FOSSIL MAMMALIA. 59 



with the base of the cranial cavity — each and all attest the close affinities of 

 the present animal to the Edentata. More decisive evidence of the same 

 relationship will be adduced from the organization of other parts of the cranium. 

 The glenoid articular surface (a, fig. 2, PI. XVI.) is an almost flattened plane, 

 wider in the transverse than in the longitudinal direction ; and, as in the genera 

 Myrmecophaga and Manis, it is not defended behind by any descending process. 

 In its general form it resembles the glenoid cavity of Orycteropus more than that 

 of the preceding Edentates ; but, in Orycteropus, the articulation is defended 

 posteriorly by a descending process of the zygoma, and it is also situated rela- 

 tively closer to the os tympanicum. 



Had the Glossotherium teeth ? The extent of the temporal muscle, which is 

 indicated by the rugged surface of the temporal fossa, and by the well-marked 

 boundary, formed by a slightly elevated bony ridge, which extends to near the 

 line of the sagittal suture, together with the size of the zygomatic portion of the 

 temporal bone, and the remains of the oblique suture by which it was articulated to 

 the malar bone, enables me to answer this question confidently in the affirmative. 

 They will probably be found to be molar teeth of a simple structure, as in the Oryc- 

 teropus. 



The evidence just alluded to of the existence of an os malse is interesting, 

 because this bone is wanting in the Pangolins ; and its rudimental representative 

 in the true Ant-eaters does not reach the zygomatic process of the temporal bone, 

 which consequently has no articular or sutural surface at its anterior extremity. 

 In the presence, therefore, of the surface for the junction of the os mala;, and the 

 consequent evidence of the completion of the zygomatic arch, we learn that the 

 Glossothere was more nearly allied to the Armadillos and Orycterope. That its 

 affinity to the latter genus was closer than to the Armadillos we have most 

 interesting evidence in the form and loose condition of the tympanic bone : it is 

 represented of the natural size at fig. 4, PI. XVI. Through the care and attention 

 devoted to his specimens by their gifted discoverer, this bone was preserved in 

 situ, as represented at d, fig. 1 ; but it had no osseous connection with the 

 petrous or other elements of the temporal bone, and could be displaced and 

 replaced with the same ease as in the Orycterope. This bony frame of the mem- 

 brana tympani, in the Glossothere, describes rather more than a semicircle, having 

 the horns directed upwards ; it has a groove, one line in breadth, along its concave 

 margin, for the attachment of the ear-drum, and sends down a rugged process, half 

 an inch long, from its lower margin. In the Dasypodes and Myrmecophagce, the 

 tympanic bone soon becomes anchylosed with the other parts of the temporal ; it 

 is only in Orycteropus, among the existing insectivorous Bruta or Edentata, that it 

 manifests throughout life the fcetal condition of a distinct bony hoop, deficient at 

 the upper part. The os tympanicum of Orycteropus, however, differs from that of 



