The course of the Shayuk was visible for several miles, running nearly due west. Beyond that distance, it disappeared among rocky hills. Fording the river, I ascended a steep bank, to get upon a stony platform, over which I proceeded in a northerly direction, gradually approaching a small stream which came from the north to join the Shayuk. Passing a low rounded hill to the right, I descended after about two miles into the ravine excavated by this little stream, and, crossing it, encamped under low limestone rocks on its right bank after a march of twelve miles. I did not ascertain the elevation of this halting-ground, but, from the result of an experiment made at a place which appeared nearly midway (in point of elevation) between it and the bed of the Shayuk, where I got a boiling-point, indicating an elevation of 17,000 feet, I estimate the bed of the river at 16,800 feet, and my encamping-ground of the 18th at 17,200 feet. The plain all round seemed destitute of vegetation, so that, as on the two last days, there was a great scarcity of fuel, which had to be collected from a distance of many miles; and consisted only of the roots of a small bushy Artemisia or Tanacetum, which rose three or four inches above the ground. During these three days, I suffered very considerably from the effects of the rarefaction of the air, being never free from a dull headache, which was increased on the slightest exertion.

On the 19th of August, leaving my tent standing, I started to visit the Karakoram pass, the limit of my journey to the northward. The country round my halting-place was open, except to the north, where a stream descended through a narrow valley from a range of hills, the highest part of which was apparently about 3000 feet above me. All the rivers had formed for themselves depressions in the platform of gravel which was spread over the plain. At first I kept on the south bank of the river close to which I had halted, but about a mile from camp I crossed a large tributary which descended from the south-west, and soon after, turning round the rocky termination of a low range of hills, entered a narrow valley which came from a little west of north-west. At the foot of the rocky point of the range were three very small huts, built against the rock as a place of shelter for travellers, in case of stormy or snowy weather; and bones of horses were here scattered about the plain in greater profusion than usual.

I ascended this valley for about six miles: its width varied from 200 yards to about half a mile, gradually widening as I ascended. The slope was throughout gentle. An accumulation of alluvium frequently formed broad and gently sloping banks, which were cut into cliffs by the river. Now and then large tracts covered with glacial boulders were passed over; and several small streams were crossed, descending from the northern mountains through narrow ravines. About eight miles from my starting-point the road left the bank of the stream, and began to ascend obliquely and gradually on the sides of the hills. The course of the valley beyond where I left it continued unaltered, sloping gently up to a large snow-bed, which covered the side of a long sloping ridge four or five miles off. After a mile, I turned suddenly to the right, and, ascending very steeply over fragments of rock for four or five hundred yards, I found myself on the top of the Karakoram pass—a rounded ridge connecting two hills which rose somewhat abruptly to the height of perhaps 1000 feet above me. The height of the pass was 18,200 feet, the boiling-point of water being 180·8°, and the temperature of the air about 50°. Towards the north, much to my disappointment, there was no distant view. On that side the descent was steep for about 500 yards, beyond which distance a small streamlet occupied the middle of a very gently sloping valley, which curved gradually to the left, and disappeared behind a stony ridge at the distance of half a mile. The hills opposite to me were very abrupt, and rose a little higher than the pass; they were quite without snow, nor was there any on the pass itself, though large patches lay on the shoulder of the hill to the right. To the south, on the opposite side of the valley which I had ascended, the mountains, which were sufficiently high to exclude entirely all view of the lofty snowy mountain seen the day before, were round-topped and covered with snow. Vegetation was entirely wanting on the top of the pass, but the loose shingle with which it was covered was unfavourable to the growth of plants, otherwise, no doubt, lichens at least would have been seen. Large ravens were circling about overhead, apparently quite unaffected by the rarity of the atmosphere, as they seemed to fly with just as much ease as at the level of the sea.

The great extent of the modern alluvial deposit concealed in a great measure the ancient rocks. At my encampment a ridge of very hard limestone, dipping at a high angle, skirted the stream. Further up the valley a hard slate occurred, and in another place a dark blue slate, containing much iron pyrites, and crumbling rapidly when exposed to the atmosphere. Fragments of this rock were scattered over the plain in all states of decay. On the crest of the pass the rock in situ was limestone, showing obscure traces of fossils, but too indistinct to be determined; the shingle, which was scattered over the ridge, was chiefly a brittle black clay-slate.

