a Sanscrit word, compounded of "hima," cold or snow; and "alaya," place (of Wilson's Sanscrit Dictionary), is the name given to the ranges of mountains which bound India on the N., from the bend of the Indus on the W. to that of the Brahmaputra on the E. On the S. they are bounded by the plains of India, and on the N. by the Tibetan courses of the above-named rivers. A transverse section of the Himalaya nowhere presents the appearance of a simple range, but of several more or less parallel chains, separated by valleys of very great depth and steepness; this is because the secondary ranges that ramify N. and S. from it are of great length, breadth, and complexity, and from bending to the E. or to the W., often run for many miles parallel to one another and to the main range, besides rising into eminences loftier than any on the latter, for which they are sometimes mistaken. The axis of the Himalaya is, moreover, not marked out by any continuous ridge or succession of peaks, but is often broad, open, and low, compared with the neighbouring isolated eminences. Hence the line of watershed becomes the only geographically determinable axis; and this, as in all mountain chains of any extent, follows an extremely sinuous course. No doubt this line, which throws the waters in two opposite directions throughout the whole extent of the range (1440 miles), is also that of greatest elevation, or that along which the land is uninterruptedly the most lofty.
Before, however, the real nature and geographical limits of the Himalaya, as above defined, can be rightly understood, it is necessary to consider this range in its relation to the little known mountain systems of Central Asia, of which it perhaps forms a less important part than is usually supposed. On reference to the map of Asia, the watershed of that continent will be found to follow a tortuous line, running diagonally from the peninsula of Gujarat to Behring's Strait. Across the plains of India this line is for the most part indicated by the Arawali chain, N. of which it crosses the Himalaya obliquely in a N.E. direction to the sources of the Indus and Brahmaputra, whence it trends westerly to the source of the Oxus, and then again north-easterly along the Altai to the S. of Lake Baikal, till it becomes the Iablonoi Mountains, and finally terminates in the prolongation of that range which traverses Himalaya the country of the Tehukitchi.
All the great rivers of Asia rise in this watershed; those from its western slope flow N. into the Polar Sea, W. into the Caspian or Aral, and S.W. into the Arabian Sea; those from its eastern slope flow E. and S.E. into the Pacific, and S. into the Indian Ocean. Enormous mountain chains branch off to the E. and W. of this main axis, inclosing the valleys of the rivers; and of these chains the southernmost is the Himalaya.
In their Tibetan courses the Indus and Brahmaputra occupy valleys of great elevation, and the opposite directions taken by them indicates the division of the Himalaya into two portions, the eastern of which stretches from their sources at the Peak of Kailas to the bend of the Brahmaputra, and the western terminates at the bend of the Indus. These limits are more natural than is usually supposed, since the prevalent idea that the Brahmaputra enters Assam through a defile caused by a break in the chain is erroneous; on the contrary, the Himalaya gradually declines in elevation in East Bhutan; and the upper valley of the Dihong (as the Brahmaputra at its bend is called), is, according to the best information hitherto procured, broad, open, and hot—rice being cultivated there on the very confines of Eastern Tibet. So also the Indus at the western extremity of the chain is usually described as flowing through a defile; but though its valley to the W. of Kashmir is contracted and rugged, and overhung by stupendous mountains, it does not in this respect differ materially, if at all, from the remainder of its Tibetan course; nor is the fall of its bed between Iskardo and the plains of the Punjab greater in proportion to the length of its course than it is above that town.
The branches or secondary ranges of the Western Himalaya are so long and lofty, that some difference of opinion exists as to which of them should be most appropriately considered as the continuation of the chain between the peak of Kailas and Kashmir; and we have considered the line of watershed between the tributaries of the Indus to the N., and the rivers that flow to the plains of India to the S., to be the axis of the chain, since it both indicates the line of mean greatest elevation, and is the only definable axis in a geographical point of view.
