1897.] MAL.VGAST GEWtrS BBACHXITROMYS . 705 



In the Antliracotherina, e.g. in Ancodus \ and in the Anci/lopoda ^, 

 we find three widely bulging cusps in the upper molars. In 

 Hyracotlieria ' it is easy to see that the two posterior of the three 

 outer cusps have nothing whatever to do with the two posterior 

 of Ancodus and the Ancylopoda ; only their anterior one, which I 

 consider as homologous with cusp 1 of Insectivora and Diddphys, 

 is the homologiie of the anterior cusp in Hyracotlieria. The two 

 posterior cusps of Ancodus and the Ancylopoda are 2 and 3 ; those 

 of the Hyracotlieria are the homologues of the two cusps more 

 internally situated in the former, Winge's 4 and 5, Osborn's 

 paracone and metaeone : the comparison with other Hyracotlieria *, 

 in which something more of these " -styles " is preserved, shows 

 this at once to be the case. The posterior outer cusp in upper 

 molars, "Winge's 3, Osborn's metastyle, which is so well developed 

 in Diddpliys and in many Insectivora \ is, when met with at all in 

 other orders, generally the least developed of the three outer cusps, 

 and there is sometimes a relation in its development with that of 

 the antero-internal cusp in lower molars, Winge's 1, Osborn's 

 jjaraconid. 



The fact, pointed out by AVinge, that these cusps of Insectivora 

 and Didelphyinse have their homologues in other orders, chiefly 

 in the older members, is of the greatest importance and a 

 notable progress in our knowledge of the homologies of the 

 Mammahan molar. It is a very remarkable fact, that this 

 outermost series of cusps is enormously developed in several upper 

 molars from the Laramie Cretaceous ^. Whether we have to con- 

 sider them, with Winge, as being the most primitive parts of the 

 molar, older than the cusps situated internally from them in upper, 

 externally in lower molai's, is quite another question, in which it 

 is not proposed to enter particularly for the present. 



As regards the Eodents, Winge points out the remnants of this 



1 Cf., e.g., Zittel, ' Handbuch d. Palaontologie,' iv. p. 329, fig. 260. 



2 Zittel, /. c. p. 314, fig. 255. 



^ J. L. Wortmann, " Species of Hyracotheriiim and allied Perissodactyles 

 from the Wabsatch and Wind Biver Beds of Nortb America," Am. Mus. Nat. 

 Hist. vol. Yiii., vi. 1896, p. 89, fig. 3, p. 95, fig. 4 ; Zittel, /. c. p. 277, fig. 219. 



•* Zittel, I. c. p. 243, fig. 179, p. 242, fig. 171 {Pachynolophus siderolithicus) ; 

 Wortmann, I. c. p. 108, fig. 18 (Oreohippiis). 



5 Winge, I. c. pi. iii. fig. \b,2h,^b. 



^ Cf. H. F. Osborn, " Fossil Mammals of the Upper Cretaceous Beds," 

 Am. Mus. Nat. Hist. vol. v. Art. xvii. pi. viii., New York, 1893. Osborn calls 

 the molars referred to " trituberculates," and expresses his belief that they 

 lend " overwhelming proof, if any more were needed, of the unity of origin of the 

 molar types of the higher Mammalia from a tritubercular stem instead of from 

 a multitubercular, as Forsyth Major has suggested "' {l. c. p. 320). For me 

 these Laramie " Trituberculates" so-called, are polybunous, as well as those 

 figured on pi. vii. ; and I have only to repeat once more that " trituberculate " 

 and " triangular" are not sj'nonymous. Prof. Osborn assures us {I. c. p. 320) 

 that these "Laramie Trituberculates" "include a variety of forms just 

 emerging from the primitive tritubercular stage." That is precisely what has 

 . vet to be proved. But even if we were inclined to take Prof. Osborn's assertion 

 for granted, it would have to be shown how it happens that several of 

 these ancient forms (E, F, pi. viii.) have in their very process of" emerging" 

 already acquired such a luxury of "-styles" and " -conules," as to constitute 

 by themselves alone half of the molar's crown. 



