OSCILLATORY CHARACTER OF CONTINENTAL SEAS 331 



continental seas were as a rule comparatively limited in extent and, fur- 

 ther, their waters were repeatedly and more or less completely withdrawn 

 into the great oceanic basins. The evidence on which these statements 

 are based may be gathered from numerous observations recorded in this 

 work. It has not seemed worth the space to repeat it here except such 

 parts as bear directly on the principles under discussion. 



Detailed paleontologic and stratigraphic investigations relating to the 

 hitherto supposed long enduring general encroachments of the Paleozoic 

 seas is proving that these were interrupted, and the direction of the move- 

 ment of the strandline reversed, at certain times that seem to have been 

 essentially contemporaneous for the whole world. These general reversals, 

 resulting, as I believe, usually in very great if not complete emergence of 

 the continents, occurred at least at the close of each of the great periods. 

 Indeed, the evidence is so abundant and convincing that I can not doubt 

 that not only the systemic boundaries, but also those of series and perhaps 

 of groups can be drawn in Europe as in America at actually correspond- 

 ing breaks in the marine sedimentary record. There is no question as to 

 the gaps in the record — experience has shown that their detection is 

 merely a matter of intelligent search in good exposures. The only grave 

 difficulty lies in their proper correlation and in the determination of their 

 respective taxonomic values. 



Of course, little that is final can be done along these lines if "matching" 

 of faunas alone is relied on. Since learning that the faunas of the sev- 

 eral oceanic basins are very distinct, and that each may continue with 

 little modification through long periods, as has been shown in discussing 

 recurrent species and faunas, also that the great changes in local succes- 

 sions of fossil faunas is more commonly due to overlapping of oceanically 

 distinct faunas than to evolution of the preceding fauna, simple matching 

 of species and genera has become but a small, and by itself a far from 

 competent factor in a modern paleontological correlation. As the pre- 

 vailing classification of formations in America and Europe is based almost 

 entirely on this inadequate old method, and as the faunal criteria have 

 not been consistently employed in all cases, some changes in definition of 

 systems and series is to be expected. It is highly probable, too, that some 

 of the most important boundaries have been overlooked, or if observed 

 that their significance has been misinterpreted and underrated. Properly 

 determined intercontinental correlations therefore are likely to result in 

 some hitherto unsuspected equivalences, and the surprises are scarcely 

 less when the correlations concern provincially distinct formations on the 

 same continent. Some of the latter are brought out in the correlation 

 charts of Part III. 



