from which it came. To reach them, a path had to be cut into the face of the cliff, and the rocks overlying the beds removed with dynamite, or with pick and bar; an operation not without some danger from rocks that from time to time fell from the crumbling cliffs above. Twice, without warning, fragments weighing some fifty pounds each fell at our feet, knocking the tools, from our hands; and where they came from there were other loose ones, at least ten times as large. Our attention was reluctantly, but impartially, divided between the maintenance of a precarious footing on the face of the sweltering cliff, the threatening rocks overhead, and the treasures at our feet.
The bed was not extensive, but it proved to be literally teeming with these extraordinary animals (Bothriolepis canadensis), and by good fortune they were in a remarkably complete and instructive state of preservation; more so, perhaps, than any other fossils found heretofore (Fig. 6).
The bed had apparently formed the bottom of a shallow, brackish water-pool in which fern-like water plants had been growing, and where many millions of years ago, with the rise and fall of the tides, these specimens had been trapped, together with other species of ostracoderms and several kinds of true fishes.
The soft mud on the bottom of the pool was now turned into a fine-grained, sandy limestone, and in it the fossilized animals were preserved in the very attitudes they had assumed when they ceased to struggle out of the enclosure. One, in its death agony, had plunged into the mud with sufficient force to remain there, head down, in a vertical position. Others were arranged in horizontal series, uniformly headed in a northeast direction. Their heads were turned against a gentle current of water, as was shown by the fact that the tops of all the ferns were pointed in nearly the opposite direction.
Many of these specimens were so well preserved that the shape of the whole body, and many details on its external surface, could be readily observed; the general character and location of the principal sense organs, jaws, gills, stomach, anus and genital openings ascertained; and the neural and hæmal surfaces identified. It was also possible to determine, beyond reasonable doubt, the mode of locomotion, the mode of feeding and the nature of the food.
Of course this message from a remote past was not off-hand legible. After the scores of specimens were safely housed in the laboratory, it required nearly three years to chisel and scratch and brush away the rocky matrix in which they were imbedded, and after that many specimens had to be cut into serial sections with a diamond saw, and the sections polished and varnished to show the arrangement of the internal organs.
