566 BAILEY WILLIS 



limit. Since that transgression took a long time, the eastern 

 Dakota sandstone is older than the western ; and as the advanc- 

 ing zone of sand accumulation, following the shore, was in turn 

 followed by a zone of shale deposit, the eastern sands were being 

 buried under Benton shale while the western sands gathered. 

 When the transgression ceased and Niobrara chalk was laid 

 down, the conditions were more nearly uniform throughout the 

 sea. Niobrara then probably marks approximately one and the 

 same epoch throughout its extent. Benton may do so in less 

 exact degree ; and Dakota corresponds at an eastern point to 

 the beginning, at a western to the closing, of the cycle of 

 deposits. 



It is convenient to conceive the time scale as marked by 

 horizontal lines. On such a scale the Dakota formation would 

 be represented b}^ a diagonal line, and the Niobrara by a nearly 

 horizontal line. 



The writer does not put these ideas forward as new; but in 

 defining a formation it is important to clear away certain mis- 

 leading conceptions that appear in the literature and in discus- 

 sion of the points at issue. Only at a particular place does a 

 formation belong to a definite age : when traced to another 

 locality it may be older or younger. 



FAUNAL UNITS. 



Classification by faunas. — The term faunal unit or individual 

 is here used as a parallel to lithologic individual, to designate a 

 set of strata characterized by a common fauna. The writer does 

 not undertake to say whether a fauna should be defined by 

 varieties, species, or genera ; by recognition of one or many 

 associated organisms. It may be granted that a fauna is some- 

 thing which each working paleontologist will define for himself 

 within certain broad limits, just as lithologic character is some- 

 thing which each stratigrapher defines for himself within certain 

 limits. But classification of strata by faunas is a different thing 

 from classification by formations. 



Professor Williams, in 1897, published a paper on dual 



