An Expedition to Mount St. Elias, Alaska

Our stay above the snow-line had lasted thirty-five days, and we were extremely glad to see the light of a camp-fire and have the trees and flowers about us once more. The vegetation indicated that the season was already far advanced. Most of the flowers had faded, and autumn tints gave brilliancy to the lower mountain slopes; salmon berries and huckleberries were in profusion, and furnished an exceedingly agreeable change in our diet. After a bath in one of the small lakelets on the island and a good night's rest on a luxuriant bed of spruce boughs, we felt fully restored and ready for another campaign.

As Kerr was anxious to get back to Port Mulgrave, it was arranged that Lindsley and Partridge should go with him, and that the rest of the men should remain. Kerr took his departure on the morning of September 7, and on the following day Christie, Doney, and myself crossed the Marvine glacier to the southern end of the Hitchcock range, and the following day made an excursion out upon the Malaspina glacier. The day of our excursion was bright and beautiful, and the mountains to the northward revealed their full magnificence. The level plateau of ice formed a horizontal plain, from which the mountain rose precipitously and appeared grander and more majestic than from any other point of view. St. Elias rose clear and sharp, without a cloud to obscure its dizzy height, and appeared to be one sheer precipice. It is doubtful if a more impressive mountain face exists anywhere else in the world. After learning all we could concerning the Malaspina glacier we returned to our camp at the end of the Hitchcock range, and the following day tramped across the extremely rough moraine-covered surface back to Blossom island.

The following morning, September 12, we started on our return trip to Yakutat bay. Two small tents and many articles for which we had no further use were abandoned, so as to make our packs light as possible. We crossed the Hayden glacier, and at night camped at the foot of Floral pass. After making two intermediate camps, traveling each day in the rain, we reached the shore of Yakutat bay on September 15.

Doney and I halted at Dalton's cabin for the purpose of seeing what we could of the openings there made for coal, while the rest of the party pressed on to our old camping place on the shore. There they found Kerr and his party still encamped, but ready to leave for Port Mulgrave early the next morning.

September 18 was occupied by us in catching salmon and trout. We were abundantly successful, as every man returned to camp with all that he could carry. These were spread out on a rack over our camp-fire and smoked for further use, as we did not know how long our stay would be extended. On the next day Stamy and Lindsley returned from Port Mulgrave, where they had left Kerr, quite recovered from his exposure on the mountain. Stormy weather continued, and a gale from the northeast piled the ice high on the beach and threatened to sweep away our tents, as has already been briefly described in earlier pages.

On September 20, our tents having been beaten in by a violent storm and our camping place overflowed by the waters from a lake above us, we removed our goods to a place of safety and went to Dalton's cabin, where we awaited better weather. The morning of September 23 dawned clear and bright, and after drying our clothes around a blazing camp-fire, we started back to our camping place on the shore. Before reaching there, however, we were rejoiced to see the Corwin coming up the bay. It took us but a short time to get on board, where Captain C. L. Hooper, her commander, did everything in his power to make us welcome and comfortable. To him we are indebted for a delightful voyage back to civilization.

After steaming up Disenchantment bay nearly to the ice-cliffs of the Hubbard glacier, and obtaining a fine view of the glaciers about Disenchantment bay, the Corwin returned to Port Mulgrave and, on September 25, put to sea. After a splendid ocean passage, we arrived at Port Townsend on October 2.

During our stay in Alaska not a man was seriously sick and not an accident happened. The work planned at the start was carried out almost to the letter, with the exception that snow-storms and the lateness of the season did not permit us to reach the summit of Mount St. Elias.

Should another attempt be made to climb Mount St. Elias, the shortest and most practicable route from the coast would be to land at Icy bay and ascend the Agassiz glacier. The course taken by us in 1890 could be intersected just north of where the tributary glacier from Dome pass joins the main ice-stream; and from there the route followed last summer would be the most practicable. A camp should be established on the divide between Mount St. Elias and Mount Newton, from which excursions to either of these peaks could be made in a single day.

In the preceding narrative many details have been omitted. One of these is that tents, together with blankets, rations, etc., were left at two convenient points between Blossom island and the Agassiz glacier, and were used by the men in bringing up supplies. In attempting to ascend Mount St. Elias from Icy bay by the route suggested, at least three such relay stations should be established between the Chaix hills, where wood for camp-fires can be obtained (as is known from the reports of the New York Times and Topham expeditions), and the high camp on the divide. The relay camps suggested should be one day's march apart, and would serve not only for stopping places while carrying rations during the advance, but would furnish a line of retreat. A party making this journey should be provided with snow-shoes, which unfortunately we did not take with us.

