to the north, and, flowing a score of miles, empties into Pyramid Lake, which constitutes its sink. Bigler is fresh, Pyramid is salt; Bigler has no inlet, Pyramid no outlet; and from Bigler to Pyramid flows ever the swift Truckee. The snows of the Sierra thus find their way into the lower basin, to be thence taken up to fall again as snows— I had almost said-upon their native peaks; but the "circuit of Nature" is not in this case so prettily completed; since the evaporations of Pyramid Lake are doubtless carried by prevailing winds to the east, to be precipitated, perhaps, on the rim of the basin of Salt Lake. It is evident that the change of level in one of these isolated lakes affords us a measurement with regard to the relative rain-fall and evaporation, provided the absence of drainage be ascertained. The Salt Lake is hundreds of feet lower than it once was; the former lacustrine beach can be distinctly traced, like a high-water mark, along the mountain-sides. But this great change of level is doubtless due to sudden drainage rather than slow subsidence of the water. Since the Mormons settled the country, the lake is said to be rising, and to have submerged already much land that was formerly dry the year round. This may, however, be but a temporary fluctuation.

The Salt Lake basin is separated by so narrow a divide from the system of the Snake River as to render it quite probable that this stream once drained it-that is to say, that the basin formerly spilled over its rim into the Snake river. This part of the country, or rather a region a little north and east, includes a curious nodal point, as it were, from which the headwaters of the Snake and the Colorado and the Missouri take their rise, and flow in different directions to the sea. It was near here, in the Wind River Mountains, that Frémont planted the Stars and Stripes upon what was supposed to be the highest point of the continent. Lewis and Clarke, who, in the time of President Jefferson, ascended the fork of the Missouri which still bears his name, bestowed by them, found two springs, but a short distance apart, which contributed their waters, the one to the Pacific and the other to the Atlantic. The former may be considered as the source of the Lewis River, now more commonly called the Snake. From this point it pursues a devious course to the southwest, then north, for perhaps nine hundred miles, until, having been strengthened on the way by its tributaries, the Owyhee, Boise, Sweetwater, etc., it joins with the northern fork of the Columbia, to make one of the stateliest rivers of the world.

The cañons of the Snake are little known. Lewis and Clarke did not follow down this stream, being deterred by the forbidding nature of the country, which was desti tute of food. The geography of the region being then wholly unknown, they were endeavoring to strike the headwaters of the Columbia, in order to follow it down to the infant settlement of Astoria; and finding no salmon in the upper course of the Snake, Lewis acutely inferred that either there were great falls in this river, or it did not belong to the Columbia system. The former of these suppositions was true; and Lewis may be said to have seen with the eye of faith the great falls of the Snake. Recrossing the continental divide and turning northward, these bold explorers finally reached their destination by an casier route. Frémont subsequently descended the stream farther than they had done; but he likewise abandoned its course, striking northward for a more favorable line, and apparently just missing by a few miles the sight of the great falls.

Like the cañons of the Colorado in the South, the channels of the Snake and other rivers of this region are carved by the streams themselves deep into the face of the country, leaving a general table-land above. The Colorado seems, however, to have had easier work, by reason of the softer nature of the predominant rocks. The characteristic formation of the Snake and the Columbia above the Cascades is basalticthe product of vast lava overflows. The sublime group of mountains in Oregon and Washington, which includes Hood, Adams, St. Helen's, Regnier, and Baker, with others scarcely less magnificent, is volcanic. Some of these peaks have been in active eruption within recent years; and all are known to have been so at some former period. But we have not to look for volcanic craters as the centers of the great lava floods which covered so much of Northern California, Oregon, Washington, and Idaho. These craters are scattered along a distance of five hundred miles in the Cascade and Sierra ranges, but it seems probable that a line connects them, upon which the eruptions of melted lava took place abundantly, repeatedly, but perhaps more silently than is apt to be the case with the explosive phenomena of volcanoes. A plutonic dike may be traced along the Sierra, and is asserted to have been observed in a considerable sectional exposure in one of the deep side-cañons of the western slope. If we are yet ignorant of the geological history of these phenomena, it is merely because we have not yet had time to collect and compare observations. There are few regions on the globe where the handwriting of nature is larger, plainer, or less obscured by nature's own subsequent efforts to erase it. Geologists who have painfully endeavored to find in railroad cuttings, wells, or rare accidental exposures, the view of the structure of the rocks which forests and alluvion otherwhere concealed, will appreciate the ease and satisfactory character of reconnaissances in a wilderness, where the mountains are bare above, and cloven with cañons that reveal their very hearts.

