40 McGee — Southern Extension of Appomattox Formation. 



ditions. The formation has indeed been connected stratigraph- 

 ically between its most widely diverse phases and throughout a 

 considerable part of its area ; but the absolute identification of 

 the various parts of a formation so diverse in composition, so 

 vast in area, and so unequal in hypsographic distribution, is 

 rendered posssible and satisfactory only by the homogenic 

 method of correlation. 



The value of the method is still better illustrated by the 

 Appomattox : This formation is frequently buried beneath 

 newer deposits, and frequently and widely divided by erosion 

 over large areas, so that connection of the exposures by strati- 

 graphic continuity is impracticable. It is essentially unfos- 

 siliferous in the exposures thus far examined, so that paleon- 

 tologic correlation is impracticable ; and while its materials, 

 texture and structure are moderately constant, they are too 

 variable to warrant correlation by petrography alone. Yet it 

 is evident that the various parts of the formation are littoral 

 or sub-littoral ; that all represent temporary incursion of the 

 sea upon a long-sculptured land surface ; that all are affected 

 by the composition of the subterrane ; that all are affected by 

 the proximity of rivers flowing along the present water lines ; 

 that the materials everywhere comprise certain constant ele- 

 ments ; and that the structural relations of the formation are 

 essentially identical throughout its extent. So this formation, 

 like the Columbia, tells of a uniform general condition and of 

 certain easily discriminated local conditions ; and its various 

 parts may thus be confidently correlated by homogeny. The 

 formation has indeed been traced and connected as far as pos- 

 sible by stratigraphic continuity through thousands of expos- 

 ures ; but the isolated knobs projecting through newer 

 deposits and the isolated remnants left by erosion, and indeed 

 the regional developments of the orange-hued deposit could 

 never have been satisfactorily identified save by homogeny. 



Eventually the formation will be confidently correlated with 

 certain topographic stages displayed in the Piedmont and 

 Appalachian regions of the Southern Atlantic and Eastern 

 Gulf slopes ; but this correlation remains for fuller develop- 

 ment through future work. 



It should be pointed out that neither the Columbia nor the 

 Appomattox adequately illustrates the value of homogenic 

 correlation, by reason of their poverty in fossils. Thus far 

 paleontologic correlation has been based upon certain explicit 

 or implicit assumptions concerning the geographic distribution 

 of organisms and the relations between organisms and environ- 

 ment during past ages; i. e., it has been commonly assumed by 

 paleontologists that the geographic distribution of organisms 

 and their relations to environmental conditions in the past 

 were much the same as those of the present. But when the 