On my return no plants were met with till I had almost reached the bank of the stream. The first species which occurred was a small purple-flowered Crucifera (Parrya exscapa of Meyer). Throughout the day the number of flowering plants observed was seventeen, of which three were grasses, three Saussureæ, and two Cruciferæ; there was also one species of each of the following genera, Aster, Nepeta, Gymnandra, Sedum, Lychnis, Potentilla, and Phaca; the dense-tufted Alsine, and a shrubby Artemisia with yellow flowers, complete the number. The only animals seen, besides ravens, were a bird about the size of a sparrow, a bright metallic-coloured carrion-fly, and a small dusky butterfly. Returning by the same road, I arrived at my tent a little after sunset, the distance from the top of the pass being about ten miles.

While travelling at these great elevations the weather was uniformly serene and beautiful. There was but little wind, and the sky was bright and cloudless. At night the cold was severe, and the edges of the streams were in the morning always frozen. On my return towards Sassar I found that the bright sunny weather which had continued since the 16th, had made a great alteration in the state of the stream in the wide gravelly valley along which the road ran. It was now impetuous and muddy, increasing considerably towards the afternoon, when it ran in several channels, which were not always easily fordable. In some places the gravel was throughout the whole width of the plain saturated with water, and gave way under the feet, so that it became necessary to ascend on the stony sloping banks on one side or other, instead of following the centre of the valley. At Murgai, on the evening of the 23rd of August, just after sunset, I felt three slight shocks of an earthquake. On that day the weather again became dull, and on the morning of the 24th there was a slight fall of snow for about an hour.

The remarkable open plain to the south of the Karakoram pass occupies a deep concavity in the great chain of the Kouenlun, which there appears to form a curve, the convexity of which looks northward. The main range to the eastward was distinctly visible, forming a range of snowless, but certainly very lofty, black peaks beyond the sources of the most eastern branch of the Shayuk; while the heavily-snowed mountains, the summits of which were seen further east, were probably also a part of the axis of the chain, which apparently bends round the sources of the river of Khoten, or of some stream draining the northern flanks of the Kouenlun. To the westward, no peaks rose behind the snowy ridge which terminated the western branch of the Shayuk a little west of the Karakoram pass, beyond which the surface probably dips, while the axis of the Kouenlun bends to the southward, towards the glaciers of the Nubra river.

In crossing the open plain on my return towards Sassar, I had the splendid snowy peaks to the south-west always in view, and was able to form a tolerable estimate of their appearance and elevation. The range was very heavily snowed, and from the lateness of the season but little additional thaw could be expected. What seemed the highest peak was very near, and its position could be determined by bearings with little risk of error. It rose abruptly in the midst of a great mass of snow, which filled the hollows and slopes of the range all around. The surface of the plain over which I was travelling sloped very gently up to the westward, and partly concealed the lower edge of the perpetual snow on the mountains behind, the limit of which was, I think, between 17,500 and 18,000 feet. To the northward and eastward the snow-line was certainly much higher. Here and there, where there was shade, there were patches below 18,000 feet, but even up to 20,000 feet there was no continuous snow. As the source of the snow-fall on these mountains is no doubt the Indian Ocean to the south-west, the gradual rise of the snow-level in advancing north-east, and the occurrence of the highest peaks, and of the greatest mass of snow on branches of the chain, and not on its main axis, are quite in accordance with what is usually the case throughout every part of the Himalaya.

The occurrence of a nearly level plain, six or eight miles in diameter, with a mean elevation of not less than 17,300 feet, is certainly very remarkable. The ridge or watershed of the plain appeared to me parallel to the deep ravine, excavated by the stream along which I had travelled on the 17th of August, and at no great distance from it, as the descent was abrupt. All the northern and western part of this level tract was composed of loosely cohering matters, and was possibly of lacustrine origin; but a much more accurate acquaintance with the outline, structure, and elevation of the plain will be necessary before any certain conclusion can be drawn as to its age or origin.

Before leaving Sassar, I visited the glaciers which descend into the valley of the Shayuk, a little to the north of that place. The path at first lay along the high platform on which I was encamped, which was precipitous towards the Shayuk; it afterwards descended to the level of the river, close to which I travelled for some distance over enormous boulders. The bluff ends of two glaciers were seen high above at the top of the precipitous alluvial bank, and after a walk of upwards of three miles, I arrived at a most superb glacier, which, descending a broad and deep valley in the mountains, and latterly in the alluvial platform, entered the bed of the Shayuk at the bottom of a deep bend, and fairly crossed the river, which flowed out below the ice. On the opposite side of the river, the mountains were precipitous a few hundred feet from the water's edge, but the stream of ice did not extend to the foot of the precipice, but stopped a very few feet up the opposite bank. I could of course only see the position of the ice at the edge of the glacier: how far it extended in the centre I could not tell.

