METHODS OF CORRELATION 1139 



which are undoubtedly due to custatic movements of the sea, be- 

 comes more and more apparent. The widespread mid-Jurassic 

 transgression of the sea over Europe is well known and the discon- 

 formity (and occasional unconformity) produced by this trans- 

 gression has been used for widespread correlation. The great 

 Mid-Ordovicic hiatus first observed in North America. (Grabau- 

 16; 18), and the similarly widespread hiatus in the Upper Ordo- 

 vicic (Weller-52), are now known to be marked in North Europe 

 as well (Bassler-5). In like manner the j\Iid-Siluric hiatus and 

 disconformity so widespread in North America appear also to be 

 present in the Baltic region of Russia and Sweden and in the 

 Bohemian Palaeozoic district. The probabilities are that in Mid- 

 Siluric time the sea left a large part of the present land area dry. 

 Similar widespread disconformities are recognized in the Mesozoic. 



6. Correlation by regional metamorphism. 



Regional metamorphism has already been defined as an altera- 

 tion or metamorphism, which affects extensive regions and which is 

 primarily due to tectonic disturbances. Such metamorphism may, 

 of course, occur at any time in the history of the earth, but when- 

 ever it does occur it will affect all the formations of the region in 

 which it takes place, though, obviously, some formations may be 

 more strongly affected than others. This being the case, it follows 

 that, wherever unaltered rocks overlie the metamorphosed ones, 

 the age of the former cannot date back of the period of meta- 

 morphism, and that, hence, the lower limit of their age is fixed 

 by this period of metamorphism. Evidently there is no guarantee 

 here, however, that the strata of the overlying series are all of 

 the same age, though within moderate limits this is probably true. 

 One general rule may, perhaps, be formulated, and within certain 

 limits applied, namely, that, of two formations in contact, the more 

 strongly metamorphosed one is the older. Here, however, the same 

 caution is necessary that is required in applying the rule of greater 

 deformation to two deformed formations in contact. Some forma- 

 tions are more subject to metamorphism than others, just as some 

 formations are more subject to deformation. 



The method of correlation by iiietamorphism is, perhaps, the 

 most applicable to the determination of the boundary line between 

 the pre-Palseozoic and later formations, though even here it seems 

 not always to be reliable. This would appear from the fact that 

 extensive sedimentary formations, such as the Belt series of Mon- 



