On the 5th of May I continued my journey to Islamabad, which is about eleven miles from Avantipura. The peak of Wasterwan is the termination of a long mountain ridge, which separates two large valleys from one another. Immediately to the eastward, therefore, the mountains recede from the river, and the road traverses a marshy tract, a great part of which, from the late heavy rains, was still under water, while the remainder was laid out in fields, prepared for the cultivation of rice. Further on, cliffs of lacustrine clay again rose perpendicularly from the river. Several streams joined the Jelam from both sides, some of them deep and sluggish, with straight banks like canals, while others were almost as large as the main stream, and broad and shallow, with a sandy bed and gently flowing current. Near Bijbehara, a considerable village, with many timber-built houses and a substantial bridge of deodar, the banks are beautifully wooded with shady trees. Above this village the Jelam is much smaller, often shallow, and the banks lower, though still eight or ten feet above the water, and not swampy, but either fringed with willow and mulberry trees, or bare and covered with fields of green corn, or of rape now in full flower. The bridge of Islamabad, which is the limit of navigation, is nearly a mile from the town, which is a considerable place, the next in importance to the capital, though very much smaller. It lies on low ground close by the river, but immediately behind it a long promontory of the lacustrine formation stretches back for several miles, rising abruptly out of the finely cultivated and well-wooded valley on the left, in steep, rugged cliffs, which are worn into irregular ravines by the action of rain. These formations attain here a thickness of at least 150 feet, and well deserve the particular attention of the geologist. The ancient temple of Martand, the most perfect of its class of ruins in the valley, is built on the upper and back part of this platform.

Leaving Islamabad, I crossed immediately one branch of the Jelam, which descends from the west. It had already lost the tranquil character of the stream lower down. There were pebbles in its bed, and it had a more rapid current. After crossing this stream, the country was for some distance quite flat, and entirely covered with rice-fields, now bare; some of them had been ploughed, but most were still just as they had been left after harvest. They were traversed by numerous ditches or canals for irrigation, in all of which a proportion of fresh-water shells, chiefly Lymnææ, were seen. Further on, the appearance of the country began to change: there were still plenty of rice-fields, but they rose in steps one above another, and the water in the irrigation canals flowed rapidly over pebbly beds. Crossing another branch of the Jelam, which had a broad channel full of large boulders, but shallow and easily fordable, the road began gradually to ascend a low range of hills covered with grass and bushes where it was dry, but still laid out in rice-fields wherever water was procurable. These hills, which are the termination of a long range which descends from the snow-clad mountains at the east end of the valley, are composed of a very hard limestone, the strata of which are much bent, sinuated, and fractured. On the south side of this ridge is the valley of Shahabad, which is watered by the principal branch of the Jelam. It contains numerous villages, surrounded with fine orchards, and its rice-fields are arranged in terraces. Water being plentiful, the whole valley is cultivated with rice, and the district appears to be one of the richest in Kashmir.

The general character of the vegetation continues the same as further west, and the more advanced season enabled me to recognize a few common Himalayan plants. The scandent white rose (R. Brunonis) was one of these, also Lonicera diversifolia and a shrubby Indigofera. I also observed Viola serpens, Thymus Serpyllum, Lactuca dissecta, and Fragaria Indica. Among the rice-fields several plains plants occurred, such as Potentilla supina, Convolvulus arvensis, Mazus rugosus, Salvia plebeia, and Marsilea quadrifolia. Nor were the plants of a Tibetan climate altogether wanting, for Rosa Webbiana was everywhere common, and a species of Myricaria grew plentifully among the boulders on the banks of all the streams.

From Shahabad I made, on the 7th, a short march to Vernag, a celebrated fountain near the bottom of the Banahal pass. Crossing the river, the road lay up the open valley of the Jelam, still among rice-fields, rising step by step behind one another, as the valley sloped upwards. Vernag lies close to the mouth of a little lateral valley, up which our further course lay. The fountain, which is built of marble, is large, contains many fish, and supplies a considerable stream. It is the reputed source of the Behat or Jelam, but the main branch of that river descends from the mountains a good way further to the south-west. The hills on both sides of the Shahabad valley are of limestone, the strike of which seemed to be west-south-west, or nearly in the direction of the valley. It is very much indurated, and its colour is bluish-grey; it has all the appearance of having been much altered by heat. The dip appeared different on the opposite sides of the valley: on the north it was east of north, on the other side southerly; the inclination of the beds varied much, and they were often very much distorted. I did not see any eruption of igneous rock on any part of the day's journey.

