1910.] Arsino'ilherium. 365 



Proboscidea, Hyracoidea and Amblypoda to the Subungulate section of the 

 Ungidata and expresses the opinion that most of the similarities in the 

 skeleton between Arsinotflierium and Elephas and the Amblypoda are 

 probably parallel modifications due to great size and weight in the two 

 grou])s. 



The following observations based on a comparison of Dr. Andrew's 

 excellent figures with specimens of Hyrax seem to lend additional support 

 to his conclusions: 



(1) The upper premolars of Arsinoitherium are derivable (in accord- 

 ance with the analogy furnisherl by the Rhinoceroses) from the type repre- 

 sented in MrgahJiijra.v. The little worn uppei- })remolar of Arsiiw'ithrriimi 

 (l c, 1)1. V, fig. 2) differs from the corresponding premolar of Mcgalohi/ra.r 

 {I. c, pi. vi, fig. 2) chiefly in the hypsodonty antl flattening of the ectoloph, 

 development of the parastyle, anterior cingulum and postero-internal 

 cingulum cusp (tetartocone). In both types the metaconule is absent and 

 there is a prominent median valley just posterior to the protoloph. In fact 

 it may confidently be stated that in respect to the unworn premolars Arsiiiui- 

 theriuni offers a closer comparison with Megalohyrax than with any known 

 Amblypod, Condylarth, Perissodactyl or other ungulate. 



(2) The lower cheek teeth of Arsinoitherium. likewise apj^ear to be 

 hypsodont modifications of the tyi)e seen in Megalohyrax. I'he fourth 

 lower premolar (/. c, \)\. iv, fig. 3) retains evidence of the reduplicate meta- 

 conid (metastylid) which is so conspicuous an element in Meniscotherinvi 

 and Hyrax (but which appears also among the Perissodactyla). The un- 

 worn lower molars, however (/. c, pi. iv, fig. 1) suggest those of the AnAly- 

 pod Coryphodon. The third lower molar lacks the third lobe (hypoconulid) 

 as in Meniscotherium, Hyrax and other forms. In general, the dentition of 

 Arsinoltheriuvi differs from the contemporary Hyracoids chiefly in the 

 kind of features which in other cases distinguish highly specialized from 

 primitive forms. 



(3) The architecture of the young skull of Arsinoitherium (/. c, \A. iii) 

 differs from that of the Dinocerata (Marsh, 1884, pi. ii) in several characters, 

 apart from the more obviously adaptive differences; e. g., the lachrymal in 

 Arsinoitherium. is practically vestigial and apjiarently without a canal; 

 in Dinoceras it is very large and spreading and is pierced by the large lachry- 

 mal foramen. 



In underlying features Arsinoitherium exhibits a much closer resemblance 

 with Hyrax. The premaxillaries, lachrymals, nasals and frontals have 

 essentially similar relations and differ chiefly in proportions; the malar 

 extends back to the anterior edge of the glenoid fossa ; the dorsal ridge of the 

 zygoma is continued above the auricular meatus, {op. cit., pi. iii, figs. 1, 2) 