The glacier was extremely rugged, being covered with huge sharp pinnacles of ice, and I was obliged to ascend a long way parallel to its side before I could find a place where it could be crossed. Near its lower extremity it rose high above the surface of the plain, and sloped rapidly down to the river: its sides were there scarped and inaccessible, but higher up it lay in a deep hollow in the alluvial conglomerate. A moment's reflection showed how impossible it was for clay and boulders to resist the friction of such an enormous mass; still I was much pleased to observe the glacier buried, as it were, in a groove of its own forming, from the light which was thereby thrown on the origin of the many broad, shallow, flat-bottomed valleys which occasionally occur in the modern alluvial and lacustrine formations in all parts of Tibet, as for instance at Karsar in Nubra, and at Bazgo below Le. An ancient moraine, deposited at a period when the glacier must have been much more bulky than it now is, skirted the edge of the high bank of alluvium, and prevented the ice from being seen till close at hand, and then only by mounting on the top of the pile of boulders. Down this moraine, which on the face towards the glacier was extremely steep and perhaps sixty feet high, I descended to the surface of the present moraine. The descent required great caution, many of the blocks being loose and easily displaced. When I had reached the surface of the glacier, the passage was not difficult. About a quarter of its width on each side was occupied by blocks of stone; the centre was almost entirely ice, extremely irregular, and here and there a little fissured. The pathway, which was only marked by the footsteps of two men whom I had sent the day before to select a place for crossing, at one time ascended to the top of a ridge of ice, at another descended into a deep hollow. At the time I crossed (about eleven A.M.) numerous streams of water had begun to flow in furrows on the surface of the ice. The whole width was close upon half a mile, and on the north side I ascended a steep moraine similar to that which I had previously descended.

From the top of the bank on which the moraine rested, a second glacier came in sight at the distance of a mile. My exploring party reported that they had been unable to find a point at which this glacier could be crossed, and as from the appearance of the mountains behind I felt certain that after crossing it I should only arrive at a third, I did not long persevere in trying to find a passage, but descended to its extremity in order to see whether or not I could walk round it, as it did not appear to enter the water. At the bottom of the valley it spread out in a fan-shaped manner to the width of at least a mile; perhaps indeed much more, for as I failed in getting round it, I was unable to ascertain precisely. At its south-east corner, where it was nearly a hundred yards from the river, a considerable stream, white with suspended mud, was rushing out from beneath an arched vault of ice, even before sunrise. To avoid fording this icy stream, the margins of which were thickly frozen, I crossed with a good deal of difficulty an angle of the end of the glacier. On its surface I found several small moraines, which had sunk down into grooves ten or fifteen feet deep, and had therefore been invisible from outside. Further progress on the ice was stopped by cliffs which were not accessible without ladders, so that I had to descend to the bank of the Shayuk. I walked along between the ice and the river, till my advance was stopped by the glacier fairly projecting into the water in such a manner that I could not see anything of what lay beyond. The icy wall being quite inaccessible, I could not get upon the surface of the glacier to attempt to advance in that way, nor could I ford the river, which was very deep.

The terminal cliff of the glacier varied in height from fifteen to thirty feet, and a talus of large stones lay in front, evidently deposited by it. Indeed, while I was there I saw several small stones which projected from the face of the cliff, drop out by the melting of the ice in which they were imbedded. Many cavities were seen in the ice, from which large stones must have dropped out no longer ago than the day before, and the stones which corresponded in size to them were seen lying close at hand. Before I left the front of the glacier, the heat of the sun having become considerable, rapid thaw had commenced; rills of water trickled down its face in every direction, and the sound of falling stones was to be heard on all sides. Now and then a report as loud as that of a cannon was heard, caused, as I supposed, by the fall of a very large boulder from one of the smaller glaciers, which stopped abruptly at the top of the high cliff of alluvium.

Before quitting finally these magnificent glaciers, I ascended to a height on the mountain-side in order to see whether or not there was any lake in sight corresponding to that laid down, from information, by Mr. Vigne as Nubra or Khundan Chu. The mountains were very steep and stony, and were covered above 16,000 feet with snow, which had fallen in a storm a few days before; I did not, therefore, get up to any great elevation, probably not beyond 16,500 feet, but at that height I could see nothing of the river beyond the second glacier, though its course through the mountains could be traced distinctly enough. It is, however, highly improbable that any permanent lake exists. Such could, I think, only be formed by the stoppage of the river by a glacier, an obstruction which could only be temporary, and would inevitably be followed by a terrific inundation, such as is known repeatedly to have devastated the valley of the Shayuk.

