they are warmed by it. Hence the precipitation is copious in the former regions and meagre in the latter. Between the two belts where these opposite effects are pronounced is a region where they shade into each other, and though this intermediate region cannot be marked out by distinct boundaries it may still be said to exist in latitudes lying within the valley of the Columbia river.

The cause of an arid climate thus indicated may be regarded as generally operative throughout the western mountain region; and it will no doubt appear upon full consideration to be much more potent and widely extended in its action than any or even all of the mountain ranges could be. It is, however, greatly modified by the intervention of local causes, which occasionally mask or obscure it. The precipitation in different portions of the region is highly irregular and several modifying causes can be indicated which, though they do not nullify the more general one here set forth, frequently become much more conspicuous in their effects. For instance, it is well known that the heaviest rainfall in the United States, excepting possibly upon some mountain tops, occurs upon the coast of Oregon and Washington Territory. But as already indicated this is the locality where we find the neutral axis, so to speak, of the alleged causes favoring respectively humidity and aridity, and where their effects are at minimum or even at zero. Moreover, the westerly winds saturated with moisture here strike the coastwise mountains, and are suddenly thrown upward several thousand feet before they have had time to feel the heating effect of the land which is here very slight; and the precipitation is thus very copious. Descending to lower levels inland they soon become dry and produce a sub-arid climate.

The most frequent variants of climate are the great differences of altitude in different portions of the west. The mountain tops and summits of the plateaus are always well watered, and in any given latitude the rainfall increases or diminishes at a fairly definite rate with the altitude. But the variation of rainfall with altitude is by no means a simple ratio. Between 4500 and 6000 feet the difference in rainfall is not great, between 6000 and 7500 feet it is very considerable; between 7500 and 9000 it is still greater.

Moreover the rainfall is greater ceteris paribus in high latitudes than in low latitudes. In passing from the southern to the northern boundary, if we compare localities of equal altitudes

along any given meridian, we shall find the rainfall steadily though perhaps not uniformly increasing. This is an obvious consequence of the theory suggested.

Although no very great effects upon the general condition of aridity are here attributed to the depletion of moisture by the passage of the winds over mountain ranges, it is still true, no doubt, that highly important local effects are thereby produced. The rainfall at the eastern base of the Sierra Nevada, and for two hundred miles east of it, is most probably reduced very greatly by this cause. In the sink of the Humboldt river, the annual precipitation seldom reaches four inches, and may average not more than three inches; but, as we pass eastward beyond the wake of this range, its effects become gradually less, and long before the Wasatch is reached they have become inconsiderable. Since the Sierra Nevada is the longest, highest and widest of the individualized ranges of the Rocky system, its local effect upon the humidity of the plains and valleys lying immediately under its lee is greater than that of any other. But the same kind of effect is perceptible in some other ranges.

The discussion of the causes of local variations in climate might be almost indefinitely extended. Nothing more is designed here than to advert to one general cause of aridity which prevails over the entire region, and which everywhere persists, though it is often obscured, sometimes reversed and sometimes reinforced by local

causes.

THE EXCAVATION OF THE GRAND CANON OF THE COLORADO RIVER. By C. E. DUTTON,' of Washington, D. C.

[ABSTRACT]

THE lecturer first exhibited lantern views of the chasm, taken at a locality near the middle of its length, which show the simplest and most typical form of the cañon. It consists of an upper chasm 2,000 feet deep, and five to seven miles wide, including a broad

Captain U. S. Ordnance Corps, U. S. Geological Survey.
This paper was given as a lecture on Wednesday evening.

and comparatively smooth floor between the palisades. Within this floor is sunk an inner and narrower gorge, 3,000 feet deeper and about 3,200 to 3,600 feet wide. The rigorous uniformity in the profiles of the palisades and their ornate architectural appearance were adverted to. The strata through which the chasm is cut are of Carboniferous age, excepting the lowest 400 or 500 feet, which are Lower Silurian or Primordial.

The audience was then conducted north of the river to the great desert-plain of the neighboring region, where the general character of the topographical features was exhibited by appropriate pictures. About forty miles north of the cañon begins a series of terraces, ascending step by step to the northward, called the Terraces of the High Plateaus. Each terrace consists of a distinct geological formation beginning below with the Permian; and, rising by successive steps upon the Trias, the Jura, the Cretaceous and the lower Eocene, we at last reach the top of the most recent formation (lower Eocene) of the region. Each terrace terminates in a gigantic cliff-wall of remarkable architectural appearance, as shown in the illustrations.

