THE PENNSYLVANIAN PERIOD. 



575 



western Pennsylvania, Ohio, and West Virginia, and to have at least 

 an equal extent where too poor to be generally worked. Many coal 

 beds, on the other hand, do not 

 occupy great areas. From their 

 thicker portions they thin out in 

 all directions, often grading into 

 black shale. From these facts it 

 is inferred that within the general 

 area of a coal-field there may have 

 been islands and peninsulas above 

 the marsh level. Such areas in- 

 terrupted the continuity of the 



Fig. 260. — Irregularity in a coal bed 



showing a " horseback " in Craig 



Slope, Kalo, la. (Keyes, la. Geol. 

 Surv.) 



swamps then, and interrupt the continuity of the coal-beds now. 



No parallelism between the coal-beds of different fields has been 

 made out, and is not likely to be. Even within the same 'field the 

 parallelism is by no means perfect, as shown by the fact that more 



Fig. 261. — Bifurcation of a coal bed. The heavy black lines = coal. (Winslow.) 



seams are found in some parts of a field than in others. On the other 

 hand, the larger coal-beds are often continuous over great areas, as 

 already noted in the case of the Pittsburg bed. 



Traced laterally, a bed of coal is sometimes found to divide (Fig. 261) 

 or, traced in the opposite direction, two beds are sometimes found to 

 unite. 1 Fig. 262 suggests a possible sequence of events by which the 



1 The Nature of Coal Horizons; see Winslow, Geol. Surv. of Missouri, Preliminary 

 Report on Coal, p. 25. Also Keyes, Jour. Geol., II, 178. 



