320 REPORT OF NATIONAL MUSEUM, 1892. 



to are now separated from one another by intervening broad bodies of 

 water beneath which there may or may not have ever been strati- 

 graphieal continuity, or by terrestrial conditions which have made it 

 impossible to truce such continuity if it ever existed. 



Thus, while correlation in this restricted sense is usually confined to 

 a comparison of great series of stratified rocks as they occur on sepa- 

 rate continents, it is sometimes quite as applicable to different parts of 

 one and the same continent as is the case, for example, with the east- 

 ern and western portions of North America. 



In considering the subject of correlation as thus restricted, we iind 

 that it not only depends quite as fully upon the study of fossil remains 

 as does the identification of formations, but having reference to regions 

 too distinctly separate from one another to have recognizable strati- 

 graphical continuity between them, such questions as those pertaining 

 to physical characteristics and identity are eliminated. We also find 

 that the manner of dependence upon fossil remains in the two classes 

 of cases is different; that is, in the one case it is mainly specific iden- 

 tification that is relied upon, and in the other reliance is placed only 

 upon the recognition of various general fauna! ami floral types, such as 

 are referred to on pages 290-300. 



As was stated in Essay n, formations being physical units of local, 

 but not universal, stratigraphic classification, the term identification is 

 especially appropriate when referring to studies of their relation to one 

 another in the held. The determination of correlation is also in some 

 sense an act of identification; that is, it is the identification, or, more 

 properly speaking, the recognition of the divisions or subdivisions of 

 the great geological scale in different and distinctly separate parts of 

 the world. The fact that the divisions and subdivisions of the scale 

 may thus be more or less completely recognized being fully admitted 

 by all geologists, the only question that need be discussed in that con- 

 nection relates to the manner of their characterization, the distinctness 

 of their limitation, the completeness of their representation, and to 

 synchronism or contemporaneity and liomotaxy. 



The idea of correlation presupposes a standard, and as every standard 

 must be either absolute or conventional it is uecessary to consider what 

 must be the character of one by which correlation as herein defined 

 maybe recognized, because upon this depends an intelligible discussion 

 of the subject. If the standard of correlation is an absolute one, there 

 can be no question as to the dehniteness and completeness of its appli- 

 cability in all cases and in all parts of the world where the necessary 

 observations can be made. If, however, the standard is a conventional 

 one, its recognition as such implies more or less uncertainty as to the 

 definiteness and completeness of its universal applicability, but a con- 

 ventional standard may be based upon such an array of admissible facts 

 that in all cases where those facts are recognizable uncertainty and 

 indetiniteness in its practical application are reduced to comparatively 



