320 pal;eontology of new-youk. 



over this extent, numerous careful collectors, it will be admitted that, if such remains do exist, 

 they are extremely rare. 



In England, however, we have positive e\'idence of the existence of such remains in strata 

 of the age of our Trenton limestone and Hudson-river group ; and although we may not feel 

 quite satisfied of the existence of such organisms at a period 'below the silurian, we are com- 

 pelled by the evidence before us to admit that fishes were coeval with the earlier fossiliferous 

 deposits of this country. From the Clinton group upwards, we have these remains in every 

 successive rock as far as the Red Sandstone of the Catskill mountains. 



Genus ONCHUS (Agassiz). 

 1. ONCHUS DEWEII. 



PL. LXXI. Fig. 1 a-d. 



Spine slender, elongated, gradually attenuated and incurved, longitudinally grooved ; an- 

 terior side marked by the bases of shorter spines ; basal portion (or part enclosed beneath the 

 surface) enlarged and projecting in the middle towards the posterior side ; lower extremity 

 constricted. The surface of this portion is ornamented by imbricating, narrow lanceolate 

 elevations which are obliquely striated. This ornamental marking terminates just above the 

 constricted base, which appears to have been more firmly fixed in the tendinous envelope, 

 while the ornamented upper portion marks the attachment of smaller muscles. 



Fig. 1 a. A fragment of shale, retaining the base of the spine, with about two inches of the lower 

 portion, and the impression beyond this nearly to the apex. 



Fig. 1 5. A fragment, retaining the impression of the base and lower portion of the spine of the 

 preceding specimen, with the higher portion of the spine itself. The part of the spine 

 remaining in these two specimens is about six inches ; and from the appearance of 

 the broken summit, it was originally more than an inch longer. 



Fig. 1 c. The basal portion of a much larger specimen, showing very distinctly, the ornamented 

 surface and constricted base. There are some slight differences in the surface markings 

 between this and the smaller one, but these are probably not of specific importance. 



Fig. 1 d. An enlargement from the surface of 1 c. 



If the spine were projected from this base, taking the sam^e proportions as in the 

 smaller specimen, its entire length could scarcely have been less tlten two feet; and 

 taking the length of the larger base alone, as compared with the length of the other, 

 the larger spine would have been about twenty-eight inches in length. The bases 

 of the shorter species, or barbs, appear as if upon one side of the specimen figured : 

 this is apparently due to pressure, as the specimen appears to be distorted, and no 

 similar markings occur on the opposite side. 



Position and locality In the shale of the Niagara group at Lockport and Rochester. 