Of the secondary chains we shall speak at length in connection with the rivers they inclose. Their direction is often perpendicular to the main chain, but they are so often oblique, and even parallel to the main chain, especially at their upper parts, that where very lofty and heavily snowed, they are frequently taken by local observers for the axis of the Himalaya itself; an error to which may be traced that misconception regarding the relative amount and duration of the snow on the northern and southern slopes of the Himalaya, which has led to so much fruitless controversy in India and Europe.
The general direction of the Himalaya throughout its length of 1440 miles, is E. and W., but it trends northwards from the centre towards its western extremity, its extremes being respectively in N. Lat. 28°, E. Long. 95°, and N. Lat. 35°, E. Long. 73°. Its breadth varies in different parts, but has been accurately ascertained in the western portion only, where it deviates but little from 190 miles.
It has been stated that the mountain ranges of the Himalaya and the Kuenlun have no special existence as chains apart from the general elevated mass of Tibet, and that that rugged country forms the summit of a great Himalaya protuberance above the level of the earth's surface, of which the two chains form the N. and S. faces. This view is derived, no doubt, from the appearance of extreme confusion that prevails in this; as in every mountainous country. In the Himalaya, as in Switzerland, the traveller, in crossing ranges and valleys at all angles, perceives no order, and finds it impossible to trace rivers or chains from elevated positions; whilst in following the courses of the valleys, their bounding mountains shut out all beyond; and it is only after a map has been constructed that the relations between the several parts of a mountainous country can be traced with accuracy. The mountain system of Central Asia differs in a physico-geographical point of view from that of Europe only in bulk. The relation of the Kuenlun to the Himalaya is similar to that of the Bernese to the Monte Rosa Alps. These have not, it is true, a separate existence apart from the general mass of the Alps, but that they have a special existence also is proved by the positions of the sources of the rivers that flow to the northward and southward from each. Taking as an example the western half of the Himalaya and the parallel range of the Kuenlun, the facts that the average elevation of the watersheds of both is continuously above 18,000 feet the whole way from the Kailas to the meridian of Kashmir, whilst innumerable peaks on each rise above 20,000 feet, and that the inclosed river-bed of the Indus falls from 15,000 to 7000 feet in the same distance, appear sufficient evidence that the ranges in question have a definable existence in a geographical point of view. The elevation of the Kuenlun and Himalaya above the bed of the Indus is continuously higher in proportion than that of the Bernese and Monte Rossa Alps above the bed of the intervening Rhone.
Before detailing the physical features of the Himalaya, it appears advisable to give some general idea of its scenery and aspect; this is derived from the impressions produced on experienced travellers who have described it. Of these there have lately been many, and although few of them have had that previous familiarity with mountain regions which would enable them to judge of the Himalaya by comparison, the narratives of Moorcroft, Thomson, the Strachey's, Cunninghams, and others, all abound in accurate and often graphic details.
Viewed from a distance, on the plains of India, the Himalaya presents the appearance—common to all mountainous countries—of consecutive parallel ridges, running E. and W.; backed by a beautiful crest of snowy peaks, with occasional breaks in the foremost ranges, through which the rivers debouch. This appearance of parallel ranges is owing to a very simple and often overlooked law of perspective; in consequence of which masses of mountains, of whatever configuration, resolve themselves into ranges perpendicular to the line of sight. Any view of the Himalaya, especially at a sufficient distance for the remote snowy peaks to be seen overtopping the outer ridges, is, throughout a great extent of the range, rare, from the constant deposition of vapours over the forest-clad ranges during the greater part of the year, and the haziness of the dry atmosphere of the plains in the winter months. At the end of the rains, when the S.E. monsoon has ceased to blow with constancy, views are sometimes obtained from a distance of nearly 200 miles. From the plains the highest peaks subtend so small an angle that they appear like white specks very low on the horizon, tipping the black lower and outer wooded ranges, which always rise out of a belt of haze; and from the density, probably, of the lower strata of atmosphere, they are never seen to rest on the visible horizon. The remarkable lowness on the horizon of the whole stupendous mass is always a disappointing feature to the new comer, who expects to see dazzling peaks towering in the air. Approaching nearer,
The snowy mountains sink behind the wooded ones long before the latter have assumed gigantic proportions; and when they do so, they too often appear a sombre, lurid, grey-green mass of vegetation, or rock, with no brightness or variation of colour.