On the hills above Vernag there was a good deal of brushwood, consisting chiefly of Fothergilla involucrata, two species of Viburnum, Cotoneaster, Lonicera, and a few trees of Pinus excelsa, yew, and deodar. The opposite hills were bare and grassy. In the forests of Kashmir (as was first pointed out by Dr. Falconer) we do not find the oak, Andromeda, and Rhododendron, which are so abundant at similar elevations in the outer Himalaya. The appearance of the woods is, therefore, remarkably different, as these trees, which, in the temperate zone of the mountains near the plains, constitute almost all the forest, give the woods there a peculiar character.

On the 8th of May I passed from the valley of Kashmir into the basin of the Chenab, crossing the Banahal pass, the summit of which is not more than 10,000 feet above the sea: it is a very narrow ridge, separating two deep valleys. Starting through rice-fields, and passing at the upper limit of cultivation a few fields of barley and rape, I soon entered brushwood, the same as on the hills above Vernag. In the ravines on the left hand, snow descended below 7000 feet. Ascending rapidly on a ridge, the brushwood gave place to a fine wood of maple, horse-chesnut, cherry, hazel, and elm, all just bursting into leaf. The dip of the limestone rocks was exceedingly variable, at one time southerly, at another northerly, but the strike was, I believe, the same as the day before. The ascent continuing rapid, the shady side of the ridge was soon covered with snow; but the road kept on the southern exposure, which was sometimes bare of forest. Birch at last appeared among the other trees, and, as the elevation increased, it began to predominate. About the same time, the limestone gave place to a slaty rock, which was almost immediately followed by an amygdaloid, which continued to the summit. Both the slate and the limestone appeared to have been upheaved by the igneous rock, and I thought the slate seemed inferior to the limestone.

On the upper part of the ascent the birch gradually became more and more stunted; it was here almost the only tree, with the exception of a few specimens of Picea Webbiana, at the limit of forest a little below the summit. Here the hills were bare and rocky; but the forest did not cease on account of elevation, because on the opposite hill, which had a northern exposure, a shady wood, chiefly consisting of pines, rose to a level considerably higher than that of the pass, which was a depression in the ridge, considerably overtopped by the hills on both sides. The crest of the pass was undulating, and covered with green-sward, among which a few spring plants were in flower; these were a Corydalis, an Anemone, and Primula denticulata. A large patch of snow occupied the northern slope, just below the top.

The view from the summit would have been magnificent had the day been more favourable; but a thick haze rested over the more distant parts of the valley of Kashmir, as well as over the southern mountains in the direction of the plains of India. The southern slope of the range on which I stood was bare, scarcely even a bush being visible; and the Banahal valley, nearly four thousand feet below, appeared as a perfectly level plain, covered with rice-fields and scattered villages, marked by groves of trees. On the descent I followed a very steep rocky ridge. About half-way down, the amygdaloid was replaced by metamorphic slate, and for the remainder of the descent the rocks were alternations of slate, very hard conglomerate, and quartz rock. The dip of these strata was very variable, and on the face of several spurs, at a little distance, sections were exposed, exhibiting enormous flexures. I saw no limestone on the southern face of the pass, except in the valley of Banahal, where there was a good deal of a horizontally stratified limestone, very different in appearance from that on the other side, which, as it was confined to the bottom of the valley, and was there very local, appeared to be of much more recent origin.

