REVISED PRINCIPLES OF CORRELATION 467 



mulating facts are unavoidable. However, if the science is not to die of 

 dry-rot, but is to live and grow, we should welcome new ideas and cheer- 

 fully accept the modifications in ©ur own method and practice demanded 

 by the march of progress. No good can come from standing aloof, nor 

 from setting ourselves up as defenders of the prerogatives of paleontology 

 in stratigraphic correlation. The use and high value of fossils in corre- 

 lation need no defense; but they do need to be accurately and minutely 

 discriminated. Paleontologists should also be more careful than hereto- 

 fore in avoiding the unconscious bias of previous opinions. In studying 

 the fossils of the Ohio shale, for instance, we can not hope to reach the 

 truth when their Devonian alliances only are pointed out, while their 

 much stronger Mississippian affinities are wholly ignored. We must look 

 at things as they are and from both sides, without regard to the dictates 

 of previous training and conviction. 



Again, if we consider the varying methods that have operated more or 

 less independently in building up the present classification of sedimentary 

 rocks, incongruous results are to be expected, for if we do not agree in 

 methods our conclusions must necessarily differ in corresponding degree. 

 And yet the arguments on the various sides may be, up to a certain point, 

 entirely logical. Each contestant may be right from his viewpoint, and 

 each may have excellent precedents for his line of reasoning. 



But this does not help us to a systematic classification of geological 

 formations. That desirable end is possible only under agreement, and 

 the agreement must be on the matter of method, rigorously and consist- 

 ently applied throughout the column. 



The element of consistency in the application of the adopted methods 

 is quite as important as any other quality, for without it a really scientific 

 classification of the sedimentary rocks, and thus of the geological ages 

 which they represent, is impossible. It should pertain (1) to the criteria 

 which shall determine why and where stratigraphic boundaries of what- 

 ever grade should be drawn, and (2) to those which shall determine which 

 combination of units is to be ranked as a group, which as a series, and 

 which as a system. In my opinion, diastrophism, in its broadest sense, 

 affords the only means of finally attaining a reasonably accurate and sys- 

 tematically constructed classification. 



THE ULTIMATE BASIS OF CLASSIFICATION 



Now, how shall we proceed in working out the desired end ? The first 



step is the selection of some principle that shall guide us in determining 



when a geologic age has ended and a new age has begun. Obviously, such 



a principle should apply similarly to all the divisions of the geologic time 



XXXIV— Bull. Gbol. Soc. Am., Vol. 27, 1915 