All rations intended for use above the snow-line should be packed in tin cans, each of sufficient size to hold between fifty and sixty pounds, and each should be securely soldered. All articles packed in this way should be thoroughly dry and should be packed in a dry, warm room. When secured in this manner they are about as easy to carry as if packed in bags, and can be "cached" anywhere out of the reach of floods and avalanches, with the certainty of being serviceable when wanted. The more perishable articles to be used where camp-fires are possible should also be secured in tin cans. Sacks of flour, corn-meal, etc., should be protected by an outer covering of strong canvas. The experience of last summer showed that the cans of rations intended for use above the snow-line should each contain about the following ration, which may be varied to suit individual taste:

Our experience with oil-stoves showed that they are serviceable. While on the march they can be carried as hand packs in gunny-sacks. Rectangular cans holding about a gallon each, with small screw-tops, were found convenient for carrying coal oil. The experience of Arctic explorers indicates that alcohol would perhaps be better than coal oil to use in snow-camps.

Among the most important articles to be provided are strong shoes or boots; of these each man should have at least two pairs. Strong hip-boots, with lacings over the instep, are exceedingly serviceable. When sleeping on the ice the boot-legs may be spread beneath one's blankets and the feet used as a pillow. The long legs are serviceable alike in the thick brush on the shore and in the deep snow on the high mountains. With their protection, many streams can be waded without getting wet. Leather, waxed ends, awls, etc., for repairing boots, and tallow mixed with bees-wax for greasing them, should be taken and distributed in part through the cans of rations. Heavy woolen socks are indispensable, and an effort should be made to have a dry pair always at hand. This may be arranged, even under the most unfavorable conditions, by drying a pair as thoroughly as is convenient and carrying them in the bosom of one's shirt.

Long alpenstocks are always necessary. My own choice is a stiff one of hickory, about six feet long and an inch and a quarter in diameter, provided with a spike and hook at one end and a chisel about two inches broad at the other. Ice axes are desirable while climbing in the high mountains, but even more serviceable are light axes of the usual pattern, but with handles about fourteen inches long; these supplement the alpenstock, and when not actually in use are carried in the packs.

Each man should be provided with a water-tight match-box, and should have, besides, a bundle of wax matches wrapped in oil-cloth and sewed in the collar of his shirt, to be held as a last reserve. Each man should also have a small water-tight bag in which to carry salt enough to last a week or ten days, in case he has to live by hunting or fishing. A heavy hunting knife is very convenient, and can be used not only in cutting trails through thick brush, but in cases of necessity is serviceable in making steps in ice. Heavy woolen clothing is preferable to furs. Sleeping bags were not used during our expedition, but are highly recommended by others. For protection at night, a thick woolen blanket with a light canvas cover and a sheet of light rubber cloth to protect it are all that is necessary. Our tents were of cotton drilling, seven feet square and about six feet high, and provided with ridge-ropes. Alpenstocks were used for tent poles. "Sou'westers" and strong water-proof coats are indispensable in a climate like that of Alaska, and at night may be used as a substratum on which to sleep. While traveling over the snow-line we used colored glasses to protect the eyes, and also found that a strip of dark mosquito netting tied across the face below the eyes afforded great protection. Some of the party found relief from the glare of the snow by blacking their faces with grease and burnt cork, but one experiment with that method is usually enough. While camping below timber-line during the months of June to September fine mosquito netting is indispensable. In carrying packs, hemp "cod-line" of the largest size was found to answer every requirement, and is preferred by expert packers to pack-straps.

It has been suggested that experienced Swiss guides are necessary to ensure success in climbing Mount St. Elias. Having never followed a guide in the mountains, I am not able to judge of their efficiency, but it must be remembered that no one can guide in a region that has never been traversed. The "guide" as understood in Europe is unknown in America. In the exploration of this country by engineers, geologists, etc., the camp hands have followed their leaders and have not shown them the way. In every frontier town there are hunters, trappers, miners, prospectors, cow-boys, voyageurs, etc.—men who have passed their lives on the plains or among "the hills" and are enured to hardship and danger. This is the best material in the world from which to recruit an exploring party. A foreigner engaging the services of such men must take into account the independent spirit that animates them and is the secret of their usefulness. They are not servants, but retainers; that too in regions far beyond the reach of civil law. They will follow their leader anywhere, support him in all dangers, and do their work faithfully so long as their rights as men are respected.

