----a thousand fearful wrecks; A thousand men, that fishes gnaw'd upon: Wedges of gold, great anchors, heaps of pearl.
Had he also beheld, amid "the dead bones that lay scattered by," the carcasses of lions, deer, and the other wild tenants of the forest and the plain, the fiction would have been deemed unworthy of the genius of Shakspeare. So daring a disregard of probability and violation of analogy would have been condemned as unpardonable, even where the poet was painting those incongruous images which present themselves to a disturbed imagination during the visions of the night.
Until lately it was supposed that the old red sandstone, or Devonian rocks, contained no vertebrate remains except those of fish, but in 1850 the footprints of a chelonian, and in 1851 the skeleton of a reptile, allied both to the batrachians and lizards, were found in a sandstone of that age near Elgin in Scotland.[212] Up to the year 1844 it was laid down as a received dogma in many works of high authority in geology, that reptiles were not created until after the close of the carboniferous epoch. In the course of that year, however, Hermann Von Meyer announced the discovery, in the coal measures of Rhenish Bavaria, of a reptile, called by him Apateon, related to the salamanders; and in 1847 three species of another genus, called archegosaurus by Goldfuss, were obtained from the coal of Saarbruck, between Treves and Strasburg. The footprints of a large quadruped, probably batrachian, had also been observed by Dr. King in the carboniferous rocks of Pennsylvania in 1844. The first example of the bones of a reptile in the Coal of North America was detected so lately as September, 1852, by Mr. G. W. Dawson and myself in Nova Scotia. These remains, referred by Messrs. Wyman and Owen to a perennibranchiate batrachian, were met with in the interior of an erect fossil tree, apparently a sigillaria. They seem clearly to have been introduced together with sediment into the tree, during its submergence and after it had decayed and was standing as a hollow cylinder of bark, this bark being now converted into coal.
When Agassiz, in his great work on fossil fish, described 152 species of ichthyolites from the Coal, he found them to consist of 94 placoids, belonging to the families of shark and ray, and 58 ganoids. One family of the latter he called "sauroid fish," including the megalicthys and holoptychius, often of great size, and all predaceous. Although true fish, and not intermediate between that class and reptiles, they seem to have been more highly organized than any living fish, reminding us of the skeletons of saurians by the close suture of their cranial bones, their large conical teeth, striated longitudinally, and the articulation of the spinous processes with the vertebrae. Among living species they are most nearly allied to the lepidosteus, or bony pike of the North American rivers. Before the recent progress of discovery above alluded to had shown the fallacy of such ideas, it was imagined by some geologists that this ichthyic type was the more highly developed, because it took the lead at the head of nature before the class of reptiles had been created. The confident assumption indulged in till the year 1844, that reptiles were first introduced into the earth in the Permian period, shows the danger of taking for granted that the date of the creation of any family of animals or plants in past time coincides with the age of the oldest stratified rock in which the geologist has detected its remains. Nevertheless, after repeated disappointments, we find some naturalists as much disposed as ever to rely on such negative evidence, and to feel now as sure that reptiles were not introduced into the earth till after the Silurian epoch, as they were in 1844, that they appeared for the first time at an era subsequent to the carboniferous.