456 PROFESSOR J. ANDERSON ON THE 
but are straight and directed forwards and downwards, the upper surface of the snout 
being flat and not convex transversely as in Zupaia. The nasals are slightly dilated in 
the middle, beyond which they contract posteriorly to a point, contracting also ante- 
riorly, but again feebly dilating at their anterior extremities. They reach back to nearly 
in a line with the anterior margin of the orbit; and their posterior fourth is invested 
externally by the frontals. The lateral concavity of the snout between the second pre- 
molar and lateral incisor of Tupaia is absent in Hylomys, the sides of its snout being 
almost straight. The maxille reach a long way posterior to the nasals; and in this the 
skull differs both from Tupaia and Erinaceus. The premaxille do not reach to the 
frontal, but their inferior margin is on a line with the first premolar. In Tupaia the 
skull from the parietal eminence slopes markedly downwards and backwards to the 
lambdoidal crest, the two sides of which meet in a point, giving a triangular outline to 
that portion of the skull; and there is a semicircular interparietal. In Hylomys there 
is only a slight depression between the vertex and the middle line of the lambdoidal 
crest, which is almost on a level with the vertex, and not considerably below it as in 
Tupaia. The lambdoidal crest is much truncated above, forming an obtuse angle with 
its sides. This portion of the cranium, therefore, of Hylomys is much broader than in 
Tupaia; and the occipital region is slightly fuller, and, instead of sloping downwards 
and forwards as in the latter genus, it is directed downwards and backwards, the 
foramen ovale looking more backwards than downwards. In the character of the 
hinder part of the cranium, Hylomys approaches more to Erinaceus’. ‘The space 
between the orbits is nearly flat, and is not so broad as in 7wpaia; neither is the orbit 
enclosed by bone behind, nor the zygomatic arch perforated; and in this it resembles 
Erinaceus. The malar is a small spicule of bone applied to the outside of the perfect 
zygomatic arch. It also approaches Erinaceus rather than Tupaia in its large infra- 
orbital foramen and shorter canal, and in the prominent ridge at the anterior margin of 
the orbit. The temporal fossa is large. The postorbital process, which, however, is 
very minute, is more distinctly marked than in EHrinaceus ; and immediately behind there 
is a slight contraction of the skull, so that that portion of the skull is narrower than the 
narrowest part between the orbits; and in this respect also it resembles Erinaceus, and 
differs from Tupaia. It is like the latter genus, however, in having the ridge from the 
postorbital process, reaching backwards within a short way of the lambdoidal crest ; 
but the sagittal ridge, which results from the union of the ridges of either side, is very 
feebly developed, and before it joins the lambdoidal ridge it forms a flat triangular 
surface, with the apex directed forwards. In Hrinaceus these postorbital ridges are far 
removed from the lambdoidal, by nearly half the length of the cranium. In Hylomys 
the skull attains its greatest breadth at the posterior roots of the zygomata. The 
palatal surface, unlike Z'wpaia and Erinaceus, has not any imperfections; but a faintly 
marked median ridge occurs as in Hrinaceus. It is deeper than in Twpaia, and of more 
' T regret that there is no skull of Gymnura in this Museum with which to compare Hylomys. 
