216 C. S. Prosser — Upper Hamilton and 



cations which when observed, enabled one to separate them. 

 There are frequently lithologic variations to assist, as in eastern 

 Pennsylvania, which were used to advantage in the work of 

 Prof. I. C. White for the 2d Geological Survey of that State, 

 and a knowledge of these two characters was found necessary 

 for the correlation of the Upper Devonian of this region. This 

 principle was tested several times by different observers and 

 it was found that those paleontologists who had simply studied 

 the species in a general way without considering the minor 

 variations due to the difference in geographic distribution or 

 stratigraphic range did not recognize the different facies of 

 these faunas but called them all Hamilton ; while two paleon- 

 tologists who had closely studied the characters of the faunas 

 and had also noted similar comparatively slight changes in 

 reference to the distribution and range of species for another 

 region, on being shown collections of these fossils without any 

 information as to their stratigraphic position, invariably recog- 

 nized those that came from the beds corresponding to the 

 typical Hamilton of New York, as well as the slightly modified 

 faunas from the overlying rocks. And this recalls the princi- 

 ple stated by Dr. H. S. Williams when describing the classifi- 

 cation of the Upper Devonian for southern and western New 

 York, viz: — " the complications arising from both geological 

 and geographical modifications of fossil faunas are so great that 

 the attempt to determine horizons by single or by roughly 

 identified fossils will certainly lead to erroneous results."* 

 The failure on the part of paleontologists to observe this rule 

 and the tendency on the part of the stratigraphic geologists 

 to base their correlations mainly on lithologic characters have 

 continually made futile all attempts toward an accurate corre- 

 lation of the terranes of the Upper Devonian for southern and 

 eastern New York, and eastern Pennsylvania. 



In a recent article is the statement that " the basal beds [of 

 the Oneonta formation J are gray flags which merge into the 

 Hamilton ;"f but no data is given in favor of such correlation 

 and I suppose the opinions of other geologists were followed. 



This particular part of the series is one to which the writer 

 of this paper has given especial attention, not only so far as 

 the fossils from the different regions and horizons are concerned 

 but he has also made cross sections at various points from 



* Proc. Am Assoc. Adv. Sci , vol. xxxiv, 1886, p. 231. 



fThe stratigraphical relations of the Oneonta and Chemung formations in 

 eastern Central New York; by N. H. Darton. This Journal. Ill, vol. zlv, 

 p. 206. In addition see on p. 207 the statement that the Hamilton is exposed at 

 Lyon Creek bridge, near Oxford; also, that "The Chemung rocks to which 

 Mather and others refer lies below the Oneonta beds or about 1.000 feet below 

 the actual base of the Chemung horizon, and are Hamilton in position. This 

 fauna is meagre and consists of species supposed by Yanuxem to be Chemung in 

 central New York, but now known to be Hamilton " 



