THE LOWER DEVONIAN EISHES OF GEMtJNDEN. 731 



figs. 1 and 2, that the ventral fulcra are shorter and smaller than those of the dorsal 

 margin. 



Both sets exhibit an ornate sculpture, which consists of the tubercles of the ordinary 

 scales, but more lengthened out, so much so in some cases as rather to be described as 

 ridges (PI. V. fig. 1). 



The sides of the tail are clothed with angular imbricating scales, the external 

 sculpture of which has been already noted ; but a fact not recorded in my previous 

 descriptions and restorations is, that between the carapace and the origin of the caudal 

 fin there is at least one longitudinal row of scales which are considerably higher than 

 broad, — the cause of my not having noted this before being that this part of the tail is 

 usually more or less covered and obscured by pyritous deposit, as seen to a marked extent 

 in PL I. fig. 1, PI. IV., and PL V. figs. 1 and 2. Further on, the scales (PL V. fig. l) 

 become more equilateral and acutely rhombic, and ever smaller as we proceed backwards. 

 The caudal fin (PL I. fig. 1, PL IV., PL V. fig. 2) is not bilobate, — in that respect 

 resembling that of Cephalaspis and Pterichthys. It is heterocercal, inasmuch as the 

 dorsal apex passes further back than the ventral one, and the fulcra along that margin 

 are larger than those along the other. There is, however, no definite line of demarca- 

 tion between the scales of the body-prolongation and those of the fin-membrane, though 

 the latter become gradually smaller. 



The plates of the carapace are thin when compared with those of most other 

 armoured fishes ; and as they are entirely converted into iron pyrites, it is un- 

 fortunately quite impossible to study their microscopic structure. The creature 

 attained a very considerable size, the largest entire specimen which I have seen being 

 that represented in PL IV., and which measures 1 85 inches in length. 



Observations. — I have nothing to add to or to alter in my opinion as to the more 

 immediate affinities of Drepanaspis as expressed in my paper on the Silurian Fishes of 

 Scotland. We have here a fish-like creature whose hard parts are entirely dermal ; 

 whose endoskeleton must therefore have been quite unossified in any part, as no traces 

 of it can be found ; and whose mouth, a simple transverse slit, shows no teeth nor 

 anything which can be called a mandible. These characters assign to Drepanaspis 

 a place in the Ostracodermi, and its indubitable resemblance to Pteraspis leads us to 

 class it in the Heterostracous subdivision, although evidence from microscopic structure 

 is unfortunately unavailable. But if it be allied to Pteraspis, it is also clearly related 

 to Psammosteus and to Thelodus ; and hence, in my " Silurian " memoir, I included in 

 the Heterostraci not merely the Pteraspidce, as formerly, but also the Drepanaspidse, 

 Psammosteidse, and Ccelolepidse, the last being looked on as the least, and the first as 

 the most specialised member of the group. To enter again into the question of the 

 origin of the Ccelolepidse does not come into the scope of this paper, which is purely 

 descriptive ; but one cannot help remarking that the structure of Drepanaspis does not 

 seem to lend much support to the idea of the evolution of the Ostracodermi from a 

 Eurypterid ancestry. 



TRANS. ROY. SOC. EDIN., VOL. XL. PART IV (NO. 30). 5 R 