By taking proper precautions while traveling across crevassed snow and ice, and guarding against avalanches and snow-blindness, an excursion can be made above the snow-line with as little danger as in better known and more frequented regions.

PART III.

SKETCH OF THE GEOLOGY OF THE ST. ELIAS REGION.

In the preceding narrative, many references have been made to the character of the rocks and to the geological structure of the region explored. It was not practicable during the journey to carry on detailed geological studies, but such facts as were noted are of interest, for this reason, if for no other: they relate to a country previously unknown.

My reconnoissance enabled me to determine that there are three well-defined formations in the St. Elias region. These are—

1. The sandstones and shales about Yakutat bay and westward along the foot of the mountain to Icy bay, named the Yakutat system.

2. A system of probably later date, composed of shale, conglomerate, limestone, sandstone, etc., best exposed in the cliffs of Pinnacle pass and along the northern and western borders of the Samovar hills, and named the Pinnacle system.

3. The metamorphic rocks of the main St. Elias range, called the St. Elias schist.

The rocks of this system are of gray and brown sandstones and nearly black shales. They are uniform in lithological character over a large area, and are usually greatly crushed and seamed. So great has been the crushing to which they have been subjected that it is difficult to work out a hand specimen with fresh surfaces. Fragments broken out with a hammer are almost invariably bounded by plains of previous crushing, and are usually somewhat weathered.

These rocks form the bold shores of Yakutat and Disenchantment bays, and were the only rocks seen along our route from Yakutat bay to Pinnacle pass. The whole of the Hitchcock range is composed of rocks of this series, as are also the Chaix hills and the hills west of Icy bay and the southern portion of the Samovar hills. North of Pinnacle pass there are rocks undistinguishable lithogically from those about Yakutat bay. These are exposed in Mount Owen and on each side of Dome pass; they also form the bold spurs about the immediate bases of Mount Augusta, Mount Malaspina, and Mount St. Elias. In the three instances last named these rocks dip beneath the schist forming the crest of the St. Elias range, and it is probable that a great overthrust there took place before the formation of the faults to which the present relief of the mountains is due.

All the mountain spurs of Mount Cook, so far as is known, are composed of sandstones and shales of the Yakutat series, with the exception of the Pinnacle pass cliffs. Nearly all the débris on the glaciers from Disenchantment bay to the Seward glacier, and probably beyond, is derived from the rocks of this system. The distribution of the rocks from which the débris was derived may be ascertained in a general way by tracing out the sources of the glaciers. Medial moraines on the Hayden and Marvine glaciers, however, have their sources on the northern slope of Mount Cook, and are composed of gabbro and serpentine. These rocks were not seen in place, and their relation to the Yakutat series can only be conjectured.

Although the rocks of this system are stratified, it is impossible to determine their thickness, for the reason that they have been greatly crushed and overthrust. This is well illustrated in the Hitchcock range, which, as already explained, trends about northeast and southwest, and is composed of strata of shale and sandstone, having a nearly east-and-west strike and a uniform dip toward the northeast. Were the rocks in normal position their thickness would be incredible. In addition to this negative evidence, there is the crushed condition of the strata to show that movement has taken place all through their mass; and in a few instances thrust faults were distinguished, dipping northeastward at about the same angle as the lines of bedding. In the crushing to which the rocks have been subjected the shales have suffered more than the sandstones, and have been drawn out into wedge-shaped masses, the sharp edges of which usually point toward the northeast, which is presumably the direction from which the crushing force acted.

The hypothesis that the rocks in the St. Elias region have been crushed and overthrust explains many otherwise inharmonious facts, and accounts for the superposition of the St. Elias schist upon rocks of the Yakutat system.

Coal has been discovered in the rocks of the Yakutat system about two miles west of the southern end of Disenchantment bay, and is reported to be of workable thickness. I saw thin lignite seams at the surface at this locality, but as the shafts were filled with water I was unable to examine the coal in the openings, and cannot vouch for its thickness. Samples obtained from the mine show it to be a black lignite which would apparently be of value for fuel. Fossil leaves are reported to occur in connection with the lignite, but these have never been seen by any one who could identify them.

