84G DR RAMSAY H. TRAQUAIR ON FOSSIL FISHES COLLECTED BY THE 



few plates of a larger size, which we may term rostral (r.). Lastly, on each lateral 

 margin, a little way in front of the termination of the postero-lateral plate, may be seen 

 a small rounded depression (x.) which may represent an orbit, though in one case I see 

 on its floor a portion of bone sculptured like the rest of the plate, but this may be 

 due to some accidental displacement. This depression or cavity is of such constant 

 occurrence that it must mean something. 



On the ventral surface there is likewise a large oblong central or median ventral 

 plate, but its posterior notch is larger, and its direction is continued forward for a little 

 distance by a slight elevation or fold of the surface — hence, when detached, the plate 

 can always be readily distinguished from the corresponding oue on the back. The 

 postero-lateral angles and margins are formed by the same plates (p. I.), which we saw 

 on the dorsal surface, but a much narrower area of each is exposed. There are also a 

 few broad plates at the anterior margin, the rest of the surface being filled in by small 

 polygonal ones, exactly as on the dorsal aspect of the carapace. 



All the plates of the carapace, both above and below, are ornamented by tolerably 

 closely-set stellate tubercles. There is no trace of jaws or of teeth, but as it is impos- 

 sible to conceive of the absence of a mouth, we must conclude that it was placed exactly 

 at the anterior margin. 



The tail is rather shorter than the carapace, and is covered with tuberculated quad- 

 rangular scales, which, becoming finer and smaller, are continued on the caudal fin, 

 which is heterocercal but scarcely bilobate. But in addition to those lateral scales, we 

 have, running along both upper and lower margins of the tail and caudal fin, a series of 

 stout elongated imbricating " fulcral scales," those of the dorsal series being the longer. 

 No trace of any other fins is to be seen, nor do we find any remains of internal skeleton. 



After this brief description of the leading points in the structure of Drepanaspis, it 

 is impossible even in the absence of evidence as to the microscopic structure of its hard 

 parts, to avoid the conclusion that it is related to the Ccelolepidse — in fact, that it forms 

 an onward stage in the evolution of a common series to which the last-named family 

 belongs. 



We cannot fail to recognise the general resemblance in form — the broad and 

 depressed anterior portion, rounded in front and truncated behind; the want of jaws 

 and teeth ; the slender tail with heterocercal caudal fin. But whereas the minute 

 shagreen-scales of Thelodus have in Ateleaspis coalesced into polygonal tesserae in front 

 and rhombic scales behind, in Drepanaspis we have, added to the rhombic scales of the 

 tail, well-developed fulcra, and as to the anterior part, we find that it is now covered by 

 a regular carapace, into which enter not merely a multitude of small polygonal plal 

 but also several large ones, among which the great median dorsal and ventral plates and 

 the falciform plates at the postero-lateral angles are conspicuous. Then, if the postero- 

 lateral lappet-like projections of the body of Thelodus and Lanarkia represent lateral 

 fin-flaps, as I believe them to do, we find these parts in Drepanaspis, though still pre- 

 serving the same general contour, rendered utterly functionless as fins by being enclosed 



