differentiated are the following, numbered in chronological order: (5) Wisconsin, (4) Iowan, (3) Illinoian, (2) Kansan, (1) Sub-Aftonian, or Jerseyan. Of these, the Kansan ice-sheet was the most extensive, and the later ones constitute a diminishing series.

Essentially all phases of glacial and aqueo-glacial drift are represented. The principal terminal moraines are associated with the ice of the Wisconsin epoch. Terminal moraines at the border of the Illinoian drift are generally feeble, though widely recognizable, and such moraines at the margin of the Iowan and Kansan drift sheets are generally wanting. The edge of the oldest drift sheet is buried by younger sheets of drift in most places.

Loess is widespread in the Mississippi River basin, especially along the larger streams which flowed from the ice. Most of the loess is now generally believed to have been deposited by the wind. The larger part of it seems to date from the closing stages of the Iowan epoch, but loess appears to have come into existence after other glacial epochs as well. Most of the fossils of the loess are shells of terrestrial gastropods, but bones of land mammals are also found in not a few places. Some of the loess is thought to have been derived by the wind from the surface of the drift soon after the retreat of the ice, before vegetation got a foothold upon the new-made deposit; but a large part of the loess, especially that associated with the main valleys, appears to have been blown up on to the bluffs of the valleys from the flood plains below. As might be expected under these conditions, it ranges from fine sand to silt which approaches clay in texture. Its coarser phases are closely associated with dunes in many places, and locally the loess makes a considerable part of the dune material.

Much interest attaches to estimates of time based on data afforded by the consequences of glaciation. These estimates are far apart, and must be regarded as very uncertain, so far as actual numbers are concerned. The most definite are connected with estimates of the time since the last glacial epoch, and are calculated from the amount and rate of recession of certain falls, notably those of the Niagara and Mississipi (St Anthony Falls) rivers. The estimate of the time between the first and last glacial epochs is based on changes which the earlier drift has undergone as compared with those which the younger drift has undergone. Some of the estimates make the lapse of time since the first glacial epoch more than a million years, while others make it no more than one-third as long. The time since the last glacial epoch is but a fraction of the time since the first—probably no more than a fifteenth or a twentieth.

Outside the region affected by glaciation, deposits by wind, rain, rivers, &c., have been building up the land, and sedimentation has Non-glacial. been in progress in lakes and about coasts. The non-glacial deposits are much like the Tertiary in kind and distribution, except that marine beds have little representation on the land. On the coastal plain there is the Columbia series of gravels, sands and loams, made up of several members. Its distribution is similar to that of the Lafayette, though the Columbia series is, for the most part, confined to lower levels. Some of its several members are definitely correlated in time with some of the glacial epochs. The series is widespread over the lower part of the coastal plain. In the west the Quaternary deposits are not, in all cases, sharply separated from the late Tertiary, but the deposits of glacial drift, referable to two or more glacial epochs, are readily differentiated from the Tertiary; so, also, are certain lacustrine deposits, such as those of the extinct lakes Bonneville and Lahontan. On the Pacific coast marine Quaternary formations occur up to elevations of a few scores of feet, at least, above the sea.

Igneous rocks, whether lava flows or pyroclastic ejections, are less important in the Quaternary than in the Tertiary, though volcanic activity is known to have continued into the Quaternary. The Quaternary beds of lakes Bonneville and Lahontan have been faulted in a small way since they were deposited, and the old shore lines of these lakes have been deformed to the extent of hundreds of feet. So also have the shorelines of the Great Lakes, which came into existence at the close of the glacial period.

Much has been written and more said concerning the existence of man in the United States before the last glacial epoch. The present state of evidence, however, seems to afford no warrant for the conclusion that man existed in the United States before the end of the glacial period. Whatever theoretical reasons there may be for assuming his earlier existence, they must be held as warranting no more than a presumptive conclusion, which up to the present time lacks confirmation by certain evidence.

The following sections from selected parts of the country give some idea of the succession of beds in various type regions. The thicknesses, especially where the formations are metamorphosed, are uncertain.

Triassic.
Chicopee shale	200	ft.	(?)
Granby tuff	580	”
Blackrock diabase (cones and dikes)
Longmeadow sandstone	1000	”
Sugarloaf arkose	4660	”
Mount Toby conglomerate
Unconformity.
Devonian.
Bernardston series	1950	ft.
Unconformity.
Silurian.
Leyden argillite	300	ft.
Conway schist ${\displaystyle \scriptstyle {\left.{\begin{matrix}\ \\\\\ \ \end{matrix}}\right\}\,}}$   Amherst schist   Brinfield fibrolite-schist	5000	”	(?)
Goshen schist	2000	”	(?)
Unconformity.
Ordovician.
Hawley schist	2000	ft.	(?)
Savoy schist	5000	”	(?)
Chester amphibolite	3000	”	(?)
Rowe schist	4000	”	(?)
Hoosic schist	1500	”	(?)
Unconformity.
Cambrian.
Becket gneiss	2000	ft.	(?)
Unconformity.
Proterozoic.
Washington gneiss	2000	ft.	(?)
(Base not exposed.)

The above section is fairly representative for considerable parts of New England.

This section is fairly representative for the Appalachian Mountain tract, though the Cambrian is often more fully represented.