40 



OLD RED SANDSTONE FISHES. 



Egerton. I was completely astonished on comparing the specimen kindly lent to 

 me by Sir Philip with the figure which has been so widely copied. The specimen 

 presents the concave surface of half of the head-shield and the inner surface of the 

 body, showing well the large lateral scales ; but none of the detail as to dorsal scales 

 which the artist has drawn is to be seen, and there is no real foundation for assigning 

 two rows to each dorsal series of scales. In Pteraspis undoubtedly there are many 

 dorsal rows of nearly equilateral rhomboidal scales, and one may be quite prepared to 

 find that such was the case in some Osteostraci. The restored outline sketch ("woodcut, 

 fig. 16) of a Cephalaspis [Eucephalasjyis LyelJU) gives a better notion of the relative pro- 

 portions of the scales than mere words can, especially when compared with some of the 

 plates. But it is necessary to state again that the locuj in which the series of scales break 



Fig. 15 </, 6. 



Loricaria platystoma, Giintlier, a recent fresh-water fish, of the Siluroid group, from Surinam.' 



up at the dorsal fin is not known. The scales are thicker and broader in C. Lyellii than 

 in C. Powriei relatively. They are deeply imbricated in each series, and have a consider- 

 able thickness of structure, histologically agreeing, as far as I have been able to examine 

 them, with that of the head-shield. The body in vertical transverse section presents a tri- 

 angular form as seen in fig. 17. It appears that the scales of the lateral series admitted 

 of considerable movement of the body, though so large and strong ; for, whereas in the 

 specimen drawn in PI. VIII they are inclined posteriorly, in that drawn in PI. XI they 

 present exactly the opposite direction. This is, no doubt, due to muscular contraction, the 

 1 I have to tliank the Secretary of the Zoological Society for the use of the cuts, figs. 14 and 15, 

 ■which illustrate a paper by Dr. Giinther. 



