dred feet below the present snow-line. Mount Rainier still carries a dozen glaciers of considerable size, and all the country is glaciated about this and the other snow-peaks of Washington Territory, Mount St. Helen's, Mount Baker, Mount Adams, Mount Olympus, etc., around Puget's Sound, and on Vancouver's Island. In British Columbia, as shown by George M. Dawson, Dr. Hector, James Richardson, and others, the signs of ancient glaciation are conspicuous in all the high country explored. Along the coast farther north the ancient glaciers have left their marks in all the fiords, and those of the present day descend lower and lower until in Alaska they reach the sea-level.

The valleys of the Wahsatch range were once filled with masses of ice as far south as Central Utah. A type of these, though not the largest, was Little Cottonwood glacier, of which the record has been carefully studied by the writer. It formed in a cirque at Alta ten to eleven thousand feet above the sea, where its bed is everywhere conspicuously glaciated. It had a length of about ten miles; its thickness, as shown by the line of granitic blocks, left along its sides, was five hundred feet or more, and its lower end protruded into the Salt Lake Valley at a level not greater than fifty-five hundred feet above the ocean. The glaciation of the Uintah Mountains has been graphically described by King, who says that all the higher portions of the range were once covered with a continuous sheet of snow and ice, and that glaciers descended through all the important valleys; also that the ancient glaciers of the Uintahs occupied a greater area than all those now existing in the Alps.

In the Rocky Mountain belt the signs of ancient glaciers abound from the northern part of New Mexico through Colorado and along the great divide in Wyoming, Montana, and Idaho. In the valley of the Arkansas, particularly about Leadville, in the Parks, on Clear Creek, in the valley of the Rio Grande, roches moutonnées, lateral and terminal moraines, embankment-lakes, etc., all the work of glaciers, have been observed by thousands of travelers. Here the mountainbelt is very broad and high, is now the great condenser from which radiate all the most important streams of the West, and in winter is covered with a heavy sheet of snow. In ancient times it played a similar role, only that the snows of winter did not melt as now in summer, but accumulated from year to year until they produced great glaciers. In Wyoming the mountains are narrower and lower, and the glacial signs are less conspicuous; but toward the British line, where the ranges multiply and the summits are higher, the records of glaciation are everywhere apparent.

To summarize the description of the glacial phenomena of this Western region, it may be said that, over all the mountain-ranges north of the limit before given, the traces of ancient glaciation are alike in character and apparently of the same date, and are evidently the effects of general and not local causes.

In the country east of the Mississippi the evidence of ancient glaciation is even more wide-spread and impressive than in the far West. The surface-rocks of Canada, New England, New York, and the greater part of Ohio, Indiana, Illinois, and Wisconsin, bear marks of ice-action, and are generally covered with a sheet of drift material which has been carried from the north southward, often many hundred miles. This

glaciated and drift-covered area extends from Maine and Massachusetts in a belt parallel with the arch of the Canadian highlands five hundred miles wide and more than two thousand miles long. Its extension northward from the head-waters of the Mississippi has not been traced farther than Lake Winnipeg, where it was studied by

Hind; but there are good reasons for believing that it reaches northward to the Arctic Ocean, and that the great lakes of the north, like those of the St. Lawrence chain, Superior, Huron, Michigan, etc., are pre-glacial river-valleys scooped out and modified by ice.*

From the facts already gathered, it is a justifiable inference that fully half the Continent of North America and nearly all north of the fortieth parallel was at one time covered with ice or perpetual snow, and, so far as we can now judge, the glaciation of all the North American localities enumerated was synchronous.

Some writers have attempted to prove that a large part of the glacial phenomena described above is really the work of icebergs and shore-ice, and one of the consequences of a great continental subsidence; but no man who has studied the inscriptions made by glaciers will hold to this theory when he has traversed much of the glaciated areas east or west of the Mississippi. To all the mountainous region of the West it is evident that the iceberg theory is inapplicable, and, when the enormous glaciers of the West are conceded, it is difficult to see why they should be denied to the East. Since, however, the iceberg theory is insisted upon for this section, it may be well enough to say that it is demonstrated untrue by four unanswerable arguments :

1. The inequalities of level in the fancied water-line formed by the margin of the drift area are irreconcilable. Mount Washington, Mount Marcy, Mount Mansfield, must have been totally submergedbecause their tops are worn and striated-while the shore-line was at New York at the sea-level, in Pennsylvania twenty-one hundred feet, and at Cincinnati three hundred feet higher. South of the drift-line, high lands and low were alike beyond the reach of the flood, while in Wisconsin it spared a special district not above the general level, and all around it the rocks are scored and strewed with débris.

2. The direction of the ice-scratches and the derivation of the bowlders would require the submergence of all the northern portion of the continent, so that icebergs (which had no land to start from, and therefore could not have existed) could float southward over all the Canadian highlands; and the local variations of direction (south west by south, in the basin of Lake Erie, south in that of Lake Huron, south-southwest in Lake Michigan, southwest in Lake Superior, and southeast in New England) show an incomprehensible tangle of ocean

currents.

3. The complete absence of marine shells from the great drift area of the interior, while they are abundant in the Champlain and bowlder clays on the coast, is incompatible with this theory.

