PRINCIPLES or STRATIGRAPHIC CORRELATIONS 533 



but he seems to underrate and misapprehend the difficulties when he 

 adds "for correlation by cosmopolitan faunas, the very best of faunas for 

 the purpose, is added to the physical correlation." Aside from the 

 pelagic faunas, and using the term in not too broad a sense, is there, or 

 has there ever been, a truly cosmopolitan fauna? For reasons given 

 on page 366, the forms comprising the so-called cosmopolitan faunas, 

 instead of being contemporaneous, are more probably either later or 

 earlier stages of slowly modifying species, and, consequently, of little 

 value in exact correlation. 



Except those afforded by pelagic and free swimming species, there are 

 no highways, no "short-cuts" to world-wide correlations. On the con- 

 trary, stratigraphic correlation, whether we use physical or biological 

 criteria, or both, is a slow, laborious process, each method requiring con- 

 stant corroboration by the other methods, with no certainty of results 

 until we have closed the circle and proved the process. As we proceed 

 from a given locality, the organic and the lithologic evidence first de- 

 pended upon grows ever weaker until one or both pass out entirely. 

 Fortunately, the continuance of diastrophic lines frequently enables us 

 to mend the broken thread; and herein diastrophism proves itself not 

 only the theoretic basis of correlation, but its criteria show themselves 

 to be eminently practical. 



Though fully endorsing Chamberlin and Willis's claims for diastro- 

 phism as the ultimate basis of correlation, and though affirming the 

 exceeding value of diastrophic criteria when abundantly checked and 

 corroborated by paleontological and lithological evidence, I must yet 

 frankly admit that by themselves these criteria are utterly valueless in 

 correlating widely disconnected outcrops. Diastrophic criteria are so 

 variable in local expression, and their successive manifestations so similar 

 in general aspect and range of variability that positive and unerring 

 discrimination seems impossible. We may determine the position of 

 certain stratigraphic boundaries, as between bodies of limestone and 

 shale or sandstone, or we may note and trace some persistent zone of 

 overlap and possible conglomerate, but what these lines mean in the 

 time scale may, as a rule, be determined only by fossils. Moreover, it 

 has frequently happened that the most important boundaries were en- 

 tirely overlooked when the geologist took into account solely the obvious 

 changes in character of deposits. That the supposed lithological unit 

 included one or more great discontinuities of sedimentation, or that the 

 formation spanned two or more systems and thus made no provision for 

 the discrimination of important, elsewhere perhaps more clearly defined, 

 geological events, was either unknown or not considered so long as the 