The mountains once entered, the appearance of parallel ridges is found to be deceptive, and due to the insinuating spurs of long tortuous secondary ranges that run N. and S. from the axis, dividing deep wooded valleys, which flank the beds of large rivers. The snowy peaks now look like a long E. and W. range of mountains, at an average distance of 30 or 40 miles. Advancing farther into the country, this appearance proves equally deceptive, and from the same cause. The snowy range is finally resolved into isolated peaks, or masses, situated on the secondary ranges; and the source of the deception is found to be that these snow-clad spurs, projecting E. and W., cross one another, and, being uniformly white, appear to connect the peaks into one grand unbroken range. The rivers, instead of having their origin in the snowy mountains, rise far beyond them; many of their sources are upwards of 100 miles in a straight line from the base of the mountains, in a very curious country, loftier by far in mean elevation than the secondary ridges which run S. from it, yet comparatively bare of snow. This rearward part of the mountain region lies in Tibet, and it is here that most of the rivers rise as small streams, which increase rapidly in size as they receive the drainage from the snowed part of the secondary ranges that bound them in their courses.
A belt of tropical forest, 10 to 20 miles in breadth, skirts Terai, the southern foot of the Himalaya throughout the greater part of its length. Its presence is due to the humidity of the climate, and to copious springs, which often give rise to marshes which cover the ground at the base of the mountains. This tract is called the Terai, and is notorious for its malarious atmosphere, which renders it almost certain death to spend any time in it during the spring and autumn (before and after the rains). The Terai belt is broadest, most luxuriant and humid in the Assam valley at the eastern extremity of the chain, narrowing with the diminished humidity and colder winters of the western part. It decreases in breadth, and partially loses its tropical character towards the central districts; whilst to the westward of the Sutlej it disappears, or is represented by a low jungle of bushes, containing but few tropical plants.
Beyond the Terai the mountains rise more or less suddenly, though seldom in precipices. Throughout the eastern parts of the chain they are luxuriantly wooded, while to the westward they are covered with a looser, drier forest, or with brushwood. The mountain region may be entered by following the course of one of the main rivers, or by ascending the outlying spurs which bound them, and which run more or less parallel to the general direction of the chain. The roads almost invariably ascend these spurs, because the malarious region extends far up the valleys, and the banks of the rivers are usually impracticable for paths. The Himalaya once entered, the traveller's route is thenceforth an uninterrupted series of ascents and descents. There are no level tracts, plains, nor flats by the streams, of any breadth or continuity; an endless succession of ridges, several thousand feet high, and as many streams, are crossed on almost every day's march towards the axis of the chain; and during the warmer part of the year the amount of forest, fog, and cloud is so great, that until the alpine regions are reached, the traveller seldom enjoys any of those magnificent panoramic views for which the Cordillera and the European Alps are so celebrated.
Throughout the temperate and wooded regions it is scarcely possible for the traveller, let his powers of obser-
---
1 Thomson's Tibet, p. 2. vation be ever so good, to understand the relations of the innumerable rivers and ridges he traverses. The country resembles a troubled ocean; and there being no apparent order, it is only by taking the main river that flows from the watershed as the starting-point, and laying down its course on paper with some accuracy, that a correct idea can be obtained of the structure of the valley which the traveller is ascending, and its relation to the secondary Himalayan chains that bound it.