One of the first things that strike the observer in these basaltic layers is their great aggregate thickness as well as superficial extent. This is finely shown just above the Dalles and in the cañon of the Des Chutes. For a considerable distance along the Dalles, the banks of the Columbia present a low volcanic escarpment, like the Pali sades of the Hudson, but absolutely barren. At some points, however, the whole thickness of the mass through which the river has carved its way stands revealed. It presents to us a series of overflows, each spreading out horizontally, and succeeded after an indefinite interval by another. The aggregate thickness, as shown in the Des Chutes, cannot fall short of 2,000 feet. In different localities it is easy to distinguish between the layers that have cooled under water (i. e., those which were erupted before the disappearance of the great inland lakes) and those which have congealed under exposure to the atmosphere. Possibly a still more important clew to the geological history is found in the fact that these basaltic layers inclose intercalated series of sedimentary strata. At Dalles City a bed of lava overlies a bed of conglomerate, and is in turn overlain by deposits of tufa-sandstones and clays, containing the most delicate plant-fossils that can be conceived. There are leaves, twigs, and buds in the greatest perfection. Perhaps they were brought down by the river in former times. It cer tainly seems less likely that they should have grown where now they are found. The mud and ashes constituting these intercalated beds may also have been erupted, since we know that nearly or quite all of the volcanic eruptions include the expulsion of vast quantities of ashes and hot water.

At the Cascades of the Columbia, the Des Chutes, and other rivers, not too much obstructed, salmon are speared or hooked in great abundance by the Indians. These fish find their way for more than a thousand miles from the sea, continually pressing up stream, and growing weaker, thinner, and more battered and lacerated by the rough and rocky trip. Those caught at the mouth of the Columbia are in far the best condition; and it is from this place that the great quantities are taken for packing and shipment to market. The Indians, less fastidious, catch their fish wherever it is most convenient, dry it in the sun, and preserve it for winter food.

Passing southeast from the Dalles, and crossing the valley of the Des Chutes, we come into the valley of the John Day's River. The walls of this cañon are of volcanic, sedimentary character, at least in many places, consisting of tufas and sandstones from volcanic materials. The effects upon these soft materials of atmospheric and aqueous agencies have been picturesque and curious in the extreme. At one point a complete ancient castle is perched, like the stronghold of a Raubritter of the Rhine, upon a commanding crag, from which its dwellers might look far up and down the valley. A little further on, a whole city, with towers and roofs of fantastic, quaint variety, attracts the wondering gaze. It is difficult, sometimes, to realize that these appearances are but the freaks of water, gnawing away into such fanciful forms the yielding layers of rock.

A brief trip into the forests of Washington Territory, up the valley of the White Salmon, and nearly to the foot of Mount Adams, gave me an opportunity to study one of the most remarkable features of these basaltic formations, namely, their subterranean passages or caves. I have elsewhere given some account of these; but I must beg your indulgence for a condensed description here.

The whole country in the neighborhood of Dallas is covered with basaltic overflows, intercalated here and there with beds of tufa, cemented ashes, and even altered clays, in some of which interesting tertiary fossils, both of plants and animals, have been discovered. Rev. T. Condon, of Dallas City, a naturalist, whose enthusiasm, patient industry, and wide acquaintance with this part of Oregon entitle him to a fame which his modesty has hitherto avoided, has brought together a large and interesting collection of these fossil remains, which he exhibits and explains with great courtesy to his friends, among whom, with catholic liberality, he appears to reckon all those who make demands upon his time and manifest interest in his pursuits.

It is in these basaltic overflows, not more than twenty miles from the base of Mount Adams, and in the heart, so to speak, of the Cascade Mountains, west of the valley of the White Salmon River, that a series of caves occur, some of which present the phenomena of perpetual ice. Nature continues the manufacture, and stores the product, year after year, though it is but occasionally that man, exhausting his own resources, falls back upon her forethought and bounty.