After joining the Banahal river, the descent became more gradual. At first, the valley was almost level and quite covered with rice-fields, all under water. The villagers were busy ploughing, both bullocks and men knee-deep in soft mud. Further on, the valley contracted, and cultivation only occurred at intervals. In the narrower parts, the stream was fringed with trees, but the hill-sides were still quite bare. Round the villages there were very fine trees, chiefly walnut, horse-chesnut, and elms, with the ordinary fruit-trees; but the plane and black poplar do not occur, nor are any vines cultivated in the valley. The winter is said to be quite as severe as in Kashmir; and the elevation, so far as I could determine it by the boiling-point of water, is a little greater, the lower villages (in one of which I encamped) being about 5500 feet, while the highest fields are about 6000 feet. In the woods, Fothergilla, cherry, sycamore, and horse-chesnut were common, just as in Kashmir. The season was much further advanced than on the north side of the pass, all these trees being fully in leaf, and the horse-chesnut in flower. The greater part of the vegetation was identical with that of Kashmir, but I saw many more species, probably only from the more advanced state of the season. The Zizyphus and rose (R. Webbiana) of Kashmir were still common, and the white poplar was wild along the banks of the stream. I did not, however, see Daphne or Myricaria. In shady lateral ravines an oak was frequent, the more interesting as I had seen none in Kashmir; it was Q. floribunda, a species of the middle zone of the outer Himalaya, which usually occurs at higher levels than Q. incana, and lower than Q. semecarpifolia.

Though the river of Banahal is a tributary of the Chenab, yet the district has always been considered as a dependency of Kashmir, from which it is only a short day's journey distant, while for several days in descending towards the Chenab, the country is almost uninhabited. Halting one day at Banahal to change my porters, I made three marches to Nasmon, on the right bank of the Chenab, following the course of the Banahal river during the first and part of the second march, but afterwards leaving it, on account of its increasing ruggedness, to cross the range on the left hand by a pass about 8000 feet above the sea, which overhangs the valley of the Chenab. The bounding spurs which hem in the Banahal valley descend almost perpendicularly upon the Chenab, and dip at last very abruptly to that river. At first, large masses of snow were visible at the sources of all the lateral valleys, but lower down the elevation was not sufficient, and the hills were bare. After leaving the last village of Banahal, the bottom of the valley was for some time level and covered with fine forest, consisting chiefly of magnificent trees of Celtis, elm, and alder; the others were two species of Acer, Fraxinus, Morus, Populus ciliata, and a willow. Fothergilla now grew to a small tree, and Marlea made its appearance, the first indication of an approach to a hot climate. Soon, the banks of the river became rocky, and left no passage, so that the road ascended on the right bank, and lay at a considerable elevation on the hill-sides, looking down upon a richly wooded and often rocky glen. The hills were steep and generally bare, but the ravines were often well wooded. Pinus excelsa occurred occasionally; Quercus floribunda was common, and Q. lanata made its appearance.

Before leaving the Banahal river, I had got down to about 4000 feet, meeting latterly with some familiar plants of the warmer zone: Pinus longifolia formed dry woods, Cedrela Toona, a fig, Albizzia mollis, and last of all, Dalbergia Sissoo. Still, most of the plants of the upper part of the valley accompanied me throughout; even the hoary oak had not disappeared, and the general appearance of the vegetation was very different from what it would have been at the same elevation further east, the plants of a hot climate being chiefly such as delight in a dry heat, and are capable of enduring a considerable amount of winter cold, provided the summer temperature be sufficiently elevated. It was evident that the temperature was considerably lower than it would have been at the same height in the Sutlej valley, and I drew the same inference with regard to the humidity, from the appearance of a number of dry-climate plants; for instance, a yellow spinous Astragalus, a Dianthus, and Eremurus, an Asphodeleous genus common in Kunawar, and other dry valleys of the Himalaya.

In the ascent of the lateral ravine, towards the pass above Nasmon, I encountered, for the first time, Rhododendron arboreum and Andromeda ovalifolia, the two trees which, with the hoary oak, form the mass of the Simla woods. The forest was now very fine, as I was on the northern slope of the range. On the upper part of the ridge by which I ascended, there was a grove of fine deodar-trees, and in the bottom of the dell a shady wood of horse-chesnut and sycamore. I had now entered a zone in which the flora was quite similar to that of Simla; Fothergilla being the only tree I observed, which is not common in that district. And it was curious that it was on the northern and most shady, as well as most humid exposure, that this identity of flora became first remarkable, and that the same trees which at Simla form the forests of the drier slopes and more exposed situations, grew in this valley low down on the hill-sides, in the most sheltered spots.