It had been my original intention, on my return from Karakoram, to follow the course of the Shayuk all the way from Sassar to Nubra, but on my return to the former place after visiting the pass, I found that there was no probability of the road along the river being practicable for at least three weeks, the depth of the stream, which requires frequently to be forded, being still much too great; I was therefore reluctantly compelled to return by the same route as that by which I had reached Sassar. Early in September, I found the crops in Nubra ripe, the barley being mostly cut; buckwheat and a few fields of millet, however, were still quite green. The Shayuk had very considerably diminished in size: one branch which in July had been three feet deep was quite dry on the 6th of September. On the 11th of that month I crossed the pass above Le, the state of which was a good deal altered. The little lake, which on the 20th of July was still frozen over, was now free of ice, nor was there any snow, except a very few small patches, below the steep snow-bank on the northern side. The snow, which had covered this steep descent, had melted away, exposing a mass of ice, which was not crossed without a good deal of difficulty and some little risk. Loaded cattle were unable to get to the top of the pass till the afternoon. The snow on the south face had almost entirely gone.

I reached Le just in time to escape some very unsettled weather, during which snow fell on the mountains down to about 13,000 feet. This was ushered in by very high wind, blowing in gusts from all points of the compass. Heavy clouds formed, but always high: on the 14th there was a good deal of thunder, and during the following night a smart shower of rain, which lasted about an hour.

The inhabitants were busy with the operations of harvest. A coarse knife or rude sickle was employed to cut the wheat and barley as close to the ground as possible; they were then tied into large bundles, each sufficient for one load, which were carried (usually by women) to the threshing-floors, not without considerable loss, from the ripeness of the ears and the great bulk of the loads, which were rubbed against every obstacle, particularly the narrow walls of the pathways between the fields. The grain was trodden out of the ear by cattle and asses, all muzzled, on small threshing-floors made of clay beaten hard. It was then winnowed, by being gently shaken out of flat vessels held as high as possible above the ground.

On the 15th of September I left Le for Kashmir. For five days my route was the same as that by which I had travelled in July. On the fourth day I reached Kalatze on the Indus, and on the 19th of September I encamped at the village of Lama-Yuru, close to which the road from Zanskar joins that along which I proposed to travel towards Dras. In the valley of the Indus a great part of the vegetation was already destroyed by the night frosts; Chenopodiaceæ were now the most numerous family, and these were rapidly ripening their seeds. In the narrow ravine of the Wandla river, on the ascent to Lama-Yuru, I found a few plants indicative of lower and hotter regions than those in which I had lately been travelling: a little wiry Lactuca with decurrent leaves, a spathulate-leaved Statice, and a small Hyoscyamus, all plants of the neighbourhood of Iskardo, were those which I noted.

On the 20th of September I crossed the Phatu pass, stated by Moorcroft to be 14,000 feet above the sea, but which Major Cunningham has ascertained to be only about 13,500 feet. The discrepancy is probably owing to some error in Moorcroft's manuscripts, from which the elevations given in his work were calculated by Professor Wilson. In the neighbourhood of Lama-Yuru lacustrine clay occurs in great abundance, and the ascent to the summit of this pass was gentle, up a gravelly valley, which was full of alluvium, almost to the very summit. The pass did not nearly attain the elevation requisite for alpine vegetation, still the flora was a good deal altered; two large-flowered thistles, Caragana versicolor, and several species of Umbelliferæ were observed, none of which had occurred in the hills to the north of the Indus; the prickly Statice was also common, but the Chenopodiaceæ of the Indus valley had entirely disappeared. The descent along the Kanji river to Karbu, at which I encamped, was long and gradual, down a wide valley skirted by gently sloping hills, which, at some distance on the left, rose into high mountains, but on the right attained only a moderate elevation, the Indus being at no great distance. Alluvium occurred throughout the descent, latterly indurated into a coarse conglomerate.

