moidal plates (j), with their rods (i), along the ventral margin, can be determined ; and some of them are in good condition. The hour- glass-shaped plates (k) are quite distinct, and are regularly arranged along the same margin ; and those below the elevated portion of the dorsal fin are likewise in excellent order.
No distinct traces of either the pectoral or ventral fins are left. The dorsal fin (a), too, is very imperfect, f of an inch only remaining. The short anterior plates, or fulcra (a"), are pretty distinct ; hut the merest traces are found of the narrow posterior part. The greater portion of the anterior division of the anal fin (b) is well preserved, and the narrow posterior part (b') is determinable throughout its length. The tail is much injured, but lies, as in all the other examples, spread out, the two lobes being strongly defined.
We have now only a few concluding words to say as to the place this curious fish should occupy in the classification. We know of no family with which it can be associated, though it undoubtedly is closely allied to the Pycnodonts ; and of these perhaps it approaches most closely to Gyrodus and Microdon: but while in Dorypterus the head-bones are smooth, they are in most of the Pycnodonts granulated or otherwise ornamented ; and our fish likewise differs from them in the absence of the usual body-scales.
The large and upward-shutting mouth, too, is not found in the Pycnodonts ; and the apparent absence of the strong characteristic teeth of that group is noteworthy. Had such teeth existed in the fish under discussion, some trace of them would assuredly have been observed, since we have seen that the jaws are pretty well preserved in two or three of our specimens. It is probable therefore that the dental organs were small and inconspicuous, if they existed at all. The enormous development of the dorsal fin and the forward position of the ventrals are significant facts. The latter is particularly worthy of notice, especially when we consider that we have in this palaeozoic species the earliest thoracic fish known in the geological series. And further Sir Philip Egerton states, in a letter with which he has lately favoured us, that he "is not cognizant of any fish in strata older than the chalk having the fins thoracic or jugular." On the whole, then, we confess ourselves at a loss to determine where to locate Dorypterus in the system. Is it not the representative of a distinct, family having a certain relationship to the Pycnodonts ? This we must leave for the determination of those more conversant with ichthyology than we ourselves are.
EXPLANATION OF PLATES.
The following letters apply to all the figures of Horypterus Hoffmanni.
a, dorsal fin ; a', narrow posterior prolongation of ditto ; a", anterior prolongation, or fin-fulcra of ditto; b, anal fin ; b', posterior, narrow prolongation of ditto ; c, pectoral fins ; c', brachial rays ; d, ventral fins ; e, rhombiform scales of tail ; f, fin-fulcra of ditto ; g, lateral plates or vertebral spines ; g' bifurcation of the inner extremity of ditto ; h, lozenge-shaped plates, or areas, — the vertebral centres ; h', ridge in the centre of ditto ; i, rods, or supplementary vertebral spines, in connexion with the lateral plates ; j, sigmoidal ridges ; j', sig-
