472 E. O. ULRICH CORRELATION OF THE STRAND-LINE 



other places in the Mississippi Valley is accepted as separating the de- 

 posits of the Devonian period from those of the Mississippian, we can 

 draw, usually with little difficulty, the same diastrophic boundary wher- 

 ever marine sediments of these two systems are in contact in southeastern 

 North America. Judging from my own experience, there are no transi- 

 tion deposits indicating continuity of marine sedimentation from one 

 geologic period to the next in any area now exposed to view. 



The cited instances are of highly important stratigraphic boundaries. 

 But equally decided improvements have rewarded our efforts in revising 

 the less important but more numerous intervening formational bound- 

 aries. The more we recognize the oscillatory character of marine depo- 

 sition in the continental basins the more accurately we learn to correlate 

 and the more exactly we mark the boundaries of stratigraphic units, 

 whether they be of high or of low rank. 



In my Eevision of the Paleozoic Systems I sought to define and apply 

 practically the principles of the diastrophic method. Because of misin- 

 formation, at other times lack of important data, the application of the 

 principles doubtless has resulted in some — I may say even many — un- 

 happy arrangements. Knowing some of these 'imperfections and expect- 

 ing to find more, I am prepared to admit the appropriateness of a recent 

 facetious paraphrasing of the title, according to which Ulrich's Revision 

 became Ulrich's "Rawvision." It is rather "raw" in spots ; but snbe- 

 quent experience has shown that the fault does not lie in the principles. 

 The errors occur chiefly in those places where the desire to present a 

 complete scheme necessitated undue haste in their application. 



Now, let us consider certain commonly entertained conceptions — all 

 bearing more or less directly on correlation by fossils — that seem to me 

 in need of revision. 



CORRELATION BY "MATCHING" OF FOSSIL CONTENTS 



Correlation by similarity of general faunal aspect has been perhaps 

 the most favored of methods for determining the relative ages of fossilif- 

 erous deposits. lists of the fossils are compared and the percentage of 

 identical and so-called "representative species" is supposed to indicate 

 the time relations of the two or more formations. Under certain condi- 

 tions, and in so far as broad stratigraphic conceptions are concerned, the 

 idea is correct enough ; but when it comes to the accurate identification 

 of minor units of the time scale it is inadequate and indeed more likely 

 to lead us astray than to the truth. 



All, or at least most of us, I believe, accept organic evolution as an 

 established fact. On the whole, too, it will be admitted that the evolu- 



