INTERNATIONAL GEOLOGIC INSTITUTE 599. 



At present the general conviction prevails that there is but 

 a single trustworthy basis of correlation between separated, 

 regions, and that even this is subject to some important qualifi- 

 cations, and in many cases is wholly unavailable. I need not 

 mention fossil contents. But while fossils are, and apparently 

 must ever remain, the chief means of correlation, it has been the 

 growing conviction of my recent years that certain important 

 improvements and extensions of the criteria of correlation are 

 possible. These embrace (a) an improvement of the paleonto- 

 logic criteria by correcring them for the uncertainties and errors- 

 introduced by migration ; and (b) the addition of physical 

 criteria to the paleontologic ones in such a way as to eliminate 

 some of the uncertainties of the latter, and to serve in their 

 place when they are not available, as is so often the case. 



It is neither appropriate nor possible to set forth these 

 adequately here, but I should fail to duly magnify the impor- 

 tance of the measure herein advocated if I did not at least try 

 to indicate the greatness of the possibilities which we might 

 hope to realize if we had the means at our command. 



(a) The line of improvement in paleontologic correlation 

 may be best illustrated by a concrete example. Let it be sup- 

 posed that a provincial fauna arises in a harbor of refuge during- 

 a time of sea-withdrawal from the continental shelf 1 on the 

 border of America in a given stage A ; that by the subsequent, 

 development of an adequate sea-shelf or a connecting series of 

 epicontinental seas, this fauna migrated to the shores of Europe 

 during stage B, and that ultimately it reached the coast of Asia 

 in stage C. By the simple application of the criterion of com- 

 munity of species, stage C of Asia would be correlated with stage. 

 A of America, and although the time-interval might not be very 

 great, the error growing out of it might seriously disconcert the 

 cor.relation of associated physical events, and prevent their cor- 

 rect interpretation. But if during the stage A there developed, 

 in a harbor of refuge on the coast of Asia another provincial 



1 " A Systematic Source of Evolution of Provincial Faunas," Jour. Geol., Vol.. 

 VI, No. 6, Sept.-Oct. 1898, pp. 604-608. 



