600 l>i; RAMSA*X H. TRAQUAIR ON TtEELODXJS PAGEI. 



At tins stage we may take notice of the fact that the head affords us no traces what- 

 ever of teeth or of jaws. Where the mouth was is not evident, and we are as little 

 informed as to the position of the eyes. 



The tail is 2|- inches broad at its origin, and becoming rapidly narrower, ex- 

 tends back for 6|- inches where it is broken off. Unlike the anterior part of the 

 animal the tail appears laterally compressed, but nothing can be seen in its structure 

 but a mass of the shagreen-bodies to be presently described. Before this tail is cut 

 off by the edge of the stone, part of a median fin is seen passing off obliquely on 

 the aspect opposite to that from which we view the fish as a whole. As there can 

 be no doubt that this is the lower lobe of a heterocercal caudal, it follows that 

 the aspect of the body, which is exposed in the fossil, is the dorsal, and not the ventral 

 as supposed by Powrie. 



There is no evidence in this specimen of the presence of ventral, dorsal, or anal fins. 



The only actual structures which are preserved in the specimen, are the scales or 

 shagreen-bodies with which the entire surface of the fossil is closely covered, from 

 the front of the head to the broken off extremity of the tail. They are mostly of 

 an oval form, as seen in figs. 2-4, though they are more nearly round at the anterior 

 margin. Their size is also pretty uniform, the long diameter measuring ^- to^y inch, 

 but on the lateral fin-expansions they are considerably smaller. The outer surface 

 (fig. 2) is brilliantly ganoid, slightly convex, and distinctly nicked or crenulated 

 round the margin ; the lower surface (fig. 3) is also convex, and shows in the middle 

 a comparatively large hole or opening passing into the interior. Most of the scales 

 in this specimen are. however, broken through horizontally, as seen in fig. 4, and here 

 the internal cavity, into which the hole in the base passes, is very obvious and of 

 considerable size. 



To complete our knowledge of the configuration of these scales, it is, however, 

 necessary to turn to a second specimen, from Canterland, Kincardineshire, consisting 

 only of a mass of the shagreen about 2 inches square, referable, I believe, to tin 1 same 

 species but to a larger and older individual. The scales are larger, their long diameters 

 often measuring y 1 ^ inch, and their form is frequently more irregular as seen in 

 figs. 5-8. 



In tig. 5 three scales are represented, magnified eight diameters, one of which, of 

 a peculiarly narrow shape, is seen from below, while the other two are seen from the 

 side. Here we see that the smooth rounded basal portion is separated by a slight 

 constriction from the upper part or crown, and that the crenulations of the margin 

 of the latter are continued down the " neck " as a sort of vertical fluting as far 

 as the upper boundary of the base. Fig. 3 represents the lower surface of a scale 

 similarly magnified from which we see that it is strongly convex and provided 

 w r ith a very small opening leading into the interior, but in many cases this hole is 

 quite imperceptible as in the two lower scales of the group of four shown in fig. 6 ; 

 in this figure we also see that the convexity of the base may be in places obtusely 



