PRINCIPLES OF STKATIGRAPHIC CORRELATION'S 559 



Mexico. Beds of corresponding ages, moreover, are not seen to the west 

 of the Allegheny basin, except ( 1 ) in south central Ohio and east central 

 Kentucky, an area that was included in this Clinton sea, and (2) possibly 

 in northern Arkansas. The Arkansas deposits thought to be of middle; 

 Clinton age are included in and commonly constitute the whole of the 

 Saint Clair limestone. 



The Saint Clair fauna as now understood is three-fold, the Brasslield 

 at the base and a very different association at the top. The latter reminds 

 in some respects of the Osgood, in others of presumably later Xiagaran 

 faunas at Chicago and in Sweden. So far it has been seen only at 

 Marble, in eastern Oklahoma. Between these two faunas is another that 

 has become, through the studies of H. S. Williams, Stuart Weller and 

 Gilbert Van Ingen, the best known of the Saint Clair faunas. It con- 

 tains some Brassfield species, but with these a greater number of forms 

 that are either the same or closely allied to British and Bohemian 

 Silurian species. 



Strictly speaking, the term Saint Clair properly belongs only to the 

 fauna and beds of this middle member. As we have seen, tilting and 

 warping occurred between its time and that of the Brassfield so that the 

 deposits of the latter were laid down in embayments of Ozarkia that did 

 not lodge the second fauna. As to the upper fauna it is not seen in 

 Arkansas ; and the typical Saint Clair species have not been found in 

 Oklahoma. T see no other way to account for these differences in geo- 

 graphic distribution than by assuming intermediate time breaks in which' 

 differential oscillation occurred. 



The point of chief interest and significance in this connection is that 

 the typical Saint Clair fauna is very different from all typical Clinton 

 faunas; and no deposit is known in which intermingling of the two is 

 even suggested. It is concluded, therefore, that if the typical Saint 

 Clair is really of middle Clinton (Wolcott) age and not an intermediate 

 stage between the Williamson and the Eochester of the New York sec- 

 tion, then its sea must have remained wholly distinct from the typical 

 middle Clinton sea which occupied the Appalachian and Allegheny basins. 

 In the succeeding Eochester-Osgood age the Atlantic connections were 

 closed and Gulf of Mexico waters captured areas previously invaded 

 from the east. 



iVs intimated the post-Clinton Niagaran oscillations are interpreted 

 largely according to theory. My views, therefore, are correspondingly 

 tentative. The only points on which T am inclined to insist are (1) 

 that the dolomitic Niagaran deposits in Wisconsin were derived from 

 waters invading from the north and that the more calcareous and shalv 



