more of each scale covered than exposed; and this is so rare a ganoidal mode of arrangement, that, with the exception of the _Dipterus_,—a genus which, though it gives its name to the Dipterian sept, differed greatly from every other Dipterian,—I know not, beyond the limits of the ancient Cœlacanth family, a single Ganoid that possessed it. The bony covering of the Cœlacanths was _farthest_ removed in character from shagreen, as that of their contemporaries the Acanths approximated to it most nearly; they were, in this respect, the two extremes of their order; and did we find the Cœlacanths in but the later geological formations, while the Acanths were restricted to the earlier, it might be argued by assertors of the development hypothesis, that the amply imbricated, slate-like scale of the latter had been developed in the lapse of ages from the shagreen tubercle, by passing in its downward course—broadening and expanding as it descended—through the minute, scarcely imbricated disks of the Acanths, and the more amply imbricated tile-like rhombs of the Dipterians and Palæonisci, until it had reached its full extent of imbrication in the familiar modern type exemplified in both the Cœlacanths and the ordinary fishes. But such is not the order which nature has observed;—the two extremes of the ganoid scale appear together in the same early formation: both become extinct at a period geologically remote; and the ganoid scales of the existing state of things which most nearly resemble those of ancient time are scales formed on the intermediate or tile-roof principle.