The roads from India to Tibet are always carried along the flanks of these broad valleys, for the ridges of the secondary chains are too lofty, rugged, and tortuous to admit of roads being constructed along their crests, while the river banks are hot, and excessively tortuous and rocky. It is the necessity for crossing the spurs from these secondary chains, with their innumerable subdivisions and contained streams, which doubles the length of the route from the plains of India to the axis of the chain; the average distance, which is only about 100 miles, usually occupying from twelve days to a fortnight to traverse; and the total ascent, which is on the average 16,000 to 18,000 feet, being performed many times over, besides involving descents which are so steep as to be hardly less fatiguing than the ascents.
Immediately within the mountains the outermost lateral valleys (containing the feeders to the main rivers) are often broad, and bounded (especially those in the western parts of the chain) on one or both flanks by low sandstone hills. The breadth and extent of these, together with the peculiarity of the rock, has given them an undue importance in some respects. Such broad open valleys are called dhuns, and the sandstone hills (sometimes called the Sewaliks) have been supposed to constitute a system distinct from the Himalaya, but skirting its base. This theory has, however, been rejected by Dr Thomson, who shows that the dhuns are valleys of precisely the same nature as the other lateral valleys; and that the sandstone ranges, however different in a geological point of view, are, in a geographical one, the terminal spurs of the ranges bounding the river valleys. Where the dhuns are very open, flat-floored, and with gradually sloping beds, their true relation to the surrounding mountain-chains is not at once apparent. Sometimes they appear to be indefinitely extended E. and W., in a direction parallel to the Himalayan chain; and, running from one great river to another, they appear to belong to a different order of valleys from those which occur further within the mountains. This arises in some cases from the slope of their beds being so extremely gradual that the watershed between the valley that ascends from the one river, and the corresponding valley that descends to the other river, can only be detected by observation of the drainage; whence the two valleys appear to form one. Such is the case with the celebrated Dehra Dhun in Kumaon, which appears to form one continuous transverse valley between the Jumna and the Ganges, but which really consists of two valleys; one descending from the village of Dehra (which occupies the col) westerly to the Jumna, and the other descending from the same spot easterly to the Ganges. Other dhuns, again, are simply very broad, open valleys, differing in no physical features from those that occur in other parts of the mountains. In the Punjab-Himalaya, where the tertiary sandstones acquire a great development, two or three such valleys occur in succession before the higher mountains begin. These dhuns are not, as is very generally supposed, continuous along the whole extent of the Himalaya, and interposed between the tertiary and secondary mountains. They are merely the outer series of lateral valleys, and are always of limited extent.
It is in the alpine and upper temperate regions of the Himalaya that the most interesting phenomena of the range are concentrated, where the climate resembles that of Great Britain and the alpine districts of Europe; and the scenery, if not picturesque, exceeds in grandeur that of Switzerland and the Tyrol. Throughout the Indian watershed of the chain, the main features, between 8000 and 14,000 feet, are more or less conspicuously European; whether or not they equal those of the Alps may be doubted. In absolute height and mass the Alps of course cannot be compared with the Himalaya; and such a view as the following extract describes (whose main features are characteristic of all parts of the range) is doubtless unrivalled on the globe. The author was at the time in Sikkim, not far from the centre of the chain, and his description embraces the snowed mountains, including the loftiest in the world, as seen from the N. flanks of the outer ranges. "The actual extent of the snowy range seen from Darjiling is comprised within an arc of 80° (from north 80° W. to north 50° E.), or nearly a quarter of the horizon, along which the perpetual snow forms an unbroken girdle or crest of frosted silver; and in winter, when the mountains are snowed down to 8000 feet, this white ridge stretches uninterruptedly for more than 160°. No known view is to be compared with this in extent, when the proximity and height of the mountains are considered; for within the 80° above-mentioned more than twelve peaks rise above 20,000 feet, and there are none below 15,000 feet, while Kinchin is 28,178, and seven others above 22,000. Kinchin-junga (45 miles distant) is the prominent object, rising 21,000 feet above the level of the observer (28,178 above the sea) out of a wilderness of intervening wooded ranges; whilst, on a line with its snows, the eye descends below the horizon to a narrow gulf 7000 feet deep in the mountains, where the great Rungeet, white with foam, threads a tropical forest with a silver line." From another point of view in the same country, and also at 7000 feet elevation, the eye surveys at one glance the vegetation of the tropics and the poles. "Deep in the valleys the river beds are but 3000 feet above the sea, and are choked with fig-trees, plantains, and palms; to these succeed laurels and magnolias; and higher up still, oaks, chestnuts, birches, &c. Pine forests succeed for 2000 feet higher, when they give place to a skirting of rhododendron and berberry. Among these appear black naked rocks, rising up in cliffs, between which are gulleys, down which the snow in January descends to 12,000 feet, ascending in one unbroken sweep to peaks of 18,000 to 25,000 feet. (Himalayan Journals, vol. I., p. 123 and 327).
