204 MR. E. T. NEWTON ON SOME [Apr. 2, 
the indentations of the margin are alike in the two forms, but the 
supraoccipital of the fossil is evidently a shorter and wider bone, 
and it is uncertain whether it was constricted in the middle or not. 
The supraoccipital of A. egertont is only slightly constricted, and 
it may be that the Barton skull, here described, more resembled 
that species. 
Several of the anterior vertebre of A. gagorides are united to 
form one mass, in a manner precisely similar to that which obtains 
in the fossil. 
The differences above noted show clearly that the reeent and 
fossil forms are specifically distinct, while at the same time the 
resemblances are sufficiently important to prevent a generic separa- 
tion; but there is still some additional evidence which supports this 
decision in an unexpected manner. 
While clearing away the matrix from the right side of the fossil 
skull, where it is broken near the back, I was fortunate enough to 
find one of the otoliths in place, and this, when extracted, proved to 
be of a remarkable form (figs. 3, a, 5, €), and quite unlike the otolith 
of any fish with which I was acquainted. 
When found, this otolith had the more pointed end directed back- 
wards and outwards, with the smooth surface upwards; and as it 
seemed to be in its natural cavity, for the present this is regarded as 
its proper position ; but seeing that the otolith of the opposite side 
is not in its place, and that in the dried skull of the recent species 
they are loose in the brain-cavity, this may not be correct. 
The otolith is proportionately large and thick, its upper surface 
(a) is smooth and convex, while its lower surface (4) is rugose and 
much more convex. At first sight there appears to be no sulcus 
acusticus, but probabiy it is represented by the sinuous groove on 
the lower surface which passes from the hinder pointed end to the 
opposite extremity, that is between the two stars in figure 36, The 
rugosity of the under surface is due to a number of concentrie 
striations, or lines of growth, crossed by several radiating ridges 
which are stronger on the inner than on the outer portion. One of 
the radiating ridges is especially strong, and forms a prominent angle 
where it reaches the inner margin, towards the front of the otolith 
(fig. 3b, 2). The poimted extremity is seen to be notched, a slight 
groove extending from this both on the upper and lower surfaces 
(figs. 3a, 6). From this bifid point a shallow groove extends along 
the outer margin, becoming a mere line towards the front ; it is seen 
in an upper view (fig. 3 a), and is separated from the smooth upper 
surface by a fine but distinct raised line. 
A similar otolith to this is figured by Herr E. Koken, from the 
Oligocene of Headon Hill, Isle of Wight (6), who, having no clue 
to its affinities, called it Otolithus (incerte sedis) crassus. 
The close relationship between the recent Arius gagorides and the 
Barton fossil skull made it particularly desirable to see whether the 
otoliths would show a corresponding resemblance, and Dr. Giinther 
very kindly had the otoliths taken out of the skull with which the 
above comparisons had been made. One of these otoliths is repre- 
