WILLIAMS] WILLIAMS. 133 



Oneonta rods and grays and (F 2) of the typical Catskill. So far as fossils have been 

 found there is difficulty in defining them. 



It is probable that Holoptychius and several allied fish, Conrad's Cypricardia cuujusta 

 and several plants, are found in both alike, but further investigation will be nec< 

 to establish any clearly distinctive characters in the fossils. 



The Waverly fauna G appears to be distinguished into three stages in other parts of 

 Ohio. But in the region comprised in these sections the stages are recognized more 

 by their lithologic than by their paleontologic characters. The general fauna may 

 be called the Syringothyris fauna, 



With G 1, the Bedford shale stage, 



With G 2, the Berea grit and sandstone, 



With G 3, the Cuyahoga shale and sandstone. 



H is the conglomerate (Olean and equivalent). 



J is the Barclay coal beds. 



From these studies the following principles of correlation were de- 

 duced : x 



First. The complications arising from both geological and geographical modifica- 

 tions rf fossil faunas are so great that the attempt to determine horizons by single 

 or by roughly identified fossils will certainly lead to erroneous results. 



Second. In classifying deposits in geologic surveys, it is of the greatest importance 

 that the actual altitude and the geographi c position of rock strata should be pre- 

 cisely defined, as well as the lithologic character of the strata themselves. And for 

 this* purpose some systematic and uniform nomenclature for the various kinds of 

 rocks should be made and adopted by all geologists in the country. 



Third. The fact that species composing the faunas and the total faunas themselves 

 are subject to constant modification, both geographical in the same horizon and geo 

 logical in the same area, is an element that paleontologists can not safely ignore. 

 These modifications, though they may be slight, can be easily recognized in the pas- 

 sage of 50 miles. 



Fourth. The actual order of faunas met with in a vertical section is not necessarily 

 expressive of biologic sequence, but signifies the sequence of the occupants of that 

 particular area. 



The change in the species from one stratum to the next may express the shifting 

 for miles of the actual inhabitants, and if the change, within a few feet of strata, is 

 to an entirely distinct group of species, the evidence should be taken as pointing to 

 a considerable shifting of condition of the bottom. If in such case each fauna is 

 kept distinct, the means of tracing the geographical distribution and modification 

 are at hand. If mingled, then the collection, though made at the same locality, will 

 only confuse. Two such faunas meet at O wego, Tioga County, in distinct strata, but 

 in rocks which are of similar lithologic character. One is a remnant of a prevailing 

 western fauna, the other is an eastern and late stage of a new fauna. 



Fifth. The classification of the rocks may receive local geographic names ; the 

 classification of the biologic series should receive names derived from the names of 

 species ; ages defined by families, periods by genera, and epochs by species, or some- 

 thing of that kind, and these periods or ages will always adjust themselves to future 

 discoveries. 



Begarding the classification of the particular formations the follow- 

 ing conclusions were reached, viz : 



(1) The Devonian black shales occur in the strata from the Genesee shales upward, 

 alternating with the normal deposits of the Portage and Cleveland shales and sand- 

 stones, and possibly higher, with modifications of the faunas, but run out at the 

 eastern extremity of the area. 



1 On the classification of the Upper Devonian, by Henry Shaler Williams, Am. Ass. Adv. Sci. Proo., 

 vol 34, 1885, pp. 232. 



