PRINCIPLES 519 



characteristic and the identification for this and other reasons more or 

 less exact and reliable. The old, and still usual, method of correlation is 

 to compare lists of faunas and to judge by the number of common species 

 how nearly the horizons correspond. It is not the best method, indeed; 

 but it does yield results that agree fairly well with the more reliable re- 

 sults of comparisons of geologic range of each species. But the method is 

 either misleading or perfectly worthless with fossil vertebrates; for most 

 of the faunal list depends on identifications of fragmentary material: 

 some may be, nevertheless, exact; most are essentially provisional. The 

 evidence of one complete skull or skeleton may well outweigh that of the 

 entire remaining list. And, on the other hand, a single very character- 

 istic tooth may be of more value than much more complete and abundant 

 material. 



In correlating marine formations vertebrate fossils are of compara- 

 tively limited use, owing to their rarity. But some of them, especially the 

 wide-ranging marine reptiles, when they can be found, are of high value 

 in comparing distant regions, owing to their wide geographic range. But 

 here, again, we must have adequate material if we desire precision in our 

 results. In the correlation of the upper chalk of Europe and North 

 America the Mosasaurs afford evidence of high value and precision, be- 

 cause they are completely known and have been carefully studied. Their 

 evidence in correlation has not indeed been fully studied, but I can say 

 with confidence that it is an important check, if no more, on the evidence 

 of the invertebrates. I have less confidence in the accuracy of results 

 obtainable from the more fragmentary Cretaceous Mosasaurs of New 

 Zealand, Australia, and South America. If the complete skulls and skele- 

 tons of these southern genera were known and carefully studied, it would 

 probably enable us to correlate the Cretaceous formations of the southern 

 continents with more certainty than we can at present. But identifica- 

 tions at present current are probably in need of revision. They are not in 

 accord with the invertebrate evidence and not positive enough to oppose 

 to it. 



Tertiary marine mammals would, if more numerous and better known, 

 afford very valuable evidence. But until we know more of their evolution 

 and geologic history I am indisposed to give much weight to them. Nor 

 do I think that enough is known about Mesozoic mammals for their evi- 

 dence to be of any serious importance in correlation of the Triassic, Ju- 

 rassic, and Cretaceous horizons. The Mesozoic land Eeptilia, on the other 

 hand, are already of high value and will be of much greater weight in the 

 near future in correlating terrestrial formations of this era. 



It is chiefly in dealing with terrestrial Tertiary formations that the 



