SUBDIVISIONS IN GEOLOGICAL HISTORY. 405 



2. Subordinate divisions should recognize the same criterion, but should 

 depend for their limits, as far as practicable, on physical breaks or events 

 registered in the rock-series, and on abrupt transitions in kinds or groups of 

 fossils. Since the latter are dependent on physical changes, they are a con- 

 venient criterion when characterizing large areas. 



3. When subordinate divisions of the higher grades have been estab- 

 lished on any continent, or part of a continent, these divisions should be 

 recognized and adopted as nearly as possible in the study of other regions, 

 and their limits determined if possible by means of the fossils ; for only in 

 this way can the history of different regions be brought together into one sys- 

 tem. For example : the Permian period, recognized and defined in European 

 geological history, should have its place in American geological history, 

 however intimately the beds and their fossils in America may blend with 

 those of the Carboniferous period. So also the Devonian of Europe should 

 be recognized and have like limits, as nearly as may be, in the Devonian 

 of America. A degree of fixedness in the higher subdivisions and their 

 names is necessary to prevent confusion in the literature of the science 

 and the frustration of its great purpose, — the production of a comprehensive 

 earth-history. 



4. Inferior subordinate divisions so far depend on local conditions, that 

 those of different continents, and even of distant parts of the same continent, 

 generally require, in the first study of a region, special designations to avoid 

 assumptions of closer relationships or equivalency than can be made out. 

 The different continents, and often also unlike regions of the same continent, 

 have had their special histories. The periods and epochs of America and 

 Europe are not in general the same in limits, and much less so in rocks. The 

 Devonian subdivisions are different on the two continents ; and it is far from 

 certain, also, that the commencement assigned to the Devonian in North 

 America is synchronous with that in Europe. In the Carboniferous, Rep- 

 tilian, and Mammalian eras the American epochs differ from the European. 

 There is much diversity between the subdivisions in New York and those of 

 the Mississippi valley, and still greater between these and the subdivisions 

 of the Pacific slope and border. Even in Pennsylvania the formations fail 

 of many of the subdivisions that are prominent in New York. 



Hence in the study of a new region it is necessary at the outset to make 

 arbitrary subdivisions of its formations, such as may seem most convenient 

 and natural, and give them local names. These names have at first only a 

 note-book value. When the relations of the beds to those recognized in 

 other regions have been ascertained through fossils, the facts begin to take 

 their places in the general geological history of the country ; and should the 

 correlation be complete, the local names may give way to those generally 

 accepted elsewhere. 



It is of the highest importance to remember that state boundaries are 

 only political limits, and not, ordinarily, at least in America, true geographi- 

 cal or geological limits ; and if the subdivisions of one state Avhich have 



