202 MR. E. T. NEWTON ON SOME [Apr. 2, 
which is also in relation with the supraoccipital (s.0.) and with a 
much smaller bone ( pt.o¢.) which lies behind it. This plate (sp.ot.) 
has on its under surface a deep depression, which evidently formed 
the attachment for the hyomandibular; the depression is not, how- 
ever, quite confined to this plate, but seems to extend for a short 
distance on to the smaller bone which lies behind it. These two 
bones, therefore, which have combined to support the hyomandibular, 
are regarded as the sphenotic (postfrontal) and the pterotic 
(squamosal). 
The last-named bone comes into relation with the supraoccipital 
on the inner side and has behind it two elements, the outer of which 
is evidently the post-temporal ( pt.tem.) and the inner one ( pa.) is 
either the parietal or epiotic. Another bone which is found behind 
these in the recent Arius has evidently been lost in this fossil. The 
post-temporal ( pt.tem.) has a peculiar and characteristic structure. 
On the upper surface of the skull it presents a flattened tuberculated 
plate, which, a comparison with the recent Arius gagorides shows, 
is composed of two parts—an anterior, united with the pterotic, and 
a hinder one attached to the bone directly on its inner side, the 
parietal (or epiotic), and extending as far as the supraoccipital. The 
post-temporal must also have been in close relation with the bone 
bebind it, which has been lost; but the smooth surface now exposed 
does not indicate a close attachment. The outer extremity of the 
post-temporal has a very distinct socket for the upper part of the 
supraclavicular, and from below this a strong bar of bone passes down- 
wards and inwards quite to the base of the skull, to which it has 
been firmly attached. There is a thin plate of bone (x) behind this 
bar, which seems to have been separated from it by the crushing of 
the skull, that passes backwards and inwards, seemingly to join the 
broad thin horizontal plate formed by the transverse processes of 
the vertebree ; indeed, it may be that this is a part of the transverse 
process, extending forward to join the post-temporal. 
The hinder part of the supraoccipital is wanting, but the front 
part is preserved and is somewhat shield-shaped, having its margins 
on each side indented by a series of curves where it is united with 
the frontal, sphenotic, pterotic, and parietal. The hindermost tem- 
poral plate, now wanting, was doubtless also in contact with the 
supraoccipital, 
The base of the skull is much crushed, but it seems quite certain 
that the brain-cavity was completely enclosed by bony walls. The 
buttress-like supports of the post-temporal bones are striking features 
of this region. Just in front of the left buttress may be seen the 
remains of an inflated bulla, which lodged one of the large otoliths, 
and still further forward there is on each side a depression for the 
attachment of the hyomandibular (Am.). In front of this again is 
another pair of elongated depressions, evidently the ethmoid attach- 
ments for the prefrontals (fig. 2, et/.). The most anterior part of 
the base, in its present condition, is formed by a fan-shaped mass, 
which is apparently in part the vomer and in part the ethmoid. 
Three or four of the anterior vertebre are firmly united to each 