The "caves" are all old lava channels through which the melted matter flowed, after the crust had cooled and hardened overhead. The same thing may be seen on any volcano; but in most cases, I presume, the molten current gradually clogs and fills the duct, and cooling leaves the solid mass. If by any cause, however, the lava should be choked at the source of the stream, I fancy that the portion already in the duct, protected to a great degree from cooling by the solid wall above and on either side would continue to flow for a considerable distance, and leave an empty space behind it. These ducts may be traced for several miles. The ground reverberates hollowly under

*Engineering and Mining Journal, vol. viii, No. 13, page 194. Overland Monthly, November, 1869.

the horses' feet; and at frequent intervals, where the crust has broken through by its own weight, the descent may be made over great blocks of basalt into the subterranean glooms. We explored one passage for a distance of 750 feet, and found the fissure still continuing at either end, though too narrow to admit of farther penetratration. But two or three of these caverns have been found to contain ice; and of these only one seems to afford it in abundance and in accessible position. This one we thoroughly examined, and found the subject, upon closer attention, divested of much of its mystery.

To make an ice-cave it is necessary to have a cave. This, as we have seen, is provided by the geological formation of the locality. The next requirement is a communication between the cave and the outer air, giving opportunity for the refrigerating draughts of winter. Finally, a slow percolation of water into the cave, which may freeze solid, in successive layers, throughout the winter. In summer the ice thus accummulated thaws slowly under the influence of warmer air entering the cavern ; but the thawing at one end produces a low temperature at the other, which preserves the great body of ice. Besides, it is a well-known fact that ice formed at very low temperatures will last much longer than that which has been barely frozen at 32°. Hence the amount of heat that would liquify river-ice only brings the cave-ice up to the neighborhood of the freezing-point. We satisfied ourselves, however, that the ice

in the cave does thaw in summer, and the water finds its way out by subterranean channels.

I am strongly inclined to see in these peculiar lava-ducts an explanation of the phenomenon, not unfrequent in this region, of "lost rivers." We camped on the excursion to the ice-caves by the side of a brisk, musical stream, which afforded us an abundance of water for our horses and ourselves. Mounting the next morning and riding away, we were surprised to find the bed of the stream less than a dozen rods below our camp perfectly dry. A closer examination showed that the water disappeared into one of the subterranean passages in the basalt. Somewhat similar must be the course of the great "lost river" which bursts out of the vertical side of the cañon of the Suake-a torrent from the solid rock; a foundling rather than a lostling, since it is the origin, not the fate of this river which is unknown. Above its strange headlong emergence from imprisonment, the black, barren desert stretches for scores of miles, treeless and waterless. Somewhere to the northward a river has been lost, and here it is. Identified it will be, so soon as its disappearance shall sufficiently interest any of its friends to make them answer the advertisement of its discovery.

The main cañon of the Snake River is cut through basalt for several hundred miles, and in following its course, by riding along the precipitous edge of the chasm, at the bottom of which flows the river, one has excellent opportunity to study both the successive deposition of the layers of basalt, with the signs of intermittence and intervals of rest, and the peculiar columnar structure, perpendicular to the planes of deposition, of each separate layer. Not even the noted example furnished by Fingal's Cave could surpass the illustrations afforded in many places by the columnar basalts of this region. The columns are usually six-sided prisms, though sometimes they have five or four sides. This variation, however, is enough to show that the cause of their formation is not analogous to crystalline force. The form of the regular pentagon does not occur in any crystalline form; nor does nature ever confound in the same substance the hexagonal and the tetragonal forms. Probably the shape of these prisms can be explained by a less difficult hypothesis. The cell of the bee, you will remember, is hexagonal in section, and was formerly cited as a wonderful example of geometrical instinct, since it possesses exactly the form which can be mathematically proved to occupy the space of the hive most economically, leaving no wasted interstices, and consuming a minimum quantity of building material. It is now seen that the cell of the bee is spherical in shape, and assumes this economical shape under pressure. It is probable that an analogous result is due in the cooling of basalt to the pressure of the whole mass. dency, apart from this pressure, would be to cool in globes; but the vertical pressure converts these into cylinders, facilitating at the same time vertical cleavages through the mass; and finally the lateral pressure through the half-solidified mass modifies these cylinders into prisms. It is natural that most of these should be hexagonal, because a circle can be tangent to six other circles of equal size, and these tangent each to each; and the transformation of these into hexagons would exactly fill the surface. But local disturbances, differences in the size of the circles forming the bases of the basalt-cylinders, and other accidental causes, would naturally lead to the formation of imperfect pristus, or such as had not their full complement of six sides. And this is the exact state of things as revealed by observation. The hexagonal type seems to be predominant, but not universal.

