 

 interesting-, as it represents a stage in the modification of a 

 more primitive type, which had the enamel arranged in the 

 form of two symmetrical Vs or crescents. Now in Cory- 

 phodon the anterior limb of each crescent is nearly reduced; 

 this applies especially to the posterior V. The portion of 

 the tooth bearing the anterior V is raised high above the 

 posterior or heel part. 



The variation in size of the teeth of the different species 

 of Coryphodon is very great, and in not a single instance 

 have I been able to find two individuals, of the same species, 

 whose teeth are of the same size. This variation is shown 

 in the form of the canines and incisor teeth; in the former 

 the difference in size is largely due to age and sex. 



The last upper molar undergoes a great amount of varia- 

 tion, it varying from the nearly quadrate form to that of an 

 elongated oval, the latter form occurring in the more modi- 

 fied species. The modification of the elements of the crown 

 of the second superior molar is interesting, as we can trace in 

 this transformation a true phyletic series, from the less spe- 

 cialized to the more modified species. The typical forms of 

 Coryphodon have the external crescent of this tooth well 

 developed. The first step towards reduction of the crescent 

 occurs where the intermediate portion of the posterior limb 

 (see upper Fig. p) disappears, leaving an external isolated 

 cusp (C. testis). This condition is found permanent on the 

 last superior molar of Ectacodon, the latter genus not hav- 

 ing advanced so far in its dental evolution as Coryphodon. 

 The species C elephantopus represents an intermediate stage 

 in its dental evolution between that of Coryphodon testis 

 and Ectacodon. 



Professor Cope established the genus Metalophodon upon 

 the character of the crescent of the second superior molar, 

 and in this genus the posterior limb of the crescent is nearly 

 reduced. As all stages exist in which this crescent is well 

 developed down to that where it is wanting, I can not accept 

 Metalophodon as a good genus, and believe it should be con- 

 sidered a synonym of Coryphodon. The most modified con- 

 dition of this crescent is where it is reduced to merely the 

 anterior limb. The latter stage is permanent in the last 

 upper molar of all the known species of Coryphodon; but 

 it is interesting to note that in a genus described by Cope, 

 called Manteodon, the last upper molar has a perfectly 

 formed external crescent. 



The genus Manteodon differs from all other genera of 

 the Coryphodontidce from the fact that the last upper molar 

 has two well-developed internal cones. Now in all other 

 forms of this family the postero-internal cone {hypocone) is 

 wanting, although traces of it occur in C elephantopus. 



It is not without considerable difficulty that the homolo- 

 gies of some of the elements of the upper molars of Corypho- 

 don are determined. The form of molar from which the 

 Coryphodon type of tooth has probably arisen, occurs in 

 the genus Pantolambda, which is from the Puerco or lowest 

 Eocene beds of New Mexico. In Pantolambda both the 

 external crescents of the superior molars are well devel- 

 oped, and the internal cone has two crests running out 

 from it. Now what are the homologies of the anterior 

 portion of tbe Coryphodon molar as compared with that 

 of Pantolambda. The postero-external crescent is equally 

 well developed in both forms, but what has become of 

 Ine anterior crescent in Coryphodon, which is so strongly 



developed in Pantolambda ? The prominent cusp (see 

 Fig. e.m.) on the external face of all the superior molars 

 of Coryphodon probably represents the reduced anterior 

 crescent of Pantolambda. This is the homology advanced 

 by Professor Cope. The anterior crest of Coryphodon has 

 arisen by the development of the crest running outwards 

 from the internal cone of Pantolambda. Thus it is by 

 studying the earlier or more primitive types of many of the 

 Mammalian phyla that we are enabled to interpret those 

 marvellous changes which different parts of the dental and 

 slseletal structures have undergone. 



The structure of the last lower molar displays considerable 

 variation ; this affects particularly the elements of the heel (see 

 lower Fig. h.). In the more primitive species the two cusps 

 forming the heel are in a straight line, whereas in other vari- 

 eties a small cusp may arise in the posterior valley of the heel, 

 internal to the postero-internal cusps (e, n, a). The growth of 

 this rudimentary cusp causes the pushing outwards of the 



A superior and inferior molar of a typical species of Coryphodon (C. radians), 

 a. e. c, antero- external cone ; a. c, anterior crest; i. c, internal cone; e. m., 

 external median cusp; e. c, postero-external crescent; a., anterior limt) of 

 crescent ; p., posterior limb ; 7i., heel of lower molar ; hy.d., external cone of 

 heel ; eji.d., Internal cone, 



internal of the two primitive cusps forming the heel; further 

 growth causes the primitive internal cusp to occupy a 

 median position, and it now fulfils the function of a fifth 

 lobe of some of the other Ungulates. This posteromedian 

 cusp is merely an analogical structure, and its development 

 proves that it is not homologous with the fifth lobe of the 

 Lophiodonts, 



The skeletal variations are many in this group, they affect 

 principally the length and heaviness of the limb bones, and 

 also the size of their articular extremities may vary a great 

 deal. 



The variations of the astragalus are particularly interest- 

 ing, as upon them in some cases new genera have been es- 

 tablished, A very primitive structure occurs in the tarsus 

 of Coryphodon, as in all the other genera of the Amblypoda; 

 that is, on the inner side of the astragalus, a separate bone, 

 or rather a facet for this bone to articulate with, is present, 

 Tbe bone articulating with this facet is generally called the 

 tibiale or internal navicular, Baur' has shown that the 



1 American Naturalist, January, 1885, p. 87, 



