BY MESSRS. R. M. JOHNSTON AND A. MORTON. 103 



scarcely visible, but appear to be small ; and the root is only 

 about 10 lines distant from a vertical drawn through posterior 

 portion of head. Pectorals about 7 lines in length, and. 

 consists, apparently, of about 15 slender rays. 



There are 56 rows of small rhomboid scales, longitudinally 

 arranged in an inclined dorso-ventral series ; the caudal series 

 being more perceptibly angled than the anterior series. The 

 inner surface of each scale is alone visible, from which it clearly 

 appears that each one is finely ridged longitudinally, as in 

 the scales of Acrolepis. There are usually 4 slightly curved 

 ridges, radiating longitudinally from posterior angle of 

 rhomboid scale to the two inner ones, almost invariably 

 becoming furcate as they approach anterior inner margin ; 

 the outside one on either side smaller and almost invariably 

 simple. The upper margin of tail is markedly serrate, 

 indicating the presence of numerous pointed fulcral scales. 



The only Austrahan fish which appears to come near it is 

 the well-known Myriolepis Clarkei, Egerton, but it is evident 

 from the description and drawings that the Tasmanian 

 Ganoid has relatively much smaller fins, and the scales, though 

 belonging to a specimen half the size, are relatively much 

 larger and consequently less numerous. 



Age of the Eocks in Which the Fish Remains Occur. 



The discovery of this interesting fossil is another proof of 

 the aquedus origin of the important series of sandstone beds, 

 of which the section from Cascades to Knocklofty affords the 

 best and most fully developed example. Although the shales. 

 contain impressions of what appear to be fucoids, the 

 evidences are not sufiicient to determine whether these basins 

 were estuarine or lacustrine; or whether the waters were 

 fresh, brackish, or salt. Ganoid fishes of the period are found 

 under all such conditions ; and therefore their discovery in 

 such deposits prove little further than to indicate the aqueous 

 origin of the beds in which such remains occur. It is most 

 probable that the waters were of the nature of brackish 

 lagoons. The exact position of these sandstones in relation 

 to the Mesozoic Coal Measures, on the one hand, and the 

 Upper Paleozoic Mudstones, on the other, has ever been one 

 of much doubt. 



It is true a similar series of sandstones at Adventure Bay 

 appear to immediately succeed the Upper Carboniferous Coal 

 Measures without any sign of stratigraphic break; and 

 again at Passage Point this succession appears to be very 

 complete in immediate relation to beds of the Upper Marine 

 series. But the absence of fossil evidence, and the manner 

 in which the several deposits are separated from each other, 

 by distance or faults and intrusive rocks, make it a doubtful 



