trace, as I have said, has yet been detected. The solitary exception in the case is, however, one of singular interest. [Illustration: Fig. 18. a. _Palatal dart-head._ b. _Group of palatal teeth._] In a collection of miscellaneous fragments sent me by Mr. Dick from the rocks of Thurso, I detected patches of palatal teeth ranged in nearly the quadratures of circles, and which radiated outwards from the rectangular angle or centre, (fig. 18, _b_.) And with the patches there occurred plates exactly resembling the barbed head of a dart, (_a_,) with which I had been previously acquainted, though I had failed to determine their character or place. The excellent state of keeping of some of Mr. Dick’s specimens now enabled me to trace the patches with the dart-head, and several other plates, to a curious piece of palatal mechanism, ranged along the base of a ganoid cranium, covered externally by a brightly enamelled buckler, and to ascertain the order in which patches and plates occurred. And then, though not without some labor, I succeeded in tracing the buckler with which they were associated to the _Dipterus_,—a fish which, though it has engaged the attention of both Cuvier and Agassiz, has not yet been adequately restored. It is on an ill-preserved Orkney specimen of the cranial buckler of this Ganoid that the ichthyologist has founded his genus _Polyphractus_; while groupes of its palatal teeth from the Old Red of Russia he refers to a supposed Placoid,—the _Ctenodus_. But in the earlier stages of palæontological research, mistakes of this character are wholly unavoidable. The palæontologist who did avoid them would be either very unobservant, or at once very rash and very fortunate in his guesses. If, ere an entire skeleton of the _Ichthyosaurus_ had turned up, there had been found in different localities, in the Liasic formation, a beak like that of a porpoise, teeth like that of a crocodile, a head and sternum like that of a lizard, paddles like those of a cetacean, and vertebræ like those of a fish, it would have been greatly more judicious, and more in accordance with the existing analogies, to have erected, provisionally at least, places specifically, or even generically separated, in which to range the separate pieces, than to hold that they had all united in one anomalous genus; though such was actually the fact. And Agassiz, in erecting three distinct genera out of the fragments of a single genus, has in reality acted at once more prudently and more intelligently than if he had avoided the error by rashly uniting parts which in their separate state indicate no tie of connection. [Illustration: Fig. 19. CRANIAL BUCKLER OF DIPTERUS.] [Illustration: Fig. 20. BASE OF CRANIUM OF DIPTERUS.]