528 DR. C, W. ANDREWS ON THE 



structure of this region of the skull may appear, it represents 

 nearly the condition in the living animal. In other Proboscidea 

 the occiput slopes strongly forwards, but this inclination is 

 masked, in the adult at least, by the enormous development of 

 cellular bone that ta]<es place in this region, except over the area 

 of the insertion of the ligamentum nuchse, which occvipies in 

 consequence a deep depression. 



In young individuals in which this development of the bone 

 has not taken place, it can be seen that the occiput is inclined to 

 the basis cranii at an angle about equal to that occurring in 

 Dinotheriu'm. Thus, in a figure of a median longitudinal section 

 of the skull of a young African Elephant given by Flower 

 (' Osteology of the Mammalia,' ed. 2, p. 181, fig. 59), the somewhat 

 convex occipital surface is inclined to the basis cranii at an angle 

 of about 40 degrees — that is, at a more acute angle than in Dino- 

 therium. In the skulls of two very young Indian Elephants this 

 angle was about 50 degrees, and in Palaeomastodon, in which in the 

 adult the spongy bone is relatively little developed, the angle is 

 only about 35-40 degi-ees in the middle line. Thus it appears 

 that the forw^ard slope of the occiput relatively to the basis cranii 

 is not really greater in Dinotherium than in the later Proboscidea, 

 but only appears so because, when the ])alate and tooth series in 

 Dinotherium is placed horizontally, the basis cranii slopes steeply 

 downwards instead of being nearly horizontal. 



In Dinotherium no development of cellular bone seems to have 

 taken place, the necessary surface for the attachment of the 

 muscles supporting the heavy head being in this case supplied by 

 the great widening of the occipital surface, which is cari-ied out 

 laterally on to the lateral extensions of the squamosals (o.s.). The 

 pit for the nuchal ligament {n.l.) is both wide and deep : behind 

 it the occiput is convex from before backwards, while at the sides 

 it is gently convex in all directions. Anteriorly the occiput is 

 bounded by the transverse poition of the lambdoidal crest {I.e.), 

 which is slightly convex forwards : in front of this the parieto- 

 frontal surface makes an angle of about 140 degrees with the 

 occiput. Laterally the lambdoidal crest reaches the upper border 

 of the temporal fossa at about its middle point, and thence is 

 continued downwards and backwards along its border to the 

 lateral processes of the squamosal. The lateral portion of the 

 occiput thus bounded, projects far over the enormous temporal 

 fossa, forming a kind of roof to its posterior portion. The fronto- 

 parietal region of the roof is short from before backwards and in 

 its middle portion is flat or slightly concave. It first narrows a 

 little, and then widens out and slopes down towards the post- 

 orbital (p.orb.) processes, which are presumably borne by the 

 frontals though no sutures are visible. A little in front of the 

 level of the post-orbital processes the fronto-parietal region of the 

 roof is bounded by a thickened border, which is concave forAvards 

 and forms the upper edge of the remarkably wide nasal opening 

 (ex.n.). There is no trace of the nasal bones ; they may have been 