The ascent towards the ridge was latterly steep, with a good deal of silver fir and deodar. The trees rose to the very top of the northern slope, but, as usual, the summit was bare and grassy, though the tops of the trees were actually higher than the crest of the ridge, and obscured the view to the north. As the elevation was only 8000 feet, there was no peculiarity of vegetation, all the plants being those of the middle zone, except the silver fir, which descended to a lower level than it usually does in the Simla hills. There was some cultivation of wheat and barley within a very short distance of the summit, which overlooked the valley of the Chenab; and as the day was fortunately clear, there was a very fine view. The ravine through which the river flowed appeared everywhere rugged, more especially towards the plains, where a succession of steep rocky hills were seen, the nearest of which surrounded the mouth of the Banahal river. Across the Chenab, a high range, beautifully wooded, ran parallel to the river, rising into a snowy peak nearly opposite to me. This peak, which concealed all view of the plains beyond, lay on my road to Jamu, and was about 9000 feet in height.

The descent to Nasmon, which is only 2700 feet above the level of the sea, was very steep. At first it led along the face of a bare hill, but soon entered a shady ravine, filled with alder, oak, walnut, and Celtis, but without any of the superb horse-chesnuts which had been so abundant in the humid valleys on the northern face of the range; nor was there any Rhododendron. Crossing a considerable stream, the road ascended through fine forest to the crest of a ridge, beyond which there was a long and steep descent of at least 1500 feet, to the village of Nasmon, on which tropical vegetation made its appearance very abruptly. Pinus longifolia grew scattered along the sides of this hill, and Daphne, pomegranate, the olive of the Sutlej valley, Vitex Negundo, Colebrookea, Rottlera, Sissoo, Adhatoda Vasica, a thorny Celastrus, Acacia modesta and Lebbek, and Bauhinia variegata, made their appearance in succession, in the order in which I have named them. Most of these are the same as the shrubby forms common in the Sutlej valley at Rampur; but the Celastrus and Acacia modesta are plants of the plains of the western Punjab, and do not extend so far west as that river. The range parallel to the Chenab on the north, which I had just crossed, has probably a granitic axis, for boulders of granite were common on the upper part of the ascent on both sides of the pass, though I did not anywhere see that rock in situ. On both sides the first rock exposed was a fine-grained gneiss, with large crystals of felspar. Lower down, on the north face, I observed mica-slate, with garnets; and in the bed of the Banahal river ordinary clay-slate occurred.

Nasmon is a very large but scattered village, with much cultivation. It lies on a high platform of alluvium, considerably above the bed of the river. Plane, orange, apricot, and pear trees grew in the gardens, with Melia Azedarach, and a few trees of the European cypress (C. sempervirens), bearing apparently ripe fruit. The day was oppressively warm, the thermometer rising above 85° in the shade.

On the 13th of May, I crossed the Chenab by a bridge about a mile above Nasmon. The descent to the bank of the river was gradual, and very bare. Rocks of a black clay-slate and of conglomerate, in nearly vertical strata, formed the bed of the river, which was as large as the Sutlej at Rampur, and very much swollen and muddy. The bridge is the simplest form of jhula, a single set of ropes, from which a wooden seat is suspended, which is pulled from side to side by means of a rope, worked from the rocks on either side of the river. The banks of the river were adorned with a profusion of bushes of Nerium odorum, in full flower, and highly ornamental. The vegetation along the river exhibited the same curious contrast of tropical and temperate forms, which I have already described as characteristic of the dry valleys of the interior of the Himalaya, at elevations between two and four thousand feet; and the tropical plants were so similar to those which I observed on the Sutlej, that I need not particularize them. There was no forest in any part of the valley near the river, but a few trees of Pinus longifolia grew scattered on the bank; and on the stony ground which skirted the stream, there was a low jungle of the same tropical shrubs as had occurred on the lower part of the descent the day before. I saw also Zizyphus nummularia, a shrub which is eminently characteristic of a dry climate, being common in the most desert and rainless districts of the Punjab. The shrubby temperate forms were not numerous, being chiefly Rosa Brunonis, and the Himalayan pear, Lonicera diversifolia, Myrsine bifaria, and Jasminum revolutum, all plants which have a very wide range in the Himalaya.