The rocks of the Yakutat system, wherever seen, dip northeastward, except when greatly disturbed near fault-lines. East of Disenchantment bay the inclination of the beds is from 15° to 20°; farther westward the dip increases gradually all the way to the Hitchcock range, where the prevailing inclination is from 30° to 40°, and frequently still greater. Beneath Mount Malaspina and Mount St. Elias the Yakutat sandstones dip northeastward at an angle of about 15°, and in the hills west of Icy bay the dip is about the same. Exceptions to the prevailing dips occur along the immediate shore of Yakutat bay, northwest of Knight island, and at the southern extremity of each of the mountain spurs between Yakutat bay and Blossom island. At these localities the rocks are frequently vertical or nearly so, owing their high dip to the proximity of lines of displacement. The faults indicated by these unusual dips also mark the boundary between the mountains and the seaward-stretching plateau of alluvium and ice.

The crushing, overthrusting and faulting that has affected the rocks of this system render it doubtful whether the coal seams which occur in it, even if of requisite thickness, can be worked to advantage. Some of the samples of coal obtained at the openings made near Yakutat bay were slickensided, showing that movements in the coal seam had there taken place.

As already stated, the rocks of the Yakutat series are remarkably uniform in character throughout the extent now known, and offer but little variety. The sandstones are intersected in every direction by thin quartz seams, which stand in relief on the weathered surfaces, giving the rocks a peculiar and characteristic appearance. The first important change in the geology along the route traversed by us was met on reaching Pinnacle pass.

The rocks of this system, as already stated, are best exposed in the great fault-scarp forming the northern wall of Pinnacle pass. They are more varied in composition and have preserved a better record of the conditions under which they were deposited than the sandstones and shales of the Yakutat system.

Only an approximate section of the rocks exposed in the Pinnacle-pass cliff was obtained.

There is also a compact, crystalline, gray limestone near the upper portion of the series, which escaped notice in the cliffs. At the end of the Pinnacle-pass cliffs, however, where the rocks are turned northward by the great fault which decides the course of the Seward glacier, and dip eastward at a high angle, the limestone is well exposed, and has a thickness of about 50 feet. In many places the surfaces of the layers are covered with fragments of large Pecten shells. Associated with the limestone there are reddish shales, much crushed and broken, and a peculiar conglomerate. The pebbles in the conglomerate are of many varieties, and were observed at places along the Pinnacle pass cliffs. Their most marked peculiarity lies in the fact that they have been sheared by a movement in the rocks and sometimes broken into several fragments which have been reunited, probably by pressure. These faulted pebbles are characteristic of the strata from which they were derived. Similar pebbles were afterward obtained in the Marvine glacier near its junction with the Malaspina glacier, thus indicating that there are other outcrops of the conglomerate about Mount Cook, near where the Marvine glacier has its source. Two quartz pebbles from the conglomerate of Pinnacle pass are shown in the accompanying illustrations. The larger pebble (shown in figure 7) is of bluish-gray quartz, and the smaller one (depicted in figure 8) is of white quartz. The fragments into which they have been broken are now firmly united. The engravings are photo-mechanical (Moss process) reproductions from the objects.

The presence of a Pecten (P. caurinus (?) Gld.) in the limestone of the Pinnacle series has already been mentioned. Other fossils were obtained from sandstones and shales at the crest of the cliffs above Pinnacle pass at an elevation of 5,000 feet. These were submitted to Dr. W. H. Dall, who kindly identified them as follows:

Mya arenaria, L.;
Mytilus edulis, L.;
Leda fossa, Baird, or L. minuta, Fabr.;
Macoma inconspicua, B. and S.;
Cardium islandicum, L.;
Litorina atkana, Dall.

All of these species are stated by Dall to be still living in the oceanic waters of Alaska. The very recent age of the rocks in which they occur is thus established.

In strata closely connected with the layers in which these shells were found there occur many fine leaf impressions, a few of which were brought away. These have been examined by Professor L. F. Ward, who has identified them with four species of Salix, closely resembling living species. The report on these interesting fossils forms Appendix D.

The age indicated by both invertebrates and plants is late Tertiary (Pliocene) or early Pleistocene. This determination is of great significance when taken in connection with the structure of the region, and shows that the mountains in the St. Elias region are young.

