REVISED PRINCIPLES OF CORRELATION 471 



Silurian system much higher in the time scale in the Ohio and Missis- 

 sippi valleys than it lies in New York and Pennsylvania. In the Eastern 

 States this boundary was properly drawn according to diastrophic cri- 

 teria. In the western areas the great Richmond overlap unconformity 

 marks the same basal limit of the Silurian. In America none of the other 

 breaks that have been suggested as serviceable in the case is so generally 

 and so definitely recognizable as the one at the base of the Richmond. 

 Paleontologists should acknowledge their error in placing the Richmond 

 fauna as older than the Oswego and Queenston sandstones of New York 

 and not persist in it to the detriment of the science by proposing to re- 

 move the unfossiliferous representatives of the Richmond in New York, 

 Pennsylvania, and elsewhere from the base of the Silurian to the top of 

 the Ordovician. The removal of the Lower Medina from the Silurian 

 would be a backward step. It would substitute indefiniteness in place of 

 defmiteness, because the base of the Lower Medina (including the Rich- 

 mond) is everywhere clearly determinable, whereas the correlates and 

 base of the Upper Medina are in many places uncertain. 



Like the Cambrian-Ordovician boundary of a few years ago, so also the 

 separation of the Silurian and Devonian in Appalachian areas containing 

 Cayugan and Helderbergian deposits remained entirely arbitrary until 

 the physical break between the two series was discovered. Even yet the 

 corresponding break in the New York sections remains to be pointed out. 

 In literature this boundary is placed differently at Manlius, Schoharie, 

 and Rondout, and at none of these places nor in New Jersey and Penn- 

 sylvania is it drawn precisely as in Maryland. Nevertheless, and despite 

 the considerable faunal transition, the same diastrophic break can and 

 has been recognized at every locality between southwestern Virginia and 

 New York where the concerned formations have been studied. 



According to prevailing practice, the Devonian-Mississippian boundary 

 is drawn variously in America. Individual authors and State surveys, 

 too, differ greatly in this matter. In New York it is fixed at the top of 

 the Chemung; in Ohio, Kentucky, and Tennessee at various horizons 

 between the tojo of the Chemung and the top of the Kinderhook; in Iowa 

 at the bottom, in the middle, or at the top of a gray shale which properly 

 constitutes the basal formation of the Kinderhookian series; in southern 

 Illinois and Missouri at some horizon in the Kinderhook above this shale. 

 In some places Mississippian fossils were found in the Upper "Devonian 

 black shale,"' and where these occurred the exact position of the systemic 

 boundary was left undecided. And yet if the break marked by the top 

 of the Chemung in New York, Pennsylvania, Ohio, and Virginia, and by 

 the base of the Kinderhook in the section at Burlington, Iowa, and at 



