METHODS OF PALAEONTOLOGICAL INQUIRY. 6 1 



involves a geological problem, and consists in determining the 

 order of formation in time of the rocks which contain the fossils. 

 It is only of late years that the extreme importance of exact accu- 

 racy in this determination has become apparent. Formerly it 

 was regarded as sufficient if we could determine a fossil as Creta- 

 ceous or Eocene, but now we need to know its precise position 

 and range in the geological column. This is because we now 

 endeavor to trace out the phylogenies step by step through 

 every gradation, and it is only this humble, plodding, step-by- 

 step method upon which any dependence can be placed. Bril- 

 liant generalizations and bold hypotheses may be of great service, 

 but if they are to endure they must be verified in every particu- 

 lar by the more laborious but surer method. Darwin's motto, 

 " It 's dogged as does it," applies here in its fullest force. 



Not to scatter our energies in attempting hurriedly to survey 

 too wide a field, it will be well for us to confine our attention 

 entirely to the Tertiary period of geological history, and in 

 what I have to say I shall deal only with the mammals, the 

 principles being the same as for other vertebrates. It is the 

 good fortune of the American palaeontologist that in the west- 

 ern states and territories occurs an almost unbroken succession 

 of fresh-water formations, from the end of the Cretaceous 

 throughout the Tertiary period, and in these hardened sands 

 and clays has been preserved a marvellous wealth of the 

 remains of the successive faunas and floras which inhabited the 

 land. Not all of these formations have yet been investigated, 

 and hardly any of them have been mapped, but enough is 

 already known to provide an accurate outline of historical 

 events in the development of the continent. 



The sure method of determining the chronological succession 

 of strata is that of observing their order of superposition, a 

 method which is not always practicable, but which is infallible 

 wherever it can be applied, and which is simplicity itself. The 

 principle is merely that in a series of undisturbed strata the 

 order of superposition is the order of relative age, the oldest 

 being at the bottom and the newest at the top. Fortunately, 

 this sure and simple method is applicable to most of the west- 

 ern Tertiaries, and hence their chronological order may be 