Not only was a part, at least, of the Pinnacle system deposited during the life of living species of mollusks, but also the whole of the Yakutat series, the stratigraphic position of which is, if my determination is correct, above the Pinnacle system. After the sediments composing the rocks of these two series were deposited in the sea as strata of sand, mud, etc., they were consolidated, overthrust, faulted, and upheaved into one of the grandest mountain ridges on the continent. Then, after the mountains had reached a considerable height, if not their full growth, the snows of winter fell upon them, and glaciers were born; the glaciers increased to a maximum, and their surfaces reached from a thousand to two thousand feet higher than now on the more southern mountain spurs, and afterward slowly wasted away to their present dimensions. All of this interesting and varied history has been enacted during the life of existing species of plants and animals.

The relative age of the Yakutat and Pinnacle series is the weakest point in the history sketched above. The facts on which it rests are as follows: At Pinnacle pass the sandstones and shales forming the southern wall belong to the Yakutat system and are much disturbed, while the northern wall, or the heaved side of the fault, is composed of the rocks of the Pinnacle system, inclined northward at an angle of 30° or 40°. North of this fault-scarp, in the foothills of Mount Owen, sandstones and shales, seemingly identical with those of the Yakutat system, again occur, although their direct connection with the rocks south of Pinnacle pass was not observed, owing to the snow that obscured the outcrops. Again at Dome pass a similar relation seems evident, but cannot be directly established. The immediate foothills of Mounts Augusta, Malaspina, and St. Elias are also of sandstone, lithologically the same as the Yakutat series. The conclusion that the Yakutat system is younger than the Pinnacle-pass rocks was reached in the field after many other hypotheses had been tried and found wanting, and to my mind it explains all the observations made. Even should the supposed relations of the two series under discussion be reversed, it would still be true that a very large part of the rocks of the St. Elias region were deposited since the appearance of living species of mollusks and plants, and that the prevailing structure of the region was imposed at a still later date. This will appear more clearly after examining the structure of the region.

The rock forming several thousand feet of the upper portion of the St. Elias range is a schist in which the planes of bedding are preserved. The dip of the strata is northeastward, and has exerted a decided influence on the weathering of the mountain crests. As the opportunities for examining this formation were unsatisfactory, a detailed account of it will not now be attempted.

The abnormal thickness of the Yakutat series, due to crushing and overthrust, has been referred to, as has also the superposition of the St. Elias schist upon rock supposed to belong to the Yakutat system.

The plane of contact between the sandstone and the overlying schist of the St. Elias range dips northeastward at an angle of about 15°, corresponding, as nearly as can be determined, with the dip of the strata in the sandstone itself. All of the observations made in this connection indicate that the schist has been overthrust upon the sandstones. After this took place the great faults to which the range owes its present relief were formed.

About Mount Cook, however, and in the elevated plateau east of Yakutat bay, the conditions are different from those observed along the base of the St. Elias range. The only displacements known in the Yakutat system south and east of Pinnacle pass is the great fault which presumably exists where the rocks of the foothills disappear beneath the gravel and glaciers of the Piedmont region, the faults referred to belonging to the same series as those which determine the southern and southwestern borders of the St. Elias range and many of the foothills south of the main escarpment. Besides the great faults which trend from St. Elias toward the northeast and northwest, there are several cross-faults, one of which determines the position of the Seward glacier through a portion of its course, while another marks out the path of the Agassiz glacier; and two others may be recognized just east of the summit of St. Elias, which have dropped portions of the eastern end of the orographic block forming the crowning peak of the range.

The southern face of Mount St. Elias is a fault-scarp. The mountain itself is formed by the upturned edge of a faulted block in which the stratification is inclined northeastward. As has just been mentioned, the mountain stands at the intersection of two lines of displacement, one trending in a northeasterly and the other in a northwesterly direction. The one trending northwestward extends beyond the end of the northeast fault. The point of union is at the pass between Mount St. Elias and Mount Newton. The upturned block, bounded on the southwest by a great fault, projects beyond the junction with the northeasterly fault. It is this projecting end of a roof-like block that forms Mount St. Elias. That this is the case may be clearly seen when viewing the mountain from the glacier near the base of Mount Owen. Such a view is shown on plate 20. The crest-line of St. Elias extends with a decreasing grade northwestward from the culminating peak, and the northern slope of the ridge is the surface of the tilted block.

PART IV.

GLACIERS OF THE ST. ELIAS REGION.

The glaciers of the St. Elias region form two groups. The ice-streams from the mountain are of the type found in Switzerland, and hence termed Alpine glaciers. The great plateau of ice along the ocean formed by the union and expansion of Alpine glaciers from the mountains belongs to a class not previously described, but which in this paper have been called Piedmont glaciers. The representative of the latter type between Yakutat bay and Icy bay is the Malaspina glacier. Both types are to be distinguished from Continental glaciers.

