its name is taken. Fig. 22 crescent-shaped blade of a saddler's cutting-knife, the body forming the handle. It is extremely broad and flat, extending on each side considerably beyond the body, and the bones appear to have been firmly soldered together, so as to form one shield, the whole head thus being apparently covered by a single plate of enamelled bone, and when seen detached from the body hardly to be distinguished from the head of a trilobite. The body compared with this singular head appears extremely diminutive; the back is arched, and gradually recedes in elevation towards the tail, which is of moderate length; the fins are few in number, and not very powerful, but appear to have possessed a bony ray in front, the rest of the fin being more fibrous. The whole body was covered with scales, which varied in shape in different parts, and seem to have been disposed in series. This fish never seems to have attained a large size; the best preserved specimen having a length of only seven inches, with a breadth of three inches between the points of the BUCKLER-HEADED FISH. * The Old Red Sandstone; or, New Walks in an Old Field, by Hugh Miller, p. 138. crescent-shaped buckler. It has been supposed by Professor Agassiz that the singular shape of the head served as a sort of defence to this animal in case of attack; and one can readily imagine that the soft substance of the orthoceratites, probably the largest and most formidable of its enemies, would be injured by any attempt to swallow so singular and knife-like an animal as the one before us. Like many, and indeed most of the species belonging to the Ganoid order of fishes, and common in the older rocks, the bones of the head, and the scales of this strange monster, were composed internally of a comparatively soft bone, but each was coated with a thick and solid plate of enamel, of extreme hardness, and almost incapable of injury by any ordinary amount of violence. The detached scales, the bucklerhead, and sometimes the complete outline of the animal, have thus been able to resist destruction, and are found in sandy rocks, composed of such coarse fragments that their accumulation would seem to have been accompanied with violence sufficient to have crushed to powder almost any remains of organized matter, and from which, indeed, we never obtain any fragments of shells or other easily injured subThe remains of this fish have been found in Herefordshire and many parts of Wales, as well as in Scotland, and lately also in Russia; but the animal was strictly confined to the period of the old red sandstone, though it is not easy to guess what may have been its habits, in what depth of water it preferred to live, or in what way it obtained its food. stances. The Pterichthys (fig. 23) is even more strikingly different from any existing species of animal than the singular monster we have just been considering. Reverting to the graphic description of Mr. Miller, we find it compared to the figure of a man, rudely drawn, Fig. 23 HORNED WING-FISH. (Pterichthys cornutus.) the head cut off by the shoulders, the arms spread at full length as in the attitude of swimming, the body rather long than otherwise, and narrowing from the chest downwards, one of the legs cut away at the hipjoint, and the other, as if to preserve the balance, placed directly under the centre of the figure, which it seems to support.* Something of this appearance is indeed presented in the fossil remains of these creatures, once * Miller, ante cit., p. 49. the tenants of the sea in our own latitude; but we are now able to describe with more minuteness, if not so vividly, the real nature of the animal. It was of small size, not more than a few inches or a foot in length; its head and body were defended by strong plates of bone coated with enamel; and its shape and proportions were singularly unlike those of ordinary fishes; the head being small, and the body much flattened, but swelling out immediately at the junction of the head and neck, and gradually tapering thence towards the tail. From the junction of the head and body there extended that pair of singular paddles or wings from which the genus has been named, and which have been supposed to answer the same purpose as the horns of the crescent-shaped shield of the Cephalaspis, and defend the animal from the attacks of its soft-mouthed enemies. Besides these paddles, which were hard and pointed, and nearly as long as the body, at least some species of Pterichthys seem to have been provided with another smaller pair, extending from the part where the body is attached to the tail; and it is thought that this second pair of wings may be the remains of anal fins, the other pair representing the pectoral fins. The body, like the head, was certainly covered on the upper side by hard plates accurately fitting one another; but the lower part both of the head and body was probably defended by tough skin, capable of distension, and enabling the creature to swallow prey of large size. The position of the mouth is not known with certainty, but it may have been formed by a transverse slit, covered by thick fleshy lips, situated round the edge of the plate which defended the head; this position, and the absence of teeth, readily accounting for the difficulty there is in discovering remains of it in imperfect specimens. The eyes and the apertures of the nostrils were probably extremely small, and placed on the edge of the broad plate, the only indication of the head hitherto met with. The tail was not long, but seems to have been thick and conical, and covered with scales, overlapping each other like the tiles on the roof of a house. The departure from the general form of most fishes in this animal is so remarkable, that when first discovered, it was looked upon by some naturalists as an insect, by others as a crustacean, and by others again it was thought to be connected with reptiles, owing to the singular resemblance of one small species to the shell of a tortoise. Strange as it undoubtedly is, however, in all respects, this genus forms one of an extinct natural family of fishes, and it is allied to the other genera of its class by the genus Coccosteus, which at one time was thought still more anomalous. Fig. 24 COCCOSTEUS. The Coccosteus (fig. 24) is entirely without the winglike projections which characterise the Pterichthys, and while when seen as in ordinary specimens, lying on its back and crushed, it appears to bear no resemblance any fish or other animal either recent or extinct, it was not in reality much unlike many well-known to |