From Karbu I marched on the 21st to Molbil, crossing the Namika pass. The previous night had been very threatening, with violent wind, and at daybreak all the hills around were covered with snow; it was still snowing slightly, but none lay in the valley, and before nine o'clock it cleared, and the remainder of the forenoon was tolerably fine. For two miles I followed the banks of the Kanji river; afterwards the road turned to the left to ascend a clayey valley, to the rounded summit of a ridge separating that river from the Pashkyum on the left. The pass has been determined by Major Cunningham, who crossed it in October, 1847, to be 12,900 feet above the sea. The descent was long, but not rapid after the first mile. The upper part was desert, but lower down villages were frequent and cultivation extensive. At first the rocks were clay-slate, but these were replaced in the lower part by a hard limestone; alluvium was everywhere plentiful, forming, near Molbil, table-topped platforms of indurated conglomerate, horizontally stratified, and faced towards the stream by scarped cliffs. The afternoon was again stormy, and a good deal of rain fell during the night.

Next day I made a long march to Pashkyum, following the course of the river of that name. The descent was very gradual, and the road varied much in character, the valley being sometimes open, at other times narrow and rocky. The villages increased in numbers as the elevation diminished, and latterly for several miles cultivation was continuous. Pashkyum is not more than 8600 feet above the sea, and accordingly the season was much less advanced than it had been three and four thousand feet higher, the weather being much milder, and the summer heat no doubt much more considerable than in the neighbourhood of Le. The crops had long been cut, except the buckwheat, the fields of which were however quite ripe; the plants were being plucked up by the roots and laid down separately in the fields to dry, previous to removal to the threshing-floor.

A remarkable change had taken place in the appearance of the country during this day's journey. The banks of the river were frequently shaded with immense willows, and the trees of the cultivated lands were numerous and of great size. Many new forms of plants were also seen, though the general character of the flora was unaltered. Shrubby Artemisiæ were extremely plentiful, and the Perowskia, Ballota, Echinops, and Iris of the Indus valley were very abundant. The new plants were all species of Kashmir or Iskardo, such as Verbascum Thapsus, Lappa, Valeriana, Swertia, and Gentiana Moorcroftiana. Trifolium repens and fragiferum grew in the pastures close to the river, and tropical species of Setaria and Amaranthus were common weeds in the corn-fields.

In the immediate neighbourhood of Pashkyum the rocks consist of coarse-grained grey or white sandstones, often containing small water-worn pebbles, and alternating with dark crumbling pyritiferous shales. These rocks, which dip to the east or south-east, at an angle of not more than 15°, rise on the north side of the valley to the summit of a long sloping ridge, which appears to overhang the Indus. As these sandstones and shales contained, so far as I could observe, no fossils, their age is a matter of complete uncertainty. They were quite independent of the modern lacustrine formation, patches of which, perfectly horizontally stratified, and therefore unconformable to the other, were seen in several places resting on the sandstone. These sandstones perhaps reach as far as the Indus, but I was not able to determine how far they extended to the southward, in which direction high and rugged mountains, now covered with snow, skirted the valley at a distance of a few miles.

On the 23rd of September, I followed the Pashkyum river to its junction with that of Dras. Crossing, at starting, to the left bank of the river, the road lay for a mile through cultivated lands; it then ascended to a platform of alluvium, which blocked up the valley, while the river disappeared in a narrow ravine far to the right. Five miles from Pashkyum, I descended very abruptly from this elevated plain, to the village of Kargil, where the Pashkyum river is joined by a large stream from Suru, called by Moorcroft the Kartse; which I crossed by a good wooden bridge, close to a small fort, occupied by a Thannadar with a small party of soldiers. The cultivated lands of Kargil, which is elevated about 8300 feet, are extensive and well wooded; but immediately below, the valley becomes narrow and rocky, and continues so for more than a mile, till the stream joins the Dras river. Nearly due south of Kargil the stratified rocks of the mountains are replaced by igneous rocks, and the point of contact of the two is well marked on the precipitous face of a lofty peak. At first the igneous rock was dark and resembling greenstone, but it soon changed to granite, which, as I had observed in April, occurs everywhere in the valley of Dras, below Karbu.

I encamped on the right bank of the Dras river, about a mile above the village of Hardas. Henceforward my route was the same as I had travelled in April. On the 24th I travelled to Tashgong, and on the 25th I arrived at Dras. In most parts of the valley I found a great deal of alluvium, but I saw none of the fine clay which is characteristic of the purely lacustrine strata above the village of Bilergu, where I had observed it in April. Gravelly conglomerate was everywhere the prevailing form,—sometimes indurated, but generally soft and shingly. Most of these deposits were unstratified, but distinct stratification was far from uncommon. The alluvium often capped low hills in the open valley many hundred feet above the bed of the river, and it was observed at frequent intervals in every part of the valley, from the junction of the Pashkyum river to Dras itself.

