The specimen under consideration was found by the Rev. Mr. Joass, of Golspie (to whom we are indebted for the discovery of the rich deposit of Devonian Ichthyolites at Edderton in Rosshire), on the beach between Culgower and Portgower, on the east coast of Sutherlandshire. It is imbedded in a dark-coloured shale perforated by recent Pholades, and incrusted in places by Balani. The head and tail are deficient ; but the trunk is tolerably well preserved, with the exception of the fins. It measures eleven inches in length by seven or more in depth : as the ventral scales are absent, the latter measurement is merely an approximation. Fortunately two of the tritoral teeth are preserved. Although not in situ, there can be no reason to doubt that they belonged to the specimen. They present all the typical characters of the teeth of the genus Gyrodus, viz. a small central papilla surrounded by two crenulated rings, the inner one rather higher than the outer one, having the appearance, viewed vertically, of a small rosette (fig. 1). Judging from their circular form and relative size, it is probable that they belonged to one of the secondary rows of the vomerine series. The column, as is the case in all the members of this family, is notochordal. The neurapophyses, especially in the nuchal region, are strong, and extend nearly to the dorsal ridge. They are spathulate at the lower extremities, where they partly embrace the chorda. The impressions of a few of the interspinous ossicles supporting the rays of the dorsal fin show that this organ occupied nearly the centre of the back. No evidence remains as to the position or characters of the other fins. The whole of the trunk is invested with a compact scaly armour, remarkable for the solidity of the component scales, and the strength of the joints by which they are united. In most of the Ganoid fishes the scales are united by a pin- and-socket lock, each scale having a process at the upper anterior angle, which is received in a depression on the lower angle of the scale immediately above it in the series ; but in this case each scale has a superior and inferior process, which are spliced to the corresponding processes in the contiguous scales above and below ; in addition to which, additional strength is imparted by a broad overlap at each joint, the upper scale covering the bevelled margin of the succeeding scale below (fig. 2). When viewed on the inner surface, these processes form the continuous rib or fillet which has been alluded to before. Gyrodus is the only genus of the secondary formations in which this structure prevails behind the dorsal and anal fins ; this peculiarity has therefore been taken as a good feature for rectifying the generic nomenclature of the several species. The genus Microdon of Agassiz (as I pointed out in 1849) comprised some species in which the peculiar scale -structure was partial, and others in which it was general ; the latter species, namely Microdon hexagonus and M. analis, are now correctly arranged by Heckel under the genus Gyrodus; on the other hand, Gyrodus micropterus of Agassiz differs from that genus in having no scale-ribs on the posterior half of the body, as also in having very large dorsal and anal fins and a rounded tail. These characters combined justify Wagner
2d2