The ten

The size of the prisms or columns is not great-seldom exceeding one or two feet as the width of a side. Although the horizontal divisions between successive layers of basalt indicate long periods, possibly, of intervening time, and we can therefore not expect the columns to be continuous through the different layers, yet this is sufficiently the case (i. e., the cleavage planes between the prisms in one layer coincide sufficiently well

with those in the layer below) to permit extensive vertical cleavages of large masses of basalt, from the surface down through a dozen layers, to the very bottom of the cañon. I have stood above, with one foot upon the solid basalt, and the other upon the upper edge of a narrow slice thus parted from the main body, while between my feet a deep, long crevice, only a few inches in width, extended down to the base of the cliffs, giving me a glimpse of the river hundreds of feet below. The manner in which the cañon of the Snake appears to have been enlarged is by the gradual wearing away of the base of its walls by the flowing river and the grinding boulders, and then the toppling or sliding off of these massive slices from the whole face. The walls remain almost everywhere vertical, and the swift stream below carries away the talus of débris. This structure of the basalt greatly facilitates the formation of cañous in it by erosion. No doubt a stream first finding its way through some of the fine crevices left in cooling enlarges these at the bottom, and thus produces, little by little, a wide, smooth chasm. Indeed, it is not impossible that the lost river to which I have alluded may some day, industriously undermining its roof, appear as a visible torrent at the bottom of a deep cañon.

Many of these features can be studied with great ease, and under the inspiration of glorious natural beauty at the great Shoshone Falls of the Snake. (The lecturer here sketched rapidly upon the black-board the outline of these falls and of the basaltic cliffs above and below them.) There are no large pictures of these falls at present available; otherwise I should not be so audacions as to attempt to convey an idea of their form-still less of their beauty-in so rude a manner. They are sitnated about twelve miles above the crossing of the Snake Cañon by the stage-road from Boise to Salt Lake They are accessible without much hardship from the Pacific Railroad, by twenty-four hours' stage-travel to the station of Rock Creek, and a ride or walk across the sage-desert of half a dozen miles. The surface of the country appears to be an unbroken, scarcely undulating plain, dotted with the gray tufts of the sage, and black, here and there, with patches of bare volcanic rock. As one approaches the cañon of the river, its presence is betrayed by the sound of its flowing, until at last one comes suddenly upon its bold brink, and sees the stream 600 feet below.

The opposite side of the cañon presents an excellent vertical section, showing both the bedded and the columnar structure of the basalt, and thin layers of sedimentary deposits between the successive overflows. The whole thickness of the basaltic formation is about 400 feet; and below it is revealed porphyry, to the further depth of 200 feet. Through both of these the cañon has been carved; but it is apparently a much slower and more difficult work for the stream to wear away by honest friction the masses of porphyry than it was to subtly undermine and overthrow the stately pillars of basalt.

The roar of the falls is heard in the distance, and a rising cloud of mist indicates their locality. Riding along the edge of the cliff for two or three miles, we come upon one of the most romantic scenes of the world. We surprise the river at its work. Its basaltic channel was long since complete, and, not satisfied with that, it is patiently sawing, foot by foot, into the porphyry. Below the falls this rock has been excavated some 200 feet in depth; and this is almost exactly the height of the falls themselves. There are almost no rapids immediately above, and none at all immediately below; the stream makes practically but one leap, in a sheet, broken at low water by projecting rocks, but majestic in times of flood as the segment of a huge, revolving wheel. Figures are per haps as impressive as words in such a description, since they leave the imagination freer play in filling up the outlines of the scene. A cañon perhaps 1,000 feet in width, 400 feet deep above the falls, 600 feet below, and a great river plunging from one bed to the other-this is the frame of the picture. As if unwilling to leave so grand a beginning without some delicate touches of milder beauty, nature has relaxed the sternness of her desolation, and clothed the gray ruins of the precipices with green trees and grass, nourished by the mists of the cataract. The sun, busy through all the surrounding plains in fierce destruction, here condescends to the graceful labor of scattering diamonds in the foam and painting rainbows on the mist. A little way above the falls the river divides, inclosing between its two arms a remnant of the overlying basalt, massive and castellated, like a great fortress, defying destruction; and almost on the edge of the falls, protected no doubt by this fortress farther up the stream, stands a still smaller relic of the former rocks, a pillar of basalt, upon the top of which was seen by the earliest pioneers who penetrated to this place an eagle's nest. For a score of years it has remained-how much longer no man can tell-and is still inhabited by the proud and solitary pair whose reign there is none to dispute.

