PRINCIPLES OF STRATIGRAPHIC CORRELATIONS 5l9 



confidence that the beds in New York, Kentucky, Tennessee, and Arkan- 

 sas, containing Tetradium cellulosum and certain equally characteristic 

 bryozoa, are referred to the same age and formation, namely, the upper 

 Lowville. Again, that beds of earliest Trenton and late Black River ages 

 in Iowa and Minnesota are exactly correlated with deposits in areas far to 

 the north and south on the ground of identity of their respective bryozoan 

 faunas, and scarcely less definitely with formations in the Baltic region of 

 Russia, because the latter contain over 60 species of bryozoa originally 

 described from Minnesota. Finally, this confidence is no less in the even 

 more extraordinary case of a Richmond coral zone, which is recognized 

 locally in America from Missouri to western Texas and New Mexico and 

 in a northerly direction to Alaska and Baffinland and in almost identical 

 development in the upper Lyckholm of Russia. 



"Stages of evolution" is used by some authorities, notably Hyatt, but as 

 a rule this class of criteria can not lead to very exact results, and at that is 

 usually too intricate to be generally applied. However, it is of excellent 

 service where the successive mutations marking the life history of long 

 enduring, broadly conceived species is concerned. The successive stages 

 of certain brachiopods, notably Plectambonites sericeus, DalmaneUa testu- 

 dinaria, and Leptcena rhomboidalis, which are usually very abundant and 

 widely distributed geographically, have been used with marked success. 



Stages indicating decadence of species or genera also have been used, 

 but as this condition may be due to purely local causes the interpretation 

 is liable to grave error. 



While the evidence of the fossils may in most cases suffice in establish- 

 ing a correlation, it is yet recommended that the organic evidence be 

 checked and supplemented by all physical evidence that may have a 

 bearing on the problem. 



CORRELATION BY LITHOLOQIG SIMILARITY 



Value of this method. — Obviously, correlation by lithologic similarity 

 is inferior in geographic extent of application, and perhaps also in general 

 reliability, to the other three methods implying more or less of deductive 

 reasoning, namely, by fossils, by criteria of active diastropliism, and by 

 evidence indicating slow, general sea transgression. Clearly, also, the 

 present method is closely connected with the other modes and more often 

 in need of their corroboration. Still, within reasonable limits, geographic 

 and otherwise, and when constantly checked by fossil and diastrophic evi- 

 dence, lithologic criteria are of great value in exact correlations. They 

 are especially serviceable and reliable when for various reasons it has be- 