The glaciers in the mountains are all of one type, but present great diversity in their secondary features, and might be separated into three or four subordinate divisions. The great trunk glaciers have many tributaries, and drain the snows from the mountains through broad channels, which are of low grade throughout all the lower portions of their courses. Besides the trunk glaciers and the secondary glaciers which flow into them, there are many smaller glaciers which do not join the main streams, but terminate in the gorges or on the exposed mountain sides in which they originate. These have nearly all the features of the larger streams, but are not of sufficient volume to become rivers of ice.

A minor division of Alpine glaciers for which it is convenient to have a special name includes those that end in the sea and, breaking off, form icebergs. These may be designated as "tide-water glaciers." Typical examples of this class are furnished by the Dalton and Hubbard glaciers, but other ice-streams having the same characteristics occur in Glacier bay, in Taku inlet, and at the heads of several of the deep fjords along the coast of southeastern Alaska.

A noticeable feature of the Alpine glaciers of Alaska is that they expand on passing beyond the valleys through which they flow and form delta-like accumulations of ice on the plains below. This expansion takes place irrespective of the direction in which the glaciers flow, and, so far as may be judged from the many examples examined, is independent of the débris that covers them. It should be remembered, however, that none of the Alaskan glaciers thus far studied show marked inequalities in the distribution of the moraines upon their surfaces. Should one side of a glacier, on leaving a cañon, be heavily loaded with marginal moraines, while the opposite border was unprotected, it is to be presumed that a deflection of the ice would take place similar to the change in direction recorded by the moraines about Mono lake, California.³⁴ The normal tendency of ice, when not confined, to expand in all directions and form a plateau is illustrated on a grand scale by the Malaspina glacier.

³⁴ Eighth Ann. Rept. U. S. Geol. Surv., 1889, part I, pp. 360–366.

The most important ice-streams about Mount St. Elias and Mount Cook are indicated on the map forming plate 8. The Tindall, Guyot, and Libbey glaciers and the lower part of the Agassiz glacier there represented are taken from a map published by H. W. Topham.³⁵ All of the other glaciers indicated on the map were hastily surveyed during the present expedition and are described to some extent in the accompanying narrative. By far the most important of these is the one named the Seward Glacier.

³⁵ Alpine Journal, London, vol. XIV, 1887, pl. op. p. 359.

The Seward Glacier is of the Alpine type, and is the largest tributary of the Malaspina glacier. Its length is approximately 40 miles, and its width in the narrowest part, opposite Camp fourteen, is about 3 miles. The main amphitheatre from which its drainage is derived is north of Mount Owen and between Mount Irving and Mount Logan. The general surface of the broad level floor of this névé field has an elevation of approximately 5,000 feet. The snow from the northern and western sides of Mount Irving, from the northern slope of Mount Owen, and from numerous valleys and cañons in the vast semicircle of towering peaks joining these two mountains, unite to form the great glacier. There is another amphitheatre between Mount Owen and the Pinnacle pass cliffs supplied principally by snows from the northwestern slope of Mount Cook, which sends a vast flood of ice and snow into the main drainage channel. Other tributary glaciers descend the steep slopes of Mount Augusta and Mount Malaspina, and a lesser tributary flows eastward from Dome pass. All of these ice-drainage lines converge toward the narrow outlet of Camp 14 (plate 8) and discharge southward down a moderately steep descent several miles in length. Below Camp 14 there are other névé fields bordering the glacier, which contribute no insignificant amount of ice and snow to its mass. Between the extremity of the Hitchcock range and the Samovar hills the path of the glacier is again contracted and greatly broken as it descends to the plateau below.

The Seward glacier, like all ice rivers of its class, has its névé region above, and its ice region below. The limit between the two is the lower margin of the summer snow, and occurs just above the ice-fall between the southern extremity of the Hitchcock range and the Samovar hills. All the névé region is pure white and without moraines, except at the immediate bases of the most precipitous cliffs. At the bases of the Corwin cliffs, which rise fully 2,000 feet above its border, no débris can be distinguished even in midsummer. An absence of moraines along the base of Pinnacle pass cliffs was also noticed during our first visit, but when we returned over the same route in September the melting of the snow had revealed many large patches of dirt and disintegrated rock. In several places near the bases of steep cliffs, strata of dirty ice, containing many stones, were observed in deep crevasses. It was evident that vast quantities of débris were sealed up in the ice along the borders of the glacier, only to appear at the surface far down the stream where summer melting exceeds the winter accumulation.

