With the beginning of the Horsetown age, however, this northern communication was interrupted doubtless by the closing of Bering Sea, and a connection was formed with the waters of southern Asia, and in that way with central Europe. The decided contrast which we find between the Lower Cretaceous faunas of California and those of Texas points to the existence of a land barrier between the seas of the two regions.
In the southern region the Lower Cretaceous was terminated by an upheaval, which caused the Comanchean Sea to withdraw from Texas and the area to the west and north of it. This mid-Cretaceous land epoch must have continued for a considerable time, permitting extensive denudation and a complete change in the fauna. Wherever the marine Upper Cretaceous is in contact with the Comanche limestones north of Mexico, the two are unconformable, and no species of animal is known to pass from one to the other. In Mexico the Lower Cretaceous passes into the Upper without a break, the disturbances there taking place at a later date.
The Upper Cretaceous rocks have a far wider distribution over North America than have those of the lower division, which is due to an enormous transgression of the sea over the land, one of the greatest in all recorded geological history. Over the region of the Great Plains the Upper Cretaceous was inaugurated by the formation of a non-marine stage, the Dakota. These strata cover much of Texas, lying unconformably upon the Comanche series, and extend northward into Canada. In Kansas, however, the connection of the Dakota with the Washita is very close, bands of sandstone carrying Dakota species of plants being interstratified with the marine beds. On the western side of the Colorado uplift, the Dakota is less distinctly a sandstone formation, and is characterized by beds of shale and even coal seams of workable thickness. In most parts of the Rocky Mountain region the Dakota rests in apparent conformity upon the lowest continental Cretaceous, and even upon the Jurassic. In the Uinta and Wasatch ranges there is no apparent break in sedimentation from the Palaeozoic to the end of the Mesozoic, though the whole Lower Cretaceous is there wanting.
From this we may infer that during the long Lower Cretaceous time all these regions had been low-lying lands, nearly or quite at base-level, and therefore not subject to profound denudation.
It was at the end of the Dakota age that the great subsidence took place which affected nearly all parts of the continent, and brought the sea in over vast areas where for ages had been dry land. South of New England the Atlantic coastal plain was submerged, and in New Jersey, at least, the waters covered even the nearly base-levelled Triassic belt, bringing the sea up to the foot of the crystalline highlands. The lowlands of Maryland, Virginia, and the Carolinas, and all of Florida were under the ocean, and the Gulf of Mexico was extended northward in a great bay (the Mississippi embayment), covering western Tennessee and Kentucky and extending into southern Illinois. In Mexico important changes occurred during the Upper Cretaceous. Early in this epoch a general elevation restricted the sea to the northeastern part of the country, but in the south was an opposite movement, the sea transgressing upon the ancient land which lay to the west of the Isthmus of Tehuantepec, finally covering it late in the period.
Elsewhere, the disturbance referred to was orogenic, resulting in the formation of most of the Mexican mountain ranges and continuing nearly or quite to the end of the period. Texas was again extensively submerged, and a wide sea connected the Gulf of Mexico with the Arctic Ocean. The eastern coast of this interior sea began in northwestern Texas, running through Kansas and Iowa nearly to the present line of the Mississippi River. Westward the coastline was the uplift which ran from the west coast of Mexico into British Columbia. The Colorado region was again converted into islands. North of the Great Basin land the interior sea was connected with the Pacific and Arctic Oceans, which united over the northwestern part of the continent.
On the Pacific side, the Sierras, which had suffered greatly from denudation, were again folded. A moderate transgression of the sea caused the Upper Cretaceous to extend farther east than the Lower. Volcanic activity continued and immense bathyliths were formed deep within the mountains. The sea extended from Lower California northward along the Sierra into eastern Oregon at the foot of the Blue Mountains.
The North American continent was thus divided into two principal land masses, the larger one to the east and comprising the pre-Cambrian and Palaeozoic areas. In the limits of the United States this land lay almost entirely east of the Mississippi, except for a southwestern peninsula, including Missouri, Arkansas, Oklahoma, and part of Texas. The western area was much smaller, extending from Mexico into British Columbia, and having its greatest width between the fortieth and forty-fifth parallels of latitude. Between the two lands lay the Colorado Islands, and doubtless many smaller ones as well.
The character of sedimentation differed so much in the various regions of the continent that the subdivisions of the Upper Cretaceous have received different names in the separate provinces, and only approximately correspond in time. The greater number of these subdivisions, which are founded chiefly upon physical changes, gives to the Upper Cretaceous the appearance of being longer and more important than the Lower, but this is only an appearance. In Europe the Lower Cretaceous has, in recent times, been divided into six series, an arrangement which proves to be of general validity.
Along the Atlantic border the Upper Cretaceous strata -are a series of marine sands and clays, which are still almost horizontal in position and of loose, incoherent texture. In New Jersey there are extensive developments of green sands locally called marl. The Appalachian Mountains, which had been subjected to the long-continued denudation of Triassic, Jurassic, and Lower Cretaceous times, were now reduced nearly to base-level, the Kittatinny plain of geographers (see p. 512). This peneplain was low and flat, covering the whole Appalachian region, and the only high hills upon it were the mountains of western North Carolina, then much lower than now. Across this low plain the Delaware, Susquehanna, and Potomac must have held very much their present courses, meandering through alluvial flats.