524 W. D. MATTHEW CORRELATION BY FOSSIL VERTEBRATES 



age found in littoral Miocene beds of New Jersey, and a Pleistocene 

 fauna said to be intercalated between Pliocene marine beds. In Texas 

 Doctor Dumble has recently discovered a Middle Miocene mammal fauna 

 in Pliocene marine beds. In California Doctor Merriam finds Middle or 

 Upper Miocene mammals in a Lower Miocene marine horizon. If the ma- 

 rine beds are to be taken as standard, then mammalian faunas are wholly 

 unreliable as evidence of age; for the evidence on the Atlantic and Gulf 

 coasts would tend to pull them up in the Tertiary succession, on the Pa- 

 cific coast to pull them down. It may indeed be supposed that the Pacific 

 coast mammal faunas, being nearer to the dominant center of dispersal 

 of the mammals, were precocious, and hence appear younger, and that 

 those of the Atlantic and Gulf coasts were primitive survivals, and hence 

 appear older than their true age. The alternative, assuming that the 

 marine correlations are accurately made, is that the Atlantic marine 

 fauna is precocious or the Pacific marine fauna retarded in the same way 

 as has been considered possible for the terrestrial faunas. But I cannot 

 find that the paleontologists dealing with marine faunas have considered 

 this point seriously; and the discrepancies will probably be reduced, if 

 not altogether removed, in most of the instances cited by a more thorough 

 study of the stratigraphic relations. 



In considering this subject of principles of correlation, I have taken 

 up a somewhat different aspect of the subject from the preceding speakers. 

 Doctor Ulrieh and Doctor Schuchert have considered especially the classi- 

 fication of the geologic record, and have laid stress on the value of the 

 great diastrophic movements, the shifting of strand-lines, and the charac- 

 teristic succession of changes in paleogeography as the fundamental basis 

 for this classification. Without question, we have all been much im- 

 pressed with the weight of the evidence and with the force of the reasons 

 they have presented for this view. While I cannot in this matter speak 

 for vertebrate paleontologists in general, yet many of us would be very 

 ready to agree that we have here, at least in theory, a sound basis for the 

 classification of the major divisions, systems, and periods. 



But the following discussion deals not with classification, but the cor- 

 relation of the sequences of strata in different regions in different parts of 

 the world. This is the preliminary work that must precede any general 

 classification. It can be accomplished only through the fossil record. 

 There are at present no other practical means of making the necessary 

 comparisons. The value of fossil vertebrates in this aspect of correlation 

 and the methods which are or should be employed in making use of their 

 evidence form the subject of the following discussion. 



