Devonian System %n North America. 59 



the series is the same for each section, but the range may 

 vary ; in one area species may be restricted in range ; in an- 

 other, species may range through a long series of deposits. In 

 other words, species which are found to have a world-wide 

 distribution, although in one area they may be restricted to a 

 particular stage of the Devonian, are likely to have a long geo- 

 logical range in other areas, not less than from bottom to top 

 of some complete Devonian sections. But a particular combi- 

 nation of species, forming a characteristic fauna of a special 

 stage in one area, occurs at the same relative position in any 

 other area in which it appears. Such faunas are, however, 

 actually more or less local, and, as far as the Devonian is con- 

 cerned, it is not practicable to form more than three subdivi- 

 sions of the Devonian to which to apply universally a uniform 

 name. These three, in their general typical faunas, can be rec- 

 ognized (so far as they are present) in the different areas of 

 America and Europe, the lower, typically seen in the Cornifer- 

 ous limestone of New York ; the middle, represented in the 

 Hamilton of New York ; the upper, represented in the Che- 

 mung fauna of JSTew York. 



Any attempt to unify in the finer details is useless for 

 America, and, of course, would be useless if attempted for all 

 countries. 



3d. In the sections of America alone there is found nothing 

 in lithological composition or sequence which is uniform for 

 the several areas. 



In seeking uniformity of nomenclature the study of the 

 American Devonian leads to the following conclusions : 



(1) That uniformity is desirable in the names and prominent 

 distinctive biological characters of the so-called systems. 



(2) That valuable results may be reached by a discussion, on 

 the part of those acquainted with the same system in the dif- 

 ferent parts of the world, as to the best biological criteria for 

 marking the boundaries of the systems. 



(3) That while uniformity is possible in subdividing a sys- 

 tem into parts, the number of such parts, and the characters 

 distinguishing them, must be determined after a wide, compre- 

 hensive and minute study of their biological characters. 



(4) That preliminary work in classifying rocks should not 

 seek uniformity, but should adopt local nomenclature, and that 

 nomenclature based upon an exhaustive comparison of repre- 

 sentative sections can alone reach a uniformity that will be of 

 permanent value. 



