ART. 14 MICHIGAN TRAVEKSE GROUP POHL 29 



yet there is not a single f aunal association common to the eastern and 

 western areas of outcrop. The individual species, too, although 

 belonging in most cases to the same generic and family types, with 

 few exceptions, are distinct in the two regions. To my knowledge 

 not a single highly diagnostic species of any portion of either the 

 eastern or western sections has been found at any place in the other. 

 To be sure, a number of so-called " species " have been identified as 

 occurring alike in the two sections, but for correlation purposes 

 these invariably turn out to be utterly useless on critical examination. 

 Like " Atrypa reticularis " they may be " species " with long life 

 histories ; or like " Pholidostrophia iowensis " they, may have insuffi- 

 cient characters for useful discrimination; or with others like 

 " Phaoops rana^'' " Spii'ifer mucronatus^'^ " Stropheodoiita conca^a^^ 

 and dozens more, the name may be merely a jack-pot fed by constantly 

 increasing numbers of indifferentlj'^ and carelessly identified forms. 

 We are never to arrive at any exact knowledge of true conditions by 

 using these " species " for otherwise careful consideration of the 

 problem. 



FALLACIES OF CORRELATION AND EVIDENCES FOR PROPER 

 POSITION IN THE GENERAL TIME SCALE 



Previous expressions of opinion concerning correlation of the 

 Traverse beds has almost uniformly favored their time equivalence 

 with the " Hamilton " of New York State. That the " Hamilton " 

 of that State comprises several distinct stratigraphic units character- 

 ized by distinct faunal associations of different time values has been 

 only recently understood.^' The combined faunas of beds ranging 

 from the Marcellus to the TuUy were grouped and the result was a 

 " typical Hamilton fauna." From this enormous association similar- 

 ities were drawn with a comparatively few Traverse representatives. 

 Accounting for the absence of diagnostic Hamilton fossils of even 

 the " percentage scale " forms in the Traverse was either disregarded 

 or explained on the basis that westward migration for these species 

 (and in the same way the vast majority) was in some way hindered. 



The presence of forms unknown to the Middle Devonian of New 

 York and in a few cases grossly simulating remains from the Wis- 

 consin or Iowa areas were designated " western migrants " com- 

 mingling with species from the east in a shallow strait opening at the 

 same time to the Atlantic and Pacific. The correlation was thus 

 completed and the " Hamilton " extended to all points of the compass 

 over half a dozen physiographic provinces of North America. 



The conception of mid-Devonian paleogeography thus reviewed is, 

 of course, extreme, and individual workers have long known that this 



"A forthcoming publication by G. A. Cooper, of Yale University, contains the results 

 of a detailed revision of the New York sections. 



