STRATIGRAPHIC CLASSIFICATION PALEONTOLOGIC CRITERIA 493 



is conspicuous chiefly because of the new generic and specific types that 

 have been introduced in the meantime from other centers of origin and 

 distribution. Now, if we look up the ancestors of the Spergen species, 

 are they not plainly recognizable in the Hamilton and Onondaga faunas ? 

 It was a long time between the Hamilton and the Spergen, but the bryo- 

 zoa, the brachiopods, and the gastropods in these two formations are sim- 

 ilar in general aspect and, indeed, include many closely allied species. 

 The principal difference lies in the species and genera which passed out of 

 existence before the Spergen began. Certainly the two faunas are more 

 closely allied than is the Hamilton to the typical N'eodevonian faunas in 

 New York. 



The Eopaleozoic faunas traced to the northern middle Atlantic basin 

 afford some striking instances of slow modification. Evidently this was 

 the center of origin of certain Cambrian trilobites and brachiopods whose 

 descendants are unmistakably recognized in faunas that spread from this 

 basin at various times from the middle Cambrian on to the close of the 

 Ordovician, Indeed, one of the trilobite genera, Agnostus, maintained 

 apparently uninterrupted existence in this basin to the close of the Ordo- 

 vician. The same may be said of certain brachiopods, like Acrotreta and 

 Lingulella. Species of these three genera are found associated in beds 

 belonging well up in the Ordovician in the Appalachian Valley. In the 

 Pacific and Arctic faunas, however, so far as known in America, these 

 genera became extinct with the close of the Cambrian. Remopleurides, 

 which is a derivative of the preceding Holmia and Paradoxides, and 

 Triarthrus, evidently descended from some relative of Ptychoparia and 

 Olenus, are almost constantly met in middle and later Ordovician trans- 

 gressions of this Atlantic fauna. Although of great value in identifying 

 the source of the invasion, these slowly modifying and generally prolific 

 species and genera have been detrimental rather than helpful in exact 

 age determinations. Triarthrus bechi, for instance, may be found in any 

 of the shaly beds of the Atlantic Ordovician from the beginning of the 

 Trenton to middle Cincinnatian, while Agnostus and Acrotreta are no 

 more Cambrian than Ordovician in this basin. It is the new life accom- 

 panying the recurrences of these long-lived types that really counts the 

 ages. 



A fair conception of the progress of geologic fauna! evolution may be 

 gained from detailed comparisons of faunas in areas repeatedly invaded 

 from the same oceanic basin. The purity of the faunal aggregates which 

 have originated and developed in these permanent basins is, of course, 

 being continually, and at times has been very greatly, vitiated by foreign 

 accessions; but during the Paleozoic, at least, the integrity of each was 



