516 W. D. MATTHEW — ^CORRELATION BY FOSSIL VERTEBRATES 



been wiped out of existence by a universal revolution. Such was the 

 testimony of the rocks as read in those clays. Such was the belief in 

 which geologic correlations were made. With this viewpoint correlation 

 was a very simple and quite exact science. You were dealing with one 

 distinct fauna or with another. Intermediates, intermixtures, survivals, 

 were not thought possible. Each formation, each fauna, was clean cut, 

 distinct, separate. A sample was sufficient to identify it. A very frag- 

 mentary record was amply sufficient to indicate which of the successive 

 distinct fossil faunas you were dealing with. Each stratigraphic unit, 

 each faunal unit, had its definite place in geologic time. Correlation was 

 a rather simple matter and involved merely a sufficient knowledge of each 

 fauna to recognize it when you saw a sample of it. 



Under the influence of these views the succession and distribution of 

 geologic formations and the sequence of faunas was worked out in detail 

 in western Europe and more sketchily in the eastern United States; the 

 foundations of our systems and nomenclature were laid. When the doc- 

 trines of evolution were first promulgated, the geologic record, interpreted 

 and recorded in terms of the older theories, seemed to furnish the strong- 

 est and most direct argument against the new views. Today it is recog- 

 nized as affording the most direct proof, the most unshakable evidence in 

 their favor. Yet the facts stand substantially the same ; we have reinter- 

 preted their meaning, broadened the scope of our data, and realized the 

 vast complexity of the life processes of which they furnish the historical 

 record. 



In practice, however, we are still controlled more than we realize by 

 the empiric methods that grew up under the older viewpoint. In theory 

 we believe in continuity; in practice we adopt correlations and map 

 strata as though catastrophism were the rule, as though the gaps in the 

 record were universal and correspondent throughout the world. 



These preliminaries may serve as an excuse for the admission that 

 theory and practice are not altogether in accord in vertebrate correlation 

 work. The practice has been, and still is in large part, to use the ordi- 

 nary current methods of stratigraphic geology, to draw up and compare 

 lists of faunas as complete as possible, and to judge of the relations of 

 the horizons by the degree of correspondence of these faunal lists. Yet 

 the work of leading paleontologists in recent years, while it may retain 

 this form, shows in its conclusions a clear recognition of the reservations 

 and modifications necessary in applying such methods. 



Fossil vertebrates have certain obvious advantages for correlation as 

 compared with invertebrates and plants. They have Other disadvantages 

 equally manifest. These advantages and disadvantages involve certain 



