laterally compressed, two rooted, teeth (« and b, fig. 3), the anterior o£ 

 which (fig. 3, a) is separated from the tusk by an interval of 21 mil- 

 limetres, while between the two is a spice of 1 1 millimetres. 



The outer surface of the bone is convex along the region occupied 

 by the socket of the tusk and terminates above in a thickened rounded 

 border which towards the hinder part of the specimen shows traces of 

 having been overlapped on its inner side by another bone (? the nasal). 

 .Towards the alveolar border behind the tusk the face of the bone is 

 concave from above downwards so that in this region it is very thin. 



Oil the inner surface opposite the alveolus of the tusk and extending 

 a little behind it, is a thickened shelf-like projection with a truncated 

 cud (fig. 3, s), which clearly united in suture with its fellow of the 

 opposite side. This formed part of the palate and seems to be the 

 palatal portion of the premaxilla: it terminates behind in a thickened 

 rounded border which probably formed the boundary of a large anterior 

 palatine foramen. The total length of this palatine process is about 

 28 millimetres and its greatest depth about 10 millimetres. Further 

 back the inner face of the bone bears alonu; its lower edge a second 

 roughened surface for union with another element. 



The interpretation of this specimen is somewhat difficult. Probably 

 the tusk is an enlarged front incisor, but in that case the nature of the 

 two smaller teeth is doubtful; it does not seem probable that they 

 represent the other incisors, but, on the other hand, if they are 

 premolars all traces of the maxillo-palatine suture seem to have been 

 obliterated, and in this region at least, the maxilla bears no palatine 

 plate. 



More complete remains of Phiomia serridens must be obtained before 

 its systematic position can be determined, but it seems probable that 

 this remarkable animal was a specialised Creodont. 



SAGHATHER1UM ANTIQUUM, Gen. et sp. nov. 



(Figure 4). 



The type of this species is a very imperfect skull, of which a portion 

 of the roof and the maxilla' with the cheek-teeth are alone preserved. 



In the parietal region there is a fairly well-developed sagittal crest, 

 which bifurcates just behind the orbits, the branches diverging at a 

 very obtuse angle and terminating in the supra-orbital processes, which 

 seem to have been very prominent and overhanging. The interorbital 

 region of the skull is very broad and flat. 



