30 FISH OF THE OLD RED SANDSTONE. 



that of the left side being in all the species larger externally than 



that of the right. 



Coccosteus pusillus (M^Coy). 



Sp. Char. Head and carapace orbicular, width about 2^ inches, 

 length 3 inches ; tail about as long as the head and carapace, 

 of very numerous small, weak ( ? slightly ossified) apophyses ; 

 dorsal fin small, weak ; dorsal plate subpentagonal, 1 inch and 

 5 lines long, greatest width (at the lateral angles) 8 lines, taper- 

 ing to a sharp point retrally, also narrowing about one-sixth 

 towards the subtruncate anterior margin, all the margins con- 

 cave, the anterior most so, middle of the plate obtusely keeled, 

 the tuberosity and fossa (for lodging the dorsal spine ?) rather 

 more than one-third the length from the posterior apex ; all 

 the plates of the carapace minutely and regularly tuberculated, 

 granules nearly equal, about their own diameter apart, four- 

 teen in the space of a quarter of an inch, intervening spaces 

 very minutely granulated ; teeth slender, cylindrical, pointed, 

 their own diameter apart, one-third of a line in diameter, 

 nearly a line long. 



The very small size and imperfect development of the ver- 

 tebral apophyses, together with the small size and orbicular 

 form of the cephalothorax, easily distinguish this species. The 

 peculiar proportion of the dorsal plate, as well as the distance of 

 its tuberosity and fossa from the apex, and the minuteness and 

 regularity of the tuberculation, distinguish it from the young of 

 the other species ; besides, I find all the characters constant in 

 five nearly perfect specimens which I have examined. 



Not uncommon in the black flags of the old red sandstone at 

 Orkney. 



[Col. University of Cambridge.) 



Coccosteus microspondylus (M'Coy). 

 This species resembles the C. oblongus (Ag.) in size and gra- 

 nulation, but has a much shorter dorsal plate, and the posterior 

 external angles of the posterior ventral plates are produced into 

 long curved processes as in the C. latus (Ag.), from which it 

 differs in its strong, regular, close granulation ; it differs besides 

 from both species in the plates of the carapace being shorter, and 

 most remarkably in the much smaller size and slight ossification 

 of the vertebrae, giving a peculiarly weak and slender appearance 

 to the tail. The bodies of the vertebrae have all left their sepa- 

 rate impressions, so that the vertebral column was certainly not 

 in this case a continuous cartilaginous cord as suggested by 

 M. Agassiz in the case of the C. latus, in which they leave no 

 trace. There is evidence of a thick spine, about an inch long, being 

 attached to the tuberosity of the dorsal plate. The dorsal plate 