The great extent and remarkable forms of alluvium which I had seen in the district through which I had travelled, between Kalatze and Dras, induced me to note with care the position and composition of the alluvial beds of the Dras valley. The known low elevation of the Zoji pass, between Dras and Kashmir, which is only 11,300 feet above the sea, made the great extent and continuity of these deposits very remarkable, and with difficulty explicable, unless on the supposition of the existence of a series of lakes separated from one another by extensive accumulations of alluvium, now to a great extent removed by denudation. The lacustrine clays of lower Dras, about Ulding, appear continuous with those of the Indus valley about Tarkata, but the clays of Pashkyum, which are separated from them by a very thick mass of alluvium, which occupies that part of the Dras and Pashkyum rivers immediately above the junction of the two, may have been deposited in an isolated lake. Further east again, at Lamayuru, there are beds of pure clay as high as the summit of the Zoji pass, so that the alluvial beds of the upper part of the Phatu ridge must have separated the lake in which these were deposited from the more western waters, which (it may be conjectured) at the same time covered the whole of the valley of Molbil and Pashkyum.

The vegetation of Dras was still very Tibetan, but transitional forms were becoming frequent. The Chenopodiaceæ (except Eurotia) had all disappeared, but Artemisiæ and Umbelliferæ were very abundant. The new forms were all Kashmirian, and indicated a considerable increase of humidity: a small white-flowered balsam was observed not far from Hardas, and Prunella, Thymus Serpyllum, an Achillea, Senecio, Galium, and Silene inflata were all seen below the fort of Dras. At that place the harvest was but just over; indeed, a field or two of wheat were still uncut.

On the 26th of September, I marched to Maten, along a road which, in April, had been entirely covered with deep snow. Part of the road was rocky, but in general the valley was open. During this day's journey, a very great change took place in the vegetation. Hitherto, Kashmirian plants had been the exception, the greater part of the species being Tibetan; to-day the reverse was the case, most of the plants seen being those common in the comparatively moist climate of Kunawar, or species new to me, but belonging to families or genera which inhabit a more humid climate than Tibet. Groves of dwarf willows lined the banks of the stream, and nearly sixty species of plants not observed in Tibet were collected during the day. Vitis, Aconitum, Hypericum, Vernonia, a prickly juniper, Convallaria, and Tulipa, may be selected as illustrative of the greatness of the change, which was particularly interesting from its suddenness. Numerous Tibetan forms no doubt still lingered, but principally such as extend into Kashmir. At Maten the barley was still uncut, notwithstanding that it is upwards of a thousand feet lower than Le, at which place harvest was nearly over at the time of my departure.

There can be no doubt that the sudden alteration in the character of the vegetation is due to the great depression in the chain separating Tibet from Kashmir, at the Zoji pass, which is far below the usual level of the lowest parts of these mountains. The access of a great amount of humidity, which would have been condensed if the moisture-bringing winds had been obliged to pass over a lofty chain, makes the autumn partially rainy, and frequently cloudy, thereby diminishing the action of the sun's rays, and lowering the mean temperature of the summer.

On the 27th of September, I crossed the pass of Zoji La, which had now a very different aspect from that which it had presented in April. From Maten the road lay up a wide open valley with a scarcely perceptible ascent, generally along the edge of a small stream, but occasionally on the slope of the hill-sides. The valley was flat and often swampy; but the mountains on both sides, more particularly on the left, were high and abrupt, not unfrequently precipitous. On that side there were in most of the ravines large patches of snow, and in one there was a fine glacier, which stopped abruptly within a hundred yards of the main valley. Latterly a few patches of snow lay even in the open valley. The vegetation was almost entirely Kashmirian, not more than six or seven out of about 110 species being otherwise; the hill-sides were covered with brushwood, at first of willow and prickly juniper, but latterly principally of birch.

Five or six miles from Maten, the main branch of the stream was found to descend from a narrow ravine on the left, at the head of which there was perhaps a glacier. In the valley along which the road lay, there was scarcely any water in the bed of the stream, and about a mile further on, without any increase in the inclination, I came to a large patch of dirty snow, beyond which there was a very evident slope to the southward. The boiling-point of water here indicated an elevation of 11,300 feet. A few hundred yards further, I arrived at a large pond (it could hardly be called a lake), into which a very small rill of water was trickling from the north, while from the opposite end a stream ran towards the south. This little lake was not, as I had expected, on the crest of the pass, but undoubtedly on the Kashmirian side of it.

