walcott.] CORRELATION. 421 



western North American sections of the Wasatch, Nevada, Highland 

 Eange, and Eureka are more or less defective, owing to the imperfect 

 interpretation of the paleontologic data. 



The next essay upon the correlation and comparison of the European 

 formations with those of America by Mr. Matthew is entitled: "How is 

 the Cambrian divided? — A plea for the classification of Salter and 

 Hicks." * He institutes comparisons of the strata in Wales, Newfound- 

 laud, and Norway, and of the characteristic trilobitic genera of each 

 horizon. From the results obtained by these comparisons he considers 

 that the Cambrian is to be divided into the Upper and Lower, and that 

 the method proposed by Mr. 0. D. Walcott, of separating the Olenellus 

 or Lower Cambrian from the Paradoxides or Upper Cambrian, is not 

 justified by the evidence afforded by the American aud European sec- 

 tions. 



METHODS OF CORRELATION. 



I. Superposition. 

 II. Organic Remains. 



III. Lithologic Characters. 



IV. Unconformity. 

 V. Miscellaneous. 



I. SUPERPOSITION. 



(a) The order of superposition of strata is the foundation of geologic 

 chronology. By its aid the stratigrapiiic succession and serial relations 

 of the fossil faunas are first determined. Undisturbed superposition 

 within a section is an absolute test of relative age of the parts of that 

 section; it is not positive evidence of the age of the various beds in 

 relation to beds sustaining the same serial relations in some discon- 

 nected section without other tests to support it, even though they 

 occur in the same geologic province : e. g., a band of sandstone beneath 

 a limestone and above a shale may be identified in one portion of a 

 geologic province; but the occurrence of a band of sandstone beneath 

 a limestone and above a shale, in a distant portion of the same province, 

 is not positive evidence that the two bands of sandstone are synchro- 

 nous or portions of one continuous formation, unless their continuity 

 can be traced. 



(b) Superposition is not a test of identity of formation except in a 

 limited basin, where the geologist has evidence to show that the forma- 

 tion is practically continuous in its stratigraphic relations. In widely 

 separated portions of the same basin, or of different basins, the order 

 of superposition must be sustained by other tests before it can be as- 

 serted positively that the formations thus correlated are identical. 



(c) In the same geographic or geologic province, fresh-water and 

 marine deposits are not to be correlated with each other, even if they 



>Am. Geologist, vol. 4,1889, pp. 139-148. 