Passing through the bush jungle which skirted the river, I entered a large tract of almost level cultivated land, covered with fields of barley, ripe and partly cut. One or two plantain-trees, and some buffaloes, were signs that we were still in a very hot region. Crossing a considerable stream, the road began to ascend rapidly on a narrow ridge. Passing some farm-houses, surrounded by fields, I entered a scattered wood of wild olive-trees (Olea cuspidata), mixed with Zizyphus and wild pomegranate. The young shoots and panicles of the olive were abundantly covered with a white floccose glutinous matter, the source of which I could not exactly determine; but I could see no trace of any insects by which it could have been formed, so that it was perhaps a natural exudation from the tree. Small woods of Pinus longifolia occurred at intervals, almost alone, for few plants seem to thrive under its shade. At 4000 feet, while the olive and pomegranate were still abundant, Quercus lanata appeared. At 4500 feet, which was about the upper limit of the olive, I re-entered a cultivated district, disposed in terraces on the slopes of the hills. The barley was quite ripe, and being cut, but the wheat, though in full ear, was still green. There were also a few fields of the opium poppy in full flower, and of safflower (Carthamus tinctorius), which was not nearly so far advanced.

I encamped at the village of Balota, elevated 5000 feet. Round the village were some very fine table-topped deodars, perhaps the relics of a former forest, though more likely planted by the villagers. The hills on all sides were richly cultivated, as far up as 6000 feet, above which elevation fine forest commenced; and the snowy top of the mountain behind, which I had seen from the pass of the 12th, was visible rising behind the forest. During the whole of the ascent from the Chenab, the rock was a coarse-grained sandstone, in highly inclined strata, generally of a reddish-brown colour, the surface of which rapidly passes into a state of decay.

The range of mountains to the south of the Chenab, by which that river is separated from the basin of the Tawi or river of Jamu, still lay between me and the plains of India. On the 14th of May, I crossed a spur from this range, descending into a valley watered by a tributary of the Chenab. This ridge, which is called Ladhe ke Dhar, rises a little above 9000 feet, that being the elevation at which the road crosses it. After leaving the cultivated lands of Balota, the ascent, which was steady, lay through fine brushwood and stunted oaks. On the banks of the stream, which occupied the centre of the valley by which I ascended, sycamore, horse-chesnut, and cherry, were abundant. On the slopes there were a few trees of Pinus excelsa and Picea, but the forest was not dense. About 7000 feet, on the north-western face of a spur, there was much cultivation of wheat and barley, hardly yet in ear. Here there was a fine view in the direction of the upper valley of the Chenab, of rugged mountains, scarcely wooded on the slope exposed to view, rising behind one another, the more distant still heavily snowed. Higher up, the forest was chiefly formed of the holly-leaved oak, but the latter part of the ascent was through a dark forest of silver fir, intermixed with a few fine yews. The underwood here was chiefly Viburnum nervosum, still in flower, though its leaves were almost fully developed. On emerging from this gloomy forest, in the upper part of which there was a thin sprinkling of snow, I found myself on the crest of the range, which was bare and rounded. Snow lay in large patches, and had evidently been till very recently continuous over the whole top, as vegetation was just commencing, and few plants were in flower. Primula denticulata was common, as well as a little gentian, which extended on both sides at least 2000 feet lower; the only alpine plant was the little Callianthemum which I had found some days before on the summit of Wasterwan in Kashmir. The distant view was unfortunately quite obscured by haze, so that I could not see, as I had expected, the plains of India.