Despite the forbidding appearance of the precipitous cliffs, it is possible to descend into the cañon below the falls. Numerous clefts and fissures extend through the basalt to the porphyry, and by scrambling, sliding, and dropping, aided by the stout branches of drooping trees, it is possible to reach the bottom. Many of these fissures are spanned above by natural bridges. Some of them are closed at the top, with openings here and

*The height of the cliffs below the falls is 620 fee; the height of the falls is said to be 216 feet.

there, through which we may look down into deep narrow caves, with glimpses, far below, of sunshine and the flowing river.

The view of the falls from beneath, though of course impressive and beautiful, is not the best that can be obtained. The characteristic feature of the scene, the great depth. of the cañon above the upper level of the stream, is dwarfed in perspective, and the beauty of the stream just before its final plunge is hid from view by the tumbling waters themselves. The most magnificent aspect, combining in one picture nearly all the elements of power and grace, of bold outline and tender shadow, of towering height above and dizzy depth below, is obtained from a jutting point of the crags below the fall and nearly on a level with its edge. It was from such a position that Mr. Clarence King's party of explorers, suspending their photographer and his instrument in mid-air, as it were, in front of the great cataract, obtained an excellent picture of this unique

scene.

I have said enough, I trust, to convince you that the far Northwest has much to show us and to tell us that will be valuable to science and to art. I hope it will not be long before its natural features will be thoroughly explored and described. Pacific Railroad parties have crossed and recrossed it, intent rather on the best way to get through its passes and across its wilderness than on the study of its character and history; the miner and the emigrant have traversed it with indifference, save where it promised fruitful farms or golden treasure; the Indian and the trapper have roamed over it, with keen but ignorant observation. But the time for all such imperfect explorations of our great interior has passed. The recent governmental survey of the belt through which the Pacific Railroad runs is a specimen of a more thorough, elaborate, and permanently valuable contribution to our knowledge of the continent. I hope it may be followed by others, as well-equipped and as well-conducted as, this has been by its young but famous leader, Clarence King, with whose name those of Hague and Gardner are worthily associated. Surely the government of this great country cannot do a wiser or a nobler thing than to give to the world a faithful picture of the mighty realm beneath its sway.

Since the foregoing description was written and made public, the rediscovery of gold in the bars of the Upper Snake has led a considerable temporary population to the neighborhood of the great falls.

THE SNAKE RIVER BARS.

The bars on the Snake River have long been the resort of placerminers at times when the lack of water caused the suspension of operations in the camps of Boise Basin and other localities, usually toward the close of the season, when the smaller streams are dry. Indeed, it is not practicable to carry on bar-mining in the Snake while the stream is high; and even under the circumstances above mentioned, it is usually expected that the bars will merely enable a few hundred men to earn "wages"-$5 to $8 per day-and so find steady employment at a time when they would otherwise be forced to remain idle. The seasons of 1869 and 1870, however, were both characterized by drought; and the unusual low stage of water early in the summer, while it impaired the productiveness of many interior camps, gave rise to increased activity along the great river. During the former year I followed the Snake cañon for some distance, visiting en route the celebrated Shoshone Falls; and it must be confessed that I found its precipitous basaltic walls, with their coarse débris, not promising for river-mining. Near these very falls, however, some discoveries of gold in the river-channel, made in the summer of 1870, created considerable excitement, and called together suddenly a temporary population, where at the time of my visit there was no sign of human habitation. From the information I have been able to gather, I am led to believe that the gold deposits of this part of the river are limited in extent and quite moderate in their yield of precious metal, while the operations of mining are difficult in the most favorable seasons, and will be quite impracticable when high water prevails. The following account of these mines, extracted and condensed from a letter addressed July 15, 1870, to the San Francisco