Beyond the lake, the descent became steep, and the valley contracted into a rocky ravine, full of snow, under which the little stream disappeared. The road was at first on the left side of the valley, but crossed on the snow at the commencement of the contracted part, and ascended rather abruptly a steep hill on the right through a very pretty grove of birch. The top of this steep ascent is usually considered by travellers as the pass, and is the place to which the name Zoji La properly belongs. The point of separation of the waters must of course, for geographical purposes, be considered as the actual pass, but this ridge, which, if not actually higher, is at all events on a level with it, and has in addition a steep ascent on both sides, has not unnaturally had that honour assigned to it. On reaching the shoulder of the ridge, the valley of Baltal came in sight, presenting, in the words of Moorcroft, "as if by magic, a striking contrast in its brown mountains and dark forests of tall pines to the bare rocks and few stunted willows to which we had so long been accustomed." The sight of a forest is certainly a great source of gratification to a traveller who has been long in Tibet; but the pleasing effect of the view from the Zoji pass is not owing merely to contrast; as the traveller looks down upon the bed of Sind river, more than 2000 feet below, and the forest in the valley is not too dense, but interspersed with open glades, while beyond rise high mountains tipped with snow. I do not think that I have anywhere in the Himalaya seen a more beautiful scene than that which then lay before me; but the effect was enhanced by the recollection of the appearance of the same spot in April, when the whole landscape was covered with snow, and I descended from the summit of the pass on a snow-bank which filled up the now inaccessible ravine, on account of which I was obliged to make a long detour. The descent was extremely abrupt, through a pretty wood, down to a log hut built for the accommodation of travellers a few hundred yards from the river, at an elevation of 9,200 feet.

The flora of the Sind valley at Baltal was very rich: the forest consisted chiefly of pine, poplar (P. ciliata), birch, and sycamore, intermixed with underwood of Ribes, Berberis, Viburnum, Lonicera, and Salix. The herbaceous vegetation had all that excessive luxuriance which characterizes the subalpine forests of the Himalaya at the end of the rainy season. Gigantic Compositæ, Labiatæ, Ranunculaceæ, and Umbelliferæ were the prevailing forms. There were several large patches of snow in the bed of the lateral torrent which descended from Zoji La, as low down as the log hut; and it was not a little curious to observe, that in spots from which the snow had only recently melted, the willows were just beginning to expand their buds, and the cherry, rhubarb, Thalictrum, Anemone, Fragaria, and other plants of early spring, were in full flower.

In descending the Sind valley towards Kashmir, my route was the same by which I had travelled in April. The mountains on the left were extremely precipitous and heavily snowed, and in a ravine a little below Sonamarg a glacier descended almost to 9000 feet. The lower part of the valley was one sheet of cultivation, chiefly of rice, which was almost ripe. In the neighbourhood of Kashmir, where I arrived on the 5th of October, the season of vegetation was almost at an end; species of Nepeta, Eryngium, Daucus, Centaurea, Carpesium, and several Artemisiæ being the most remarkable of the herbaceous plants remaining. In the lake there were vast groves of Nelumbium leaves, but the flowers and fruit were both past; Salvinia was everywhere floating in great abundance; while the other aquatic plants were species of Bidens, Stachys, Mentha, Scutellaria, Hippuris, and Typha, all European or closely resembling European forms.

Besides rice, which constitutes the staple crop of the valley, the principal grains cultivated in autumn appeared to be different kinds of millet, and a good deal of maize; Indian species of Phaseolus also were common, now nearly ripe. The wheat and barley, which are much earlier, were already above ground. I saw a few fields of Sesamum (the Til of India), and in drier spots a good deal of cotton, which was being picked by hand, but appeared a poor stunted crop, much neglected.

On the high platforms between Pampur and Avantipura the saffron was in flower, and its young leaves were just shooting up. This crop seems a very remunerative one to the Raja, who retains the monopoly in his own hands, compelling the cultivators to sell the produce to him at a fixed price. The bulbs are allowed to remain in the ground throughout the year, and continue in vigour for eight or ten years, after which the produce diminishes so much in quantity that the beds are broken up, and the bulbs separated and replanted. The flowers are picked towards the end of October, and carried into the town of Kashmir, where the stigmas are extracted.