The Alps nowhere present panoramas so remarkable as these; but when once amongst the snowy mountains, the traveller's position, in respect of proximity to the snowy masses, as well as in elevation, is analogous in both ranges; the absolute heights of the principal objects being nearly the same, and strictly comparable. The apparent elevation of a mountain range is merely relative; and throughout the enormous arc of horizon embraced in a Himalayan view, its apparent vertical height is much diminished by the great distance of the nearest objects. To view Kinchin-junga as Mont Blanc is viewed from Chamonix, the traveller must ascend to 12,000 feet, for it is at that elevation that the vegetation and physical features of the valleys, caused by the moraines, &c., are analogous to those of 3000 feet in the Alps; and from such positions Kinchin-junga being but 3000 or 4000 feet higher above the observer than Mont Blanc (a difference not appreciable by the eye amid such scenes), and being further from the spectator, is not a more strikingly grand mountain than Mont Blanc. In the long extent of the Himalaya there are alpine scenes of unrivalled grandeur; but owing to the rarefaction of the atmosphere, and other causes, these regions will always remain inaccessible to any but the most hardy seekers of the
---
1 Tibet, p. 314. Introductory Essay to the Flora Indica, p. 169. Himalaya picturesque, for they can only be viewed under circumstances of extreme physical discomfort. In certain respects, again, the Himalayan valleys are greatly inferior to the Alpine; for they want both lakes and cascades; and though Himalayan travellers may find scenes more awful, and solitudes immeasurably more impressive than any the Alps present, there are none known which, in grandeur, beauty, and picturesqueness combined, are to be compared with Lauterbrunnen or the valley of Chamonix. Nowhere in the Himalaya do blue glaciers, descending from mountains towering 10,000 feet above, pour their icy streams on to the flat floors of green valleys covered with corn, flocks, and villages; whilst lakes reflecting both the forest-clad base and snow-clad summits of one and the same mountain are, we believe, wholly unknown in Northern India.
Immediately beyond the most heavily snowed ranges of the chain the still ascending traveller enters on the loftiest, coldest, and windiest desert to be found in the temperate zone of either hemisphere, a country contrasting quite as much with the alpine country he has just left, as with the tropical regions at the base of the mountains. This is the axis of the Himalaya, where all its rivers have their rise, and which owes its freedom from snow in part to its distance from the sources of humidity and the mass of intervening ranges, but in part also to its great elevation; for it is in the lower part of the atmospheric column that most vapour is suspended; and as the humid wind only blows from the S., the bulk of its moisture is deposited in the form of rain on the southernmost parts of the mountain range, and of snow on the secondary ranges, 15,000 to 20,000 feet high, which extend many miles S. from the axis. Hence the clouds get absolutely dispersed before they can reach the latter point, and the traveller who has crossed that part of the range where enormous snow-beds and glaciers descend to 15,000 or 16,000 feet, and thence ascends to 18,000 or 19,000, finds no snow at that level; whilst the surrounding mountains also are so bare of snow that it is difficult to conceive that he has not descended an opposite slope.