The surface of the glacier below the lower fall is composed of solid ice with blue and white bands, and has broad moraines along its borders. The course of the glacier, after entering the great plateau of ice to which it is tributary, may be traced for many miles by the bands of débris along its sides. These moraines belong to the Malaspina glacier, and have already been referred to.

At the outlet of the upper amphitheatre, about 6 miles above Mount Owen, there is an ice-fall which extends completely across the glacier. Below the pinnacles and crevasses formed by this fall the ice is recemented and flows on with a broad, gently descending surface, gashed, however, by thousands of crevasses, as shown in plate 20, to the end of the Pinnacle pass cliffs. It there finds a more rapid descent, and becomes crevassed in an interesting way. The slope is not sufficient to be termed a fall, but causes a rapid in the ice-stream.

The change of grade in the bed of the glacier is first felt about a mile above Camp 14. A series of crevasses there begins, which extends four or five miles down-stream. At first the cracks are narrow, and trend upstream in the manner usual with marginal crevasses. Soon the cracks from the opposite sides meet in the center and form a single crevasse, bending upstream in the middle. A little lower down, the crevasse becomes straight, showing that the ice in the center of the current flows more rapidly than at the sides. The more rapid movement of the center is indicated by the form of the crevasses all the way down the rapid. After becoming straight they bow in the center and form semi-lunar gashes, widest in the center and curving up-stream at each extremity. Still farther down they become more and more bent in the center and at the same time greatly increased in breadth. Still lower the curve becomes an angle and the crevasses are V-shaped, the arrow-like point directed down-stream. These parallel V-shaped gashes set in order, one in front of the other, are what gives the glacier the appearance of "watered" ribbon when seen from a distance.

With the change in direction and curvature of the crevasses, there is an accompanying change in color. The cracks in the upper part of the rapid are in a white surface and run down into ice that looks dark and blue by contrast. Lower down, as the cracks increase in width, broad white tables are left between them. Cross-fractures are formed, and the sides of the table begin to crumble in and fill up the gaps between. As the surface melts the tables lose their pure whiteness and become dust-covered and yellow; but the blocks falling into the crevasses expose fresh surfaces, and fill the gulfs with pure white ice. In this way the color of the sides of the crevasses changes from deep blue to white, while the general surface loses its purity and becomes dust-covered. Far down the rapid where the V-shaped crevasses are most pointed, the tables have crumbled away and filled up the gulfs between, so that the watered-ribbon pattern is distinguished by color alone. The scars of the crevasses formed above are shown by white bands on a dark dust-covered surface. Before the lower fall is reached nearly all traces of the thousands of fissures formed in the rapids above have disappeared.

On looking down on the rapids from any commanding point, the definite arrangement of the crevasses along the center of the ice-stream at once attracts attention, and their order suggests a rapid central current in the stream.

Below Camp 14, for at least two or three miles, as well as at many places above that point, the Seward glacier flows between banks of snow. Along its border there are marginal crevasses trending up-stream, and in the adjacent banks there are similar breaks trending down-stream. Where the two systems meet there is a line of irregular crevasses, exceedingly difficult to cross, which mark the actual border of the flowing ice. A similar arrangement of marginal crevasses and of shore crevasses has been referred to in connection with the Marvine glacier, and was observed in many other instances.

While occupying Camp 14 we could hear the murmur of waters far down in the glacier below our tent, but there were no surface streams visible. Crashing and rumbling noises made by the slowly moving ice frequently attracted our attention, and sometimes at night we would be awakened by a dull thud, accompanied by a trembling of the rocks beneath us, as if a slight earthquake had occurred. Occasionally a pinnacle of ice would fall and be engulfed in the crevasses at its base. These evidences of change indicated that movements in the Seward glacier were constantly in progress. A short base-line was measured and sights taken to well-marked points in the Seward glacier for the purpose of measuring its motion. The angles between the base-line and lines of sight to the chosen points were read on several successive days, but when these observations were compared they gave discrepant results. The measurements which seemed most reliable indicate that the central part of the ice-stream has a movement of about twenty feet a day. This is to be taken only as an approximation, which needs to be verified before much weight can be attached to it.