Another very important product of Kashmir is hemp, which grows spontaneously along the banks of the river, forming dense thickets often twelve and fifteen feet in height, and almost impenetrable. It is only used in the manufacture of an intoxicating drink, and for smoking; and the plant is preserved entire, in store-houses, in the town of Kashmir, till required for consumption.

From Kashmir I proceeded towards the plains of the Punjab by the same route by which I had travelled in May. During my absence in Tibet, the second Sikh war had broken out, and as it was then at its height, it was not easy to reach the British territories. I was therefore detained a good while, first in Kashmir, and afterwards at Jamu, and did not reach Lahore till the 16th of December.

CHAPTER XV.

General description of Tibet—Systems of mountains—Trans-Sutlej Himalaya—Cis-Sutlej Himalaya—Kouenlun—Four Passes across Kouenlun—Boundaries of Western Tibet—Height of its mountain ranges and passes—Climate of Tibet—Clouds—Winds—Snow-fall—Glaciers—Their former greater extension—Elevation to which they descend—Snow-level—Geology—Lacustrine clay and alluvium.

The elevated country of Central Asia, situated to the north of the lofty snowy mountains which encircle India from Kashmir to Assam, is familiarly known to Europeans by the name of Thibet or Tubet,—most properly, I believe, Tibet. This name is also commonly employed by the Mohammedan nations to the north and west to designate the same country, but is not, so far as I am aware, known in the language of the Tibetans themselves, among whom different portions of the country are usually known by different names.

The whole of Tibet (as far as our present very limited knowledge of the south-east portion enables an opinion to be formed) appears to be characterized by great uniformity of climate and productions, and perhaps also of natural features, on which account it appears convenient to retain the name for the whole country, although, as has already been pointed out by Baron Humboldt[28], it is naturally separable into two grand divisions. One of these, the waters of which collect to join the Sanpu, which in India becomes the Brahmaputra, is still scarcely known; the other, drained principally by the Indus and its tributaries, has been repeatedly visited by European travellers. The line of separation between these two portions lies a little to the east of the great lakes[29], from the neighbourhood of which the country must gradually slope in both directions towards the sea.

If the whole of western Tibet formed (as it does, according to the popular opinion on the subject of the countries to the north of the Himalaya) an extensive plain bounded on the south by the great chain of the Himalaya, and on the north by the lofty mountains of Kouenlun, it would be an easy task to define its limits. This is, however, so far from being the case, that the greater part of the surface of the country is traversed in all directions by ranges of mountains in every respect similar to the Himalaya, of which in fact those south of the Indus are ramifications, while those on the north are branches of the snowy chain of Kouenlun.

If, again, the Himalaya formed an uninterrupted chain along the southern border of Tibet, broken only by the passage of the Indus at one extremity and by that of the Brahmaputra at the other, the mountainous nature of the interior would be no obstacle to the existence of a clear and distinct boundary. Unfortunately, however, for simplicity of definition, no such chain exists. A line of high snowy peaks may doubtless be traced in a direction nearly parallel to the plains of India, but these are separated from one another by deep ravines, along which flow large and rapid rivers, and therefore afford no tangible line of demarcation between the two countries.

Between the river Indus and the plains of north-west India is interposed a mountain tract which has a breadth of about 150 miles in linear distance. This tract is everywhere (with one exception) extremely rugged and mountainous, nor is it at all an easy task to convey an idea of the extreme complication of the ramifications of the numerous ranges of which it consists. No wide plain (Kashmir alone excepted) is interposed between these ranges, so that the only feasible mode of division which appears to be applicable to them is afforded by the course of the different rivers which traverse them in various directions. If these be taken as a guide, the mountains will be found to resolve themselves into two great systems connected to the eastward, but otherwise independent of, though nearly parallel to, one another.

From the sources of the west branch of the Chenab or Chandrabhaga river, a range of very great elevation runs in a north-west direction as far as Kashmir, and, after reaching the north-east corner of that valley, assumes a more westerly direction so as to encircle the whole of its north side, bending at the same time gradually towards the south. This chain forms the line of separation between the waters of the Indus and those of the Chenab and Jelam. To the eastward of the Baralacha Pass it ramifies to a considerable extent, its different branches including between them several depressions quite unconnected with the general drainage of the country, and surrounded on all sides by ranges of hills which prevent any exit of their waters. The principal of these depressions is that of lake Chumoreri; another is occupied by the little salt lake first visited by Trebeck, and called by him Thogji[30].