The features of the main axis of the Himalaya differ greatly at different points of the chain; it is usually broad, and is always characterized by extreme vicissitudes of temperature. As with the Cordillera, the Norwegian Alps, and many other mountain ranges, the line of the watershed is not marked by any continuous ridge or succession of ridges; loftier eminences oftener rise in its proximity, from the spurs that branch off from it than from itself; and as the southernmost of these are (as has been already shown) always very heavily snowed, it has been usual to consider the belt of the most snowed peaks (which, as seen from the southward, shuts out the view of the loftier rearward axis), as the crest of the Himalaya. Dr Thomson has, however, pointed out that the deceptive effects of false perspective, the rarity of snow, the absence of a defined ridge on the axis, and the occasionally greater elevation of isolated peaks on the subsidiary ranges, have led to the transference of the true axis from the watershed to an imaginary line cutting across the valleys of all the rivers. As even the most recent maps of India represent the Himalayan rivers as cutting through the axis of the chain, it is necessary to bear in mind upon what slender authority this is done, and that throughout two-thirds of the extent of the Himalaya, there is no foundation whatever for the position either of the rivers or of the axis, as laid down on our maps. From Kumaon the whole way to the eastern termination of the chain, with the exception of the small province of Sikkim, nothing whatever is even approximately known of the relation of the rivers to the snowy mountains, or of these to the axis. Many lofty peaks are seen from the plains of India, and the positions of a few of these have been determined by triangulation; and in the earlier maps these were fictitiously represented as forming eminences on a continuously snowy ridge. The information was afterwards supplied by the natives that the Himalaya sources of the rivers were far beyond these snows; and to reconcile these phenomena the streams were invariably made to intersect the ridge.
The fact, that in that portion of the Western Himalaya which lies between Nepal and the Sutlej, the loftiest eminences are situated on the subsidiary ranges, was first stated by Col. Herbert; but his observation, that the line of the great peaks intersects the river basins, and therefore is not the true axis of the Himalaya, has never yet been fully appreciated. More recent geographers have, however, so multiplied the number of these peaks in that portion of the Himalaya to which Col. Herbert's observation extended, that no single line of great peaks can now be traced in reality; and with regard to the explored portions of the Himalaya W. of it, and to the Sikkim portion E. of it, it is certain that the great peaks do not follow any definite line; added to which, the number of peaks on the axis itself, attaining elevations of 20,000 to 25,000 feet, is very great.
Reverting to the physical features of the loftiest regions, the suddenness with which the dry Tibetan climate and its concomitant features are often encountered, by crossing a lofty pass over a lateral heavily-snowed spur of a secondary range, is very remarkable, and contrasts with the slowness of the same change when a river is followed to its source; and as the short cuts over these spurs are generally preferred to the winding courses of the rivers, a false impression has been conveyed of the definition of the boundary between the dry and humid regions. This erroneous impression is strengthened by another fact, that the political boundary between the Tibetan and Indian states is often determined by the position of the greatest quantity of snow—a physical obstacle to intercourse of far more importance than the greater elevation of the comparatively snowless watershed N. of it. Recent discoveries have, however, shown that this boundary is neither so straight nor so natural as has been supposed, or as is represented in maps. In the small state of Sikkim, for instance, it zigzags so much that it is 50 miles further S. in one meridian than in another. Again, in the western part of Nepal, it is said to follow the true axis, or watershed; whilst in the more snowy eastern half of the same kingdom, it is traced along the most snowed regions; and it has in many places been repeatedly altered in the course of the last century by the Tibetans and Chinese.
A few other points in the physical geography of the axis of the Himalaya and of its loftier valleys are worthy of note. Such are the prevalence of lakes, never of any great depth, and of tertiary deposits, often extending for many miles, and forming undulating expanses, through which the rivers cut deep gorges, and which also form flats and terraces along the banks of the streams. These have given rise to the supposition that Tibet is a vast plain or plateau—an error which the fact that the roads in that usually difficult country make long detours to take advantage of these flats, has tended to confirm. That Tibet is however the most rugged and mountainous country in the world, the united testimony of all travellers assures us; and that no such level expanses, or lakes (with the solitary exception of Kashmir), or flat-bottomed valleys, occur in the equally mountainous, though perhaps less rugged and precipitous temperate and tropical valleys of the Indian watershed, is a very remarkable fact.
