254 REPORT ON THE STATE CABINET. 



and although it was shown that the graptolites were all of different 

 species from those previously described, yet they appeared to ofi'er only 

 corroborative evidence in support of the previousl}^ entertained opinion 

 regarding the age of the strata. It was only at a later period, and from 

 the discovery of numerous other fossils in the same formation, some of 

 them having a primordial aspect, that its higher antiquity was suspected. 



The shales of Point Levis, with their associated limestones and sand- 

 stones, since termed the Quebec group, are now regarded as embracing 

 the period from the Calciferous sandstone to the Chazy limestone, inclu- 

 sive. This epoch, therefore, is entirely anterior to that of the Hudson 

 river formation, and a careful comparison of all the species of graptolites 

 has shown that no identical species occur in the two series of rocks. 



In the present state of our knowledge, we recognize the Graptolitidse 

 as beginning their existence at the period of the Potsdam sandstone. 

 The greatest development of the family, both in genera and species, is 

 found to be at the epoch of the Quebec group. Several genera and a 

 few species are known in the Trenton formation ; and a greater develop- 

 ment, embracing most of the genera and many species, occur at the 

 period of the Hudson river formation in Canada and the United States. 

 In the Clinton strata we have a single species of Graptolithus and a 

 Retiolites ; while Digtyonema and Inocaulis occur in the Niagara beds. 

 In all the subsequent geological formations we have found no true grap- 

 tolites, and the only representatives of the family consist of fragments 

 of Digtyonema, belonging to a few species. These occur in the Upper 

 Helderberg and Hamilton formations, above which we do not yet know 

 a species of any genus referable to this family of fossils. The Grapto- 

 lites proper, therefore, have their upper limit in the shales of the Clinton 

 formation, and all other genera of the family, except Digtyonema, are 

 restricted to the Silurian system. 



The geographical distribution of the graptolites is not in all respects 

 coincident with the extent of the geological formation. Dendbograptus 

 occurs in the Potsdam sandstone of the St. Croix valley ; but neither 

 this nor any other graptolite is known in other localities of the sandstone, 

 so far as I am aware. The species of the Quebec group, numbering 

 more than all the other formations together, have been identified for a 

 longitudinal extent of about 900 miles; Point Levis, Orleans Island, St. 

 Anne's River (Gaspe), and the western part of Newfoundland, being 

 the principal localities. But although the Quebec group is known to 

 extend into Vermont and along the eastern counties of New York, I am 



