portion which projected into the mould had become brittle and porous, and presented the ordinary appearance of an old churchyard bone. And what the adipocere had done for the human bone in this case, seems to have been done for the bones of the _Asterolepis_ by the animal bitumen. [Illustration: Fig. 45. HYOID PLATE OF THURSO ASTEROLEPIS.[20] (One fifth the nat. size, linear.)] The size of the _Asterolepis_ must, in the larger specimens, have been very great. In all those ganoidal fishes of the Old Red Sandstone that had the head covered with osseous plates, we find that the cranial buckler bore a certain definite proportion,—various in the several genera and species,—to the length of the body. The drawing-master still teaches his pupils to regulate the proportions of the human figure by the seven head-lengths which it contains; and perhaps shows them how an otherwise meritorious draftsman,[21] much employed half an age ago in drawing for the wood-engraver, used to render his figures squat and ungraceful by making them a head too short. Now, those ancient Ganoids which possessed a cranial buckler may, we find, be also measured by head-lengths. Thus, in the _Coccosteus decipiens_, the length of the cranial buckler from nape to snout equalled one fifth the entire length of the creature from snout to tail. The entire length of the _Glyptolepis_ was equal to about five one half times that of its cranial buckler. The _Pterichthys_ was formed in nearly the same proportions. The _Diplopterus_ was fully seven times the length of its buckler: and the _Osteolepis_ from six and a half to seven. In all the cranial bucklers of the _Asterolepis_ yet found, the snout is wanting. The very fine specimen figured in page 99 (fig. 28) terminates abruptly at the little plate between the eyes, the specimen figured in page 98 (fig. 27) terminates at the upper line of the eye. The terminal portion which formed the snout is wanting in both, and we thus lack the measure, or _module_, as the architect might say, by which the proportions of the rest of the creature were regulated. We can, however, very nearly approximate to it. A hyoid plate in my collection (fig. 45) is, I find, so exactly proportioned in size to the cranial buckler, (fig. 28,) that it might have belonged to the same individual; and by fitting it in its proper place, and then making the necessary allowance for the breadth of the nether jaw, which swept two thirds around it, and was surmounted by the snout, we ascertain that the buckler, when entire, must have been, as nearly as may be, a foot in length. If the _Asterolepis_ was formed in the proportions of the _Coccosteus_, the buckler (fig. 28) must have belonged to an individual five feet in length; if in the proportions of the _Pterichthys_ or _Glyptolepis_, to an individual five and a half feet in length; and if in those of the _Diplopterus_ or _Osteolepis_, to an individual of from six and a half to seven feet in length. Now I find that the hyoid