PURBECK FORMATIONS. 43 



concave. It plainly indicates a produced and slender fore-part of the head, recalling the 

 physiognomy of the insectivorous marsupial Bandicoots [Perameles] and JMyrmecobians. 

 This mandibular character suggested the ' nomen triviale ' proposed for the present 

 species. 



Of the incisors (PI. II, fig. 11 a, n, 2, 3) it can only be said that their crowns have been 

 slender, and were probably long ; the first or foremost procumbent, the second and third 

 progressively, though slightly, rising from the horizontal position. The canine (c) is 

 rather more raised, and shows its characteristic curvature to a sharp summit. A slight 

 interval separates each of tliese front teeth from the other. At a similar interval from the 

 canine rises the compressed sharply conical crown of a small premolar {p 1). The next 

 [p 2) has been forced forward and turned round, the curved hind slope of the cone being 

 pushed behind the base of j» i : the divided socket of this two-rooted tooth is exposed. 

 The third premolar is in place, and well shows the shape of the crown, with the convex 

 front border and concave hind one, produced at the base without any cusp-like elevation 

 of this part. The socket of the fourth premolar is vacant. 



The fifth tooth {jj 5) shows a change of shape, approaching that of the first of the retained 

 molars in figs. 10 and 10 a. The fore and hind borders of the compressed cone slope or 

 diverge from the sharp apex at almost the same angle, and each end of the base developes a 

 minute cusp or talon. The body of the cone is markedly convex antero-posteriorly, as in the 

 hinder molars. There is no indication, impressional or alveolar, of any premolars between the 

 last {p 6, if it be a premolar ?) in place and the molars in situ (in fig. 10 a), suggesting 

 the character shown in both upper and lower jaws of Peralestes. Otherwise it might be 

 questioned whether the type of upper molars in Peralestes (PI. II, figs. 3 and 3 a) might 

 not probably be that of Peramus. But, even if Peramus had shown the marked superiority 

 of height and size of the last two premolars over the succeeding true molars, the present 

 must have been a different and smaller species from the Peralestes longirostris represented 

 by figs. 8 and 4 in PL II. 



Peramus tenuirostris (?). Plate II, figs. 12, 12 a. 



This is represented by the anterior portion of a left mandibular ramus, with the outer 

 surface exposed, and the symphysial end with its alveoli and teeth broken away : it also 

 shows the premolars, which have suffered some displacement in the foregoing specimen. 

 They accord in proportions and shape too closely with the dental characters of Peramm 

 to justify generic separation. 



The first tooth in place is a premolar, two-rooted, with a simple subcompressed conical 

 crown. The next tooth seems to show a lower conical crown, with the same fore-and-aft 

 basal extent ; but there is evident trace of mutilation by fracture. The third tooth gains 

 in height and fore-and-aft extent. The fourth is similar in size to the third. The fifth 



