Coal beds, Devonian, Carboniferous, Mesozoic and Tertiary, alike, are associated with shales, sandstones and, in many cases, with calcareous beds, the last often containing a marine or a fresh-water fauna. Interior or limnic basins frequently bear close resemblance to paralic or coastal basins, so that distinction between the types becomes arbitrary in some great areas. In the Indiana-Illinois field, wide invasions of the sea appeared again and again throughout practically the whole period of accumulation. On the other hand, the Appalachian basin, almost land-locked during most of its history, experienced few invasions and those, of comparatively small extent, were confined to the earlier periods; in the later stages, the whole region was practically limnic.

Study of reports by observers in the several countries makes certain that conditions needed for formation of coal beds were to

'Part I. appeared in these Proceedings, Vol. L., pp. 1-116; Part II. in same volume, pp. 519-643.

PROC. AMER. PHIL. SOC., LI. 207 A, PRINTED DEC. 13, 1912.

all intents and purposes the same in all regions and in all periods, from the Devonian coals of Bear Island in the Arctic to the Tertiary coals of Wyoming or Trinidad; but the varying descriptions and explanations presented by students make equally certain that one cannot ascertain what the essential conditions are, if his investigation be confined to areas embracing a score or even several hundreds of square miles. The investigation must cover a great area, in which merely local features do not obscure those which are general and which actually bear upon the problem in hand. Such an area is the Appalachian Basin of the eastern United States, where one finds the Pennsylvanian or Coal Measures divided into

The Appalachian coal field, now embracing approximately 70,000 square miles of almost continuous deposits, occupies only a part of the original area. The deep synclinal basins of anthracite in eastern Pennsylvania are separated by 50 to 100 miles from the great bituminous region at the west, while southwardly one finds insignificant fragments along the eastern side until he comes to Georgia and Alabama. The greatest extent of the area of deposit was probably at the close of the Pottsville, when it reached from southern New York in west southwest direction to beyond central Alabama, more than 800 miles; at the north, it spread from the old Appalachian land, at the east, westward to beyond Newark in

J. J. Stevenson, "Carboniferous of the Appalachian Basin," Bull. Geol. Soc. Amer., Vol. 18, 1907, p. 178. The Pottsville is subdivided in this paper into Beaver and Rockcastle. I. C. White, in West Virginia Geol. Survey, Vol. Ia, 1908, p. 13, has suggested that Rockcastle be replaced by New River and Pocahontas; this should be accepted, as Stevenson did not assign proper significance to Pocahontas, regarding it as merely a subordinate stage.

Ohio, while at the south it reached the western boundary of Alabama. The area of deposit at that time embraced not less than 200,000 square miles. The present outcrop approaches the western border at a few localities in Ohio and Kentucky as well as in Alabama, but for the most part it is two score or more miles east from the original limit. The eastern border is approached in the southern anthracite field of Pennsylvania and apparently it was not far eastward from the Pocahontas outcrop in Virginia; in Alabama, the eastern outcrop is not more than 25 miles from the original border on that side. But, in most of the space north from Alabama, the present continuous outcrop is from 30 to 100 miles west from that border as it probably existed at the close of the Pottsville. The Appalachian field included a small part of New York, more than two thirds of Pennsylvania, the western third of Maryland, nearly the whole of West Virginia, the eastern third of Ohio and Kentucky, with southwestern Virginia, eastern Tennessee, the northern half of Alabama as well as northwestern Georgia. Here then is an area of sufficient extent to provide ample illustration of purely local features and their relations to the effects of widely acting agents.

THE APPALACHIAN BASIN.

The Appalachian basin, from its origin to the close of the Paleozoic, was the scene of frequent changes in the relations of land and water. Schuchert and Ulrich have shown that such changes were merely commonplaces in the earlier periods. Those students are not in agreement respecting several matters, which have much interest from a philosophical standpoint, but they are in full agreement respecting all matters which concern the questions at issue here. As Schuchert has shown, the Appalachian basin. originally was continuous with the broad Mississippi region and much of it was covered with sea. Toward the close of the Ordovician the Taconic revolution began, which, at the east, widened

C. Schuchert, "Palæography of North America," Bull. Geol. Soc. Amer., Vol. 20, 1910, pp. 427-606; E. O. Ulrich, "Revision of the Paleozoic Systems," ibid., Vol. 22, 1911, pp. 281-680.