In classifying the geographical features of the Himalaya, Divisions as they occur, in belts parallel to the axis, we recognize the Himalaya in four distinct latitudinal zones. 1. That which extends from the plains of India to those parts of the chain where the elevation is sufficient for snow to lie upon the ground during the winter months; within this the bases of the hills and all the valleys are tropical, and the upper portions tem- The surface, though steep, and cut up into innumerable ravines, is seldom absolutely precipitous, or continuously rugged. This belt is from 40 to 60 miles broad, and presents eminences of all heights below 13,000 to 14,000 feet—the mean being perhaps from 5000 to 8000 feet. 2. In the snowy belt, which is from 30 to 40 miles broad, the climate is temperate in the valleys, and tropical where these are very deep and damp; alpine on the heights. The mean elevation is probably under 13,000 feet, though innumerable peaks rise to 20,000, a few to 25,000, and some to 28,000 feet; while the river beds are often only 3000 feet above the level of the sea. The surface of this belt is very rugged and often precipitous. 3. The axis of the chain is from 10 to 20 miles broad, probably 15,000 feet in mean elevation; and the line of watershed itself seldom sinks below 17,000 feet, except at the two extremities of the chain. The surface is very rocky and often precipitous, but varied with occasional undulating expanses; the climate is alpine throughout. 4. The northern slope consists of rugged rocky valleys, with occasionally flat floors; a temperate climate, but one of excessive vicissitudes, prevails below 13,000 feet elevation; and great drought and a total absence of forest vegetation distinguish it from the analogous elevations of the southern side.
These general remarks indicate three principal series of divisions of the Himalaya, viz., according to length, breadth, and height; these are:
1. Longitudinally; an eastern and western Himalaya. 2. Latitudinally; an exterior or damp, an interior or snowy, and a Tibetan or dry Himalaya. 3. Altitudinally; a tropical, temperate, and alpine Himalaya.
The secondary ranges, which originate in the axis, and descend on the S. slope to the plains of India, and on the N. to the Tibetan Indus and Brahmaputra, separate great rivers, and may consequently be conveniently used to divide the whole chain into a succession of river basins.
The great rivers flowing to the S. are thirteen in number; advancing from W. to E., they are,—the Jelam, Chenab, Ravee, Byas, Sutlej, Jumna, Ganges, Gogra, Gandak, Kosi, Tista, Monas, and Subansiri. The directions of these rivers vary a good deal, being often extremely oblique, especially in the upper part of their courses, to the axis of the chain. As a general rule, those that rise nearest to the centre of the chain (as, for instance, the Tista) have the straightest courses; those that are situated to the eastward flow first S.E., and then turning S., follow a western course, to the Assam Brahmaputra; those in the W. of the chain, flow first S.W., and then turning S., flow easterly. Of these rivers, the two eastern, the Monas and Subansiri, flow through Bhutan; the Tista drains Sikkim; and the three next, the Kosi, Gandak and Gogra, which are Nepalese, all flow into the Ganges; the two following, the Jumna and Ganges, water the British hill states of Kumaon, Garhwal, and Simrore, W. of Nepal; the remaining five water the Punjab, and flow into the Indus, draining amongst others the hill states of Mandi, Chamba, Jamu, and Kashmir, respectively.
