4 ^Ir. A. S. Woodward on the 



Limestone (Lower Devonian) of Ohio. The type specimens 

 comprise a series of detached plates, no-y in the museum of 

 Columbia College, New York, some of these exhibiting the 

 lateral appendages which suggested to Dr. Newberry the 

 generic and specific name. Li the original description the 

 appendages were compared with the cornua of typical Cepha- 

 laspidian fishes, and the new genus w.is tlius supposed to 

 pertain to the latter group, differing from all known forms in 

 having a cephalic shield composed of several distinct plates. 

 If, however, the iSpitzbergen fossils prove to be correctly 

 interpreted below, Acanthaspis is most nearly related to the 

 Asterolepida3, and its spine-like processes are the homologues 

 of the well-known pectoral paddles of the latter. 



Acanthaspis decipienSj sp. n. (PI. I.) 



1884. Cephalaspis (cf. C. Agassizii), E. R. Lankester, op. cit. p. o, pi. i. 

 fig. 4. 



Pectoral appendages comparatively broad and very gently 

 arched, ornamented in the proximal two thirds by closely 

 arranged longitudinal series of fine tuberculations. Median 

 ventral plate relatively large, about as broad as long, occu- 

 pying more than half the width of the ventral shield at its 

 middle point. Tubercular ornament very fine and closely 

 arranged. 



This species is based upon a small slab of red sandstone 

 exhibiting remains of dermal armour that are at first sight 

 somewhat difficult of interpretation. Before the investing 

 matrix was completely removed, one portion of the fossil wjis 

 briefly noticed and figured by Lankester {loc. cit.) as indicating 

 the occurrence of a species of Cephahn^pis related to the 

 British C. Agassizii ; but it is now evident that the organism 

 in question is quite distinct from any hitherto satisfactorily 

 determined and cannot be referred even to the family to which 

 Cephalaspis belongs. 



As shown in the drawing of the natural size the fossil 

 exhibits two distinct portions of armour, each provided on one 

 side with a large, curved, spinous process. As remarked by 

 Lankester, the two spines ajipear to be essentially identical, 

 and hence it is reasonable to infer that both parts of the tbssil 

 pertain at least to the same species, if not to the same indi- 

 vidual. The one shield (a) is distinctly convex on the external 

 aspect, for the exposed concave face of the specimen is smooth 

 and has the characteristic appearance of the visceral as[)ect. 

 The other shield (ll) is chiefly shown as an imj)rossion ; but 

 it is proved to be nearly iiat, and some iVagments ot' the 



