PRINCIPLES OF STKATIGKAPJIIC CORRELATIONS 543 



irregular shifting of continental seas. The evidence on which this 

 principle is postulated is unusually abundant and when fully published 

 it can not fail to be convincing. Much of it is described on pages 411 to 

 430. This discussion, however, is of necessity largely confined to gen- 

 eralized statements, the detailed observations on which many of the 

 asserted correlations are based being reserved for publication elsewhere. 

 (15) Correlation by probabilities depending on rhythm of move- 

 ments. — Breaks in local stratigraphic sequences having been establisheJ 

 the time relations of the hiatuses in different areas, as in adjacent struc- 

 tural basins or in distinct faunal provinces, are usually determined by 

 fossil evidence. However, often this evidence is inconclusive and some- 

 times even misleading, and provisional conclusions pending discovery 

 of localities where such relations are positively manifested are frequently 

 desirable. In such cases we may correlate on the basis of probable 

 changes in the attitude of continental basins with respect to sealevel 

 during the rhythmic course of diastrophic movements. When the crude 

 and regionally varying rhythm has been sufficiently ascertained it seems 

 possible to predict which trough or troughs in the Appalachian Valley, 

 for instance, are likely to have been submerged at a certain time and 

 which remained emerged. The stratigraphic hiatus in the latter is thus 

 correlated with a deposit in the former. The principle also contemplates 

 correlation by probabilities suggested directly and solely by the known 

 or inferred extent of continental seas at given times. 



The application of this principle in correlating disconnected deposits 

 and stratigraphic breaks in adjacent troughs presupposes detailed knowl- 

 edge concerning the distribution of beds and faunas in the area affected 

 by the oscillatory movements. Such an area would be the Appalachian 

 Valley in east Tennessee, where the areal distribution of fifteen or six- 

 teen successive Ordovician zones and formations, beginning with the 

 base of the Stones Eiver and ending with the calcareous Cincinnatian 

 sandstone commonly referred to as the Bays, is fairly well worked 

 out. In this area the age of a l^ed lying between two previously recog- 

 nized formations may be quickly decided. For instance, in the west 

 Knoxville and Pearisburg troughs, a fine grained limestone, in contact 

 with the Knox beneath and the Holston above, would be at once referred 

 to the Mosheim. And this reference would be justified not only because 

 no other limestone is known to occur in this position in those troughs, 

 but also because in the development of the Stones Eiver oscillations only 

 the lower or Mosheim division spread entirely across the valley. Follow- 

 ing the Mosheim the middle troughs were emerged so as to separate the 



