CORRELATION OF THE INVERTEBRATE FAUNA. L63 



An analysis of the fauna will not be attempted at present, as the col- 

 lections now being made will enlarge the data for comparisons, and the 

 final study of the fauna will result in the identification of a greater 

 number of species. I think sufficient data are given clearly to prove that 

 the invertebrate fauna of the Harding sandstone corresponds to that of 

 the lower Trenton of the New York section or the lower Bala of Wales. 

 The fauna of the two limestones is to be compared to that of the middle 

 and upper Trenton of America or the Bala of Europe. It is not to be 

 expected that an absolute correlation can be made of all the genera and 

 species common to the Colorado, Mississippi valley and New York sec- 

 tions. The vertical range of some genera and species will be found to 

 vary, but as a whole the succession is the same in the several sections. 



The discovery of so huge and varied a fauna of Trenton fades is of 

 great interest, irrespective of its bearing on the stratigraphic position of 

 the ichthyic fauna, it clearly proves the continuation of the fauna of the 

 Trenton sea from Wisconsin, Iowa and Missouri to the western side of 

 the great interior sea.* 



The range of Halysites catenulatus has hitherto been considered to be 

 limited to the Niagara terrane of the American Silurian, and it has often 

 been the sole means of identifying that horizon. With the extended 

 ranse it is now known to have in the Ordovician fauna of Colorado we 

 can speak less confidently of the stratigraphic horizon identified by its 

 presence. In Wales and England it ranges from the Llandeilo through 

 the Bala or Caradoc. 



Tin'. Vertebrate Fauna. 



General Character. — The evidence of the existence of vertebrates at this 

 early epoch is limited to the plates and scales of ganoid fishes and what 

 appears to lie the ossified chorda] sheath of a fish allied to the recent 

 Chimzera. The latter correlation is based entirely upon the resemblance 

 between the fossil form and thecalcified chorda! sheath of Chimxra mon- 

 strosa. This resemblance is too striking to be passed over, although there 

 are certain differences thai render it of less value in classification than at 

 first appears. The Holoptychius-like scales and the Asterolepis-like plates 

 are their own interpreters and prove their connection with the lower 

 Devonian types with which they are compared. They are clearly the 

 diminutive. ancestral types of the greal fishes thai ;it a later date swarmed 

 in the Devonian sea and left their remains in the classic u Old Red 

 sandstone." 



* Quito recently I received from Professor F. II. Carpenter Haclureo nd Endo 



latum that were collected from a hand 6f limestone beneath tli it ol the Blael Hills 



of South Dakota, thus establishing unothi i owtposl in the Trontou 