Reaching the summit of the Eocene, and looking backwards and downwards over the stairway of terraces, the facts thus briefly glanced at were discussed from a geological standpoint. The formations consisting of Permian, Mesozoic and Eocene strata, disclosed and abruptly cut off in the terraces, are inferred to have reached formerly much farther southward, and in fact to have extended over the entire Carboniferous platform (13,000 to 15,000 square miles) now drained by the Grand Cañon. The total thickness of the beds thus denuded was on an average very nearly 10,000 feet.

It was then shown that this denudation began about Middle Eocene time and has continued until the present. By far the greater part of it was effected during Eocene and Miocene time.

The great amount of uplifting undergone by the Grand Cañon platform was then shown, being in the aggregate (during Tertiary and Quaternary time) 16,000 to 19,000 feet. This amount, though great, is surpassed by other localities of the west. The present altitude is the difference between the total uplifting and the thickness of strata denuded, and it varies from 4,700 to 9,200 feet. The elevation was not at a uniform rate,

but was broken by periods of repose, the evidences being recorded in the drainage channels and faults.

The excavation of the present Grand Cañon began after the greater part of the denudation had been accomplished. The river began to flow in Eocene time, and had to cut through 10,000 feet or more of later formations before it entered the summit beds of the Carboniferous which now form the highest crest of the cañon wall. The beginning of the present Grand Cañon was in Pliocene time. The evidence of this is cumulative. Collectively it is overwhelming, but as it consists of many discrete parts, no one of which by itself alone would be conclusive, it will not be advisable to attempt to present the evidence for want of time.

The mechanical process of excavating the cañon was then alluded to. It is customary to say that the river cut it. This is so far true, but it is less than half the truth. The processes may be referred to two groups, corrasion and weathering, both of which are highly complex. Corrasion is the action of a river upon its bed by which, under certain conditions (the general nature of which was briefly explained), it scours down its channel in the rocks to lower and lower depths. But the cut sawed by the river cannot be wider than the water surface of the stream, and the Grand Cañon is five to eleven miles wide. The cañon is widened by the weathernig of its walls. This is an extremely complex process, though now it has come to be pretty well understood, and under the conditions prevailing in the Plateau country, it gives rise to many curious, beautiful and instructive results. A single instance was given by way of example, viz.: the operations by which the remarkable cliff-profiles of the cañon are established and maintained. The discussion of the details of natural architecture would require a large volume.

The hearers were then conducted to the grandest portion of the chasm, the Kaibab Plateau, and a description of its superlative scenery was given, illustrated with lantern pictures.

ORIGIN OF THE GREAT LAKES; BY J. W. SPENCER.

131

A SHORT STUDY OF THE FEATURES OF THE REGION OF THE LOWER GREAT LAKES DURING THE GREAT RIVER AGE; OR NOTES ON THE ORIGIN OF THE GREAT LAKES OF NORTH AMERICA. By J. W. SPENCER, of Windsor, N. S.

I PURPOSE bringing before this section a few notes on the physical features of the Great Lake Region, which have a bearing on the origin of the lakes themselves, with a few deductions therefrom. Although the bibliography of the subject is scanty, I will not detain the Association with a notice of what has been written. While working out the origin of the Dundas valley, at the extreme western end of Lake Ontario, the discovery, that the present great rock-bound valley is only one of insignificance compared with the buried channel of Preglacial date, led to the broader study of the origin of the lake-basins themselves, as the buried channel in the Dundas valley appeared to form a portion of the Preglacial outlet of the basin of Lake Erie into that of Lake Ontario. On this subject my first paper was read last March before the American Philosophical Society, and was published in the last volume issued by it. The same paper has been subsequently reprinted in volume Q, of the Reports of the Geological Survey of Pennsylvania. To this paper frequent reference will be made. During the present summer further details have been worked out, and observations have also been extended to the more important small lakes of Central New York. But only some of the results of these observations can here be noticed.

Theories of the origin of the Lakes.-Of these there are three :1, the basins of the lakes are geological valleys; 2, the basins. were excavated wholly or partly by glacier-action; 3, the basins were excavated by atmospheric and fluviatile erosion, with their outlets closed by the drift of the Ice Age, assisted probably by geological uplifts. The relative value of these explanations will be seen in the succeeding pages.

Features along the Preglacial Outlet of the Erie Basin into the Basin of Lake Ontario.-The Niagara escarpment encloses the

1 The title of the above mentioned paper, read March 18, 1881, is-"Discovery of the Preglacial Outlet of the Basin of Lake Erie into that of Lake Ontario; with Notes on the Origin of our Lower Great Lakes. By Prof. J. W. Spencer, B. A. Sc., Ph. D., F. G. S. King's College, Windsor, N. S.