Along the Atlantic and Gulf borders the coast-line of the Eocene closely follows that of the Cretaceous, of which only a narrow strip separates the Eocene from the Triassic and crystalline rocks of the Piedmont plain. The unconformity between the Cretaceous and Eocene indicates that along this coast the latter period had been inaugurated by an encroachment of the sea upon the land. The Mississippi embayment had nearly the same size and form as before, extending up to the mouth of the Ohio. Florida was entirely submerged, as was most of Central America, cutting off the northern from the southern continent. On the Atlantic coast the Eocene rocks are unconsolidated sands and clays, with some glauconitic greensand, particularly in New Jersey. They form a narrow belt through New Jersey, Maryland, and Virginia, widening into a quite broad band through the Carolinas and the Gulf States, and extending around the borders of the Mississippi embayment into Texas. In the Gulf region the rocks are more consolidated, and are quite hard limestones, sandstones, and shales, with extensive deposits of lignite, formed in ancient peat bogs which followed the low-lying Gulf shores; some of these lignitic formations may be more properly referable to the Paleocene.

On the Pacific coast a long, narrow arm of the sea occupied the great valley of California, extending northward into Oregon and Washington; its deposits are at present principally displayed along the eastern flank of the Coast Range. These deposits form a single series, the Tejon, which lies upon the Chico in apparent conformity; but the lowest Eocene is not represented in the Tejon, and in Oregon an unconformity between the two series has been detected. The presence of Atlantic species in the Eocene of California indicates that the barrier between the oceans which had existed in the Cretaceous, was submerged for a time.

In the western interior the Eocene formations, with the exception of a few small areas in Colorado, are all confined to the plateau region west of the Rocky Mountains. Formerly, they were considered to be lacustrine deposits, but this explanation is applicable to only a small part of them; they are much more generally due to the subaerial agents, rivers, rain, and the wind, in the work of filling basins of slow depression. In some instances, at least, the material of these beds is principally fine volcanic ash and dust, so extensively distributed as to prove great volcanic activity in the region. It has not yet been determined how generally the formations are built up of these volcanic materials. Whatever the mode of their accumulation, the Eocene stages are remarkably complete and have preserved a marvellous record of the successive faunas of the epoch, which registers not only the changes in animal life, but also the shifting land connections with the other continents.

The oldest and most widely spread of these Eocene stages is the Wasatch, the principal area of which extends from New Mexico, over eastern Utah and western Colorado, to the Uinta Mountains, around the eastern end of which it passes in a narrow band, expanding again north of the mountains and covering southwestern Wyoming. As the Wasatch is in many places buried under newer rocks, it is not certain that the beds are continuous over this great area of 450 miles long by 250 miles wide in the broadest part, but there is no reason to suppose that they were not originally so. Another great body of Wasatch strata occupies the Big Horn Basin of northwestern Wyoming, extending into southern Montana, and two small areas in southern Colorado are referred to the same date. The thickness of the Wasatch varies from 1500 to 2500 feet.

The Wind River, the second of the Eocene stages, appears to be present in two distinct facies, though, as the two contain no common fossils, the correlation is open to some doubt. The beds of the typical facies occupy the Wind River Basin, north of the mountains of that name, in central Wyoming, and contain many remains of mammals, and also occur in the Huerfano Canon, Colorado. The second facies (Green River Shales) occupies the same position stratigraphically that the Wind River does palaeon-tologically, following upon the Wasatch and overlaid by the Bridger. This facies is typically displayed in southern Wyoming in the valley of the Green River and is a thick body of very finely laminated "paper shales," which seem to have been deposited in a very shallow lake, and have preserved an extraordinary number of plants, insects, and fishes, but no mammals have been found, except in the form of footprints. Occasional crystals of gypsum indicate that the water became salt, at least locally, while the presence of such fishes as the Rays points to communication with the sea.

The third of the Eocene stages is the Bridger, of southwestern Wyoming and northeastern Utah, which lies upon the Green River Shales, overlapping these east and west and extending over upon the Wasatch. The Bridger beds are very largely made up of volcanic ash and dust deposited partly on land and partly in shallow or periodical lakes.' The remains of fishes, crocodiles, and non-marine shells scattered through the beds is a proof of subaqueous deposition, and the large selenite crystals (see p. 20), which frequently occur, indicate that the water was occasionally strongly saline.

These Eocene continental deposits are principally sands and clays, with occasional banks of conglomerate. They are more or less indurated, but usually quite soft and weather rapidly, giving rise to the characteristic bad-land scenery so frequently mentioned.

While the mammals of the Wasatch age are so largely identical with those of Europe that an easy way of land migration must have been open between the continents, the similarity becomes less and less and the two faunas follow such divergent lines of development through the Wind River and Bridger ages, that the connection between America and Europe must evidently have been interrupted. A brief connection with South America is suggested by the very unexpected discovery of an Armadillo in the Bridger, though it is possible that this animal was a survival from the connection which existed between North and South America in late Cretaceous or Paleocene times. If so, the ancestors of the Bridger armadillo yet remain to be discovered in the Wind River and Wasatch.

The Eocene epoch was brought to a close by a series of movements which added a narrow belt of land to the Atlantic and Gulf coasts. In the interior, the plateau region was elevated and drained and did not again become an area of extensive deposition. A great mountain-making disturbance followed the Bridger age, elevating the mountain ranges to which the post-Cretaceous disturbances had given birth and, in the neighbourhood of these mountains, tilting and upturning the Eocene strata.