4. The inscription left by the eroding agent is altogether sui generis, and characteristic of glacial action, and not at all that which could be effected by dragging masses of ice over the sea-bottom. This in itself is a conclusive refutation of the theory. The record made by a glacier

*See position of northern lakes on map.

is unmistakable, and no one who has not learned the language in which it is written is warranted in taking part in the discussion; but he who has done so will find graven on the rocks of the Alps, the hills of New England, the basins of the Great Lakes, and the mountains of Colorado and Oregon, an inscription which is every where the same, which can have but one meaning, and bears a signature that can not be counterfeited.

While it is hopeless to expect that all men will agree upon thisor any other-subject, I think I am justified in saying that the facts which have been stated, and others of like import, constitute an indisputable record, not necessarily of the former existence of a great icecap over all the northern regions, but of the simultaneous prevalence of sheets of land-ice, i. e., glaciers, over great areas of our continent; and that these glaciers, forever in motion, holding imbedded in their substance sand, gravel, and bowlders, pressed against the underlying rock by their enormous weight (probably averaging fifty thousand pounds to the square foot),* became powerful agents of erosion; general and uniform when they were broad, narrow and special when they were local. This is the reading of the facts now given by those who are best qualified to judge of the import of the phenomena in the Old World and in the New. Already the belief in an ice period and ancient glaciers is general-hereafter, with more complete knowledge of the subject, it must become universal.

Accepting the facts cited above as demonstrating the truth of the glacial hypothesis, and as proving beyond cavil the reality of an ice period, we now pass to consider the proximate and remote causes of the distinctive phenomena of this remarkable chapter in geological history.

With characteristic conservatism Lyell endeavored to account for the prevalence of glaciers over the northern hemisphere by supposing them to be due to a peculiar arrangement of land and sea; broad and elevated areas of land in the Arctic regions, low and narrow land surfaces in the tropics. I have elsewhere † discussed this question at some length, and have shown that this theory is untenable, because: First, during the Tertiary age the land was high at the north, no marine Tertiary deposits being found there; Asia, Europe, and America were then connected by land, and the tropical currents were excluded from the Arctic Ocean, but in that age a warm climate

* Fifty-four thousand eight hundred and ten pounds for one thousand feet in thickness; in some cases (around Mount Washington), probably two hundred and fifty thousand pounds.

+ "Popular Science Monthly," July, 1876.

At least through the channels of the North Pacific or North Atlantic. It has been suggested that in the Tertiary ages a communication existed between the Mediterranean and the Arctic Ocean by way of the Caspian Sea, Sea of Azov, etc.; but if there was an open channel across Western Asia at that time-which has not been proved-it could hardly have been broad and deep enough to permit a flow through it both ways (for no other channel is known) of sufficient volume to modify the climate of the Arctic regions.

prevailed over all the Arctic regions. At the same time the tropical lands were locally if not generally lower than now, since in the West Indies and on the borders of the Gulf of Mexico are broad sheets of marine Tertiary; Second, because all evidence is wanting of high northern lands during the ice period; and, at least during a portion of the time when an arctic temperature prevailed from New York northward, the sea stood much higher than now, receiving and precipitating the Champlain clays-the fine flour ground by the land glaciers-and burying in them arctic shells.*

In the article referred to I have shown that no terrestrial causes yet suggested are adequate to produce an ice period, and that we are compelled to look to some cosmical cause for an explanation of its

Occurrence.

Recently a voluminous and elaborate review of the subject has been published by Professor J. D. Whitney, with the title of "Later Climatic Changes," the object of which is to prove that there has never been an ice period, properly speaking. To establish this, it is claimed that ice has little or no eroding power; and the few ancient glaciers, of which the evidence can not be ignored or sophisticated, are considered as the products of local causes. Following Lecoq and others, Professor Whitney claims that since snow and ice are forms of moisture evaporated elsewhere by heat, the extension of glaciers at any time or place is simply an effect of increased evaporation-of heat and not cold-and hence if there ever was an ice period, meaning a time when glaciers were more widespread than now, it must have been a warmer period than the present; forgetting, apparently, that increased congelation is the only necessary feature in the increase of glaciers, for, without this, increased evaporation and precipitation would be inoperative. Only a few of many facts need be cited to show that this theory is untenable 1. Glaciers are now confined to altitudes and latitudes where the temperature is low-Alpine summits and the Arctic and Antarctic Continents. To extend the reach of the glaciers now existing, and to reproduce them where they existed formerly but are now absent, it would be only necessary to widen and intensify the conditions upon which their existence depends, viz., lower the temperature and cause the present precipitation to be more generally fixed in ice and snow. A single example will be sufficient to prove the truth of this statement. On the Cascade Mountains in Oregon we find a copious precipitation of rain and snow, but no ice where great glaciers formerly existed. The snow-fall is so heavy that the snow-line is brought down to an altitude of seven thousand feet above the ocean, and there the temperature is high enough to permit the vigorous

* The Champlain clays about New York are near the present sea-level. At Croton Point they are 100 feet higher; at Albany, 200 feet; on Lake Champlain, 350 feet; at Montreal, 500 feet; on Labrador, 800 feet; on Davis Strait, 1,000 feet; and at Polaris Bay, 1,600 feet above the ocean.