The fall of the rivers varies with the length of their course; that of the Tista is perhaps the most rapid, the river descending from 17,000 feet to 300 (at its exit on the plains), with a fall of 85 feet per mile; this, however, varies in different parts of its course; thus, between 17,000 and 15,000 feet, it is 60 feet per mile; between 15,000 and 12,000 feet, 140 feet per mile; between 12,000 and 5000 feet, 160 feet per mile; and between 5000 and 300 feet, 50 feet per mile. Analogous differences have been observed in other Himalayan rivers, as in the Tambur, a tributary of the Himalaya Nepalese Kosi, and in some of the western rivers; as the Sutlej, which, for the first 60 miles of its course, falls 35 feet per mile; in the following 50 miles, 82 feet per mile; and in the succeeding 68 miles, 25 feet per mile. The Indus again, for the first 90 miles, falls about 3 feet per mile; in the next 115 miles, 28 feet per mile, and in the remaining 345 miles, falls 18 feet per mile. The prevalence of this gradual fall at the upper part of the rivers has been attributed mainly to the fact, that in the dry climate of the axis, the floors of the valleys are raised and levelled by glacial and other accumulations, which have been removed by denudation from the lower parts of the same valleys.
All the Himalayan rivers increase enormously in volume during the summer months; in the eastern portions, where the rains are very heavy, the increase is more due to this cause than to the melting of the snow. The individual feeders of the main streams also differ much; those that rise from glaciers being in summer opaline and full, in winter clear and less in volume. Again, the main body of the streams is in summer swollen and muddy, owing to the excessive rains, but in winter clear and diminished to a fraction of its summer volume. In summer the glacier-fed feeders are always much the coldest; in winter all approximate more in temperature. In summer the volume of all the snow-fed feeders of the rivers is much diminished at night, owing to the less rapid melting of the snow, and the frequent freezing of their head waters; but this is scarcely appreciable in the central and eastern parts of the chain, where most rain falls at night, and where the effect of the sun's rays in melting the snow is less than that of the rain and corroding fog.
The rivers flowing northward from the axis have much shorter courses; and, owing to the dryness of the climate, they carry comparatively little water. The principal of those that flow into the Indus are the rivers of Dras and Zanskar; of those that flow into the Tibetan Brahmaputra none are known to geographers except the Painom, which flows from the N. of the province of Bhutan, and falls into the Brahmaputra near the holy city of Teshoo Loombo.
The rarity of cascades, on a scale at all commensurate with the grandeur of the mountains and the volume of the rivers, is a peculiarity worthy of observation. Tibet, where absolute precipices are more numerous, is better adapted for waterfalls, but the dryness of its climate sufficiently accounts for their paucity.
The Himalayan lakes are almost confined to Tibet, where many of the rivers have their apparent sources in lakes fed by innumerable very small glacier streams. Some of these lakes are very extensive, as that of Yeumtso in Eastern Tibet, whose drainage is unknown; the Ramchoke lakes, which are said to give rise to the Painom; and the Mansarowar or Tso Mapham, and Rakas Tal or Tso Langkak lakes; which are sheets of water 20 to 30 miles long, and of considerable breadth, elevated 15,200 feet above the sea; they give rise to the Sutlej. In the Western Himalaya the principal lake is the Tso Moriri, also of 15,200 feet elevation, 30 miles long, and placed on the axis of the chain; it has now no exit, but was once of greater dimensions, when its superfluous waters (which, no doubt, were then fresh) formed a tributary of the Sutlej. The waters of the Tso Moriri, and of many other Tibetan lakes, are very salt; a fact attributed to the evaporation and drainage exceeding the influx in the case of those which have outlets, and by the further concentration of the salts in the instances of lakes without drainage, which are hence always the saltiest. The rationale of the formation of the Tibetan lakes
---
1 The geographical importance of these divisions was first indicated by Mr Hodgson in a valuable paper printed in the Journal of the Asiatic Society of Bengal. 2 The Indian source of the Brahmaputra is amongst the Mishmi Mountains of Upper Assam. These mountains form part of a system quite unconnected with the Himalayas, and which divide the waters of the Brahmaputra from those of the Irrawaddi and other rivers of Burma, China, &c. 3 II. Strachey, Phys. Geog. of W. Tibet, p. 45.