34 Proceedings of the Royal Physical Society. 



therefore clear that no comparison can be instituted between 

 Pterichthys and Aster olepis so far as the tail is concerned. 



There remains the body-carapace. This is more depressed 

 than in Pterichthys, but the number and general arrangement 

 of the plates are the same. As regards their mode of articu- 

 lation Pander does not enter into any great detail either in 

 his figures or text ; but he makes one important statement 

 reofardinor the anterior median dorsal which demands atten- 

 tion, namely, that its lateral margins have on the under side 

 narrow squamous surfaces which overlap both the anterior 

 and posterior dorso-laterals (" unter welche sich die beiden 

 seitlichen Schilder 12 und 13 unterschieben "), a statement 

 borne out also by his figures of the plates in question. Now 

 we have seen that in Pterichthys the anterior median dorsal 

 plate does not overlap the posterior dorso-lateral, but is 

 certainly overlapped by it, so that w^e have in this circum- 

 stance a quite tangible mark of distinction between the two 

 genera. 



I have not seen the anterior dorso-median plate of Astero- 

 lepis ornatus ; but in the Upper Old Eed Sandstone of 

 Nairnshire remains of a large Asterolepid are not uncommon 

 in which this plate certainly had the same relations to the 

 surrounding ones as Pander has described in the Russian 

 form. This is the Coccosteus maximus of Agassiz (4, p. 137, 

 tab. XXX. a., figs. 17 and 18), who supposed the plate in 

 question was a median ventral, while Hugh Miller, with a 

 better conception of its real nature, wished to consider it the 

 dorsal plate of " Pterichthys " major. Having now got 

 together a very instructive set of its plates, I find that this 

 creature is not Pterichthys major, which is in reality a 

 Bothriolepis, but a species closely resembling the Pterichthys 

 of the lower beds in all essential respects save its depressed 

 form and the mode of articulation of its anterior median 

 dorsal plate. In PI. XL, Figs. 1 and 2, I have given outlines 

 of the upper and lower aspects of this plate, the articular 

 surfaces being shaded by horizontal lines. There it will be 

 observed that on the outer aspect (Fig. 1) there is no articular 

 surface but the one, z, at the posterior margin which is over- 

 lapped by the posterior median dorsal, while on the under 