In descending the southern face of this mountain, the road at once entered a forest of silver fir, in the upper part of which I saw one tree of Quercus semecarpifolia, a species which I had not met with on the Kashmir passes, or anywhere since leaving the Sutlej. About 8000 feet, the pines were replaced by the holly-leaved oak, forming open woods, in the glades of which patches of cultivation soon occurred; I encamped at about 7000 feet, at the village of Katti. During the day the sandstone rock occurred uninterruptedly, partly, as the day before, of a reddish-brown colour, partly grey, or nearly white. On the descent large angular fragments of this rock were everywhere scattered over the surface, almost always more or less imbedded in the soil: these had somewhat the appearance of a former moraine, but the surface was so much covered with wood, and the boulders were so much buried, that I could not trace their arrangement in a satisfactory manner.

Next morning I continued the descent, which was rapid, so that I soon arrived at tropical vegetation. There was but little forest, except in ravines, and the heat soon became very great. About three miles from Katti I passed the fort of Landar, built on an almost isolated cliff, overhanging the ravine; and a little further on I descended abruptly to a small stream, running towards the Chenab, the elevation of whose bed was about 3000 feet. The descent, which was almost precipitous, led down the face of a mass of clay, in some respects like the alluvial deposits so common in Tibet. Similar masses of alluvium, all table-topped, and very steep, and much worn by ravines, had occurred throughout the whole of the descent from Katti. A few pines grew on this steep bank, and all the shrubs which I had found on the banks of the Chenab at Nasmon were again met with. After crossing this stream, the bed of which was filled with large water-worn boulders, I again ascended to about 5000 feet, chiefly among cultivation, and encamped at Mir, a small village close to the crest of the main range south of the Chenab, the elevation of which was now very inconsiderable.

Next day, a gentle ascent of half an hour brought me to the crest of this range. The mountain slopes were bare and grassy, but in the ravines there was now and then some brushwood. Andromeda ovalifolia and Rhododendron arboreum were both noticed; and, much to my surprise, I observed at intervals a few trees of Fothergilla, for I had not expected to find this Kashmir tree so close to the plains, and in a district the flora of which was so completely that of the Simla hills. On the summit of the pass, which was not more than 6000 feet, I found a beautiful gentian (G. Kurroo of Royle) and a yellow spinous Astragalus, seemingly the same species which I had found at Nasmon, on the Chenab. It was curious to find a representative of the spiny-petioled group of this genus in so hot a climate and so near the plains; for in the rainy parts of the mountains, and in the more humid parts of the Indian plain, the genus is almost wanting, and this particular section entirely so.

From the summit I descended at once through a pine-wood to the bottom of a valley, the course of which I followed throughout the day in a southerly direction. It gradually widened as I advanced; villages became frequent, and were surrounded by extensive cultivation, and all temperate vegetation disappeared. I encamped at the village of Kirmichi, where the valley which I was following appeared to expand into an open plain of some width. Here oranges and mulberries were cultivated in gardens, and the toon and mango, pipal and banyan (Ficus religiosa and Indica) were planted in groves round the houses.

On the 17th of May, I continued my journey towards the plains of the Punjab. An open, somewhat undulating valley lay before me, appearing to stretch from east to west, and to be bounded by two ranges of hills which had the same direction. Trikota Debi, a curious three-peaked hill, the last culminating point of the range separating the Chenab from the Tawi, rose some miles to the westward. To the eastward the valley of the Tawi was open as far as Ramnagar, which was distant about twenty miles. In crossing this open plain, or dhun, I nearly followed the course of a little stream which had excavated for itself a deep channel in the soft sandstone of which the plain was composed. This rock was very different in appearance from the red or grey sandstone which had accompanied us from Balota; it was pure white, and almost horizontally stratified, while that was always highly inclined. During the latter part of my journey of the 10th I nowhere saw rock in situ, so that I had no opportunity of ascertaining the contact of these two formations, which are probably of very different epochs, the sandstone of the open plain being certainly the Sewalik tertiary formation, while the red sandstone of the higher mountains, which in the total absence of all organic remains is as yet of uncertain age, is perhaps the same as the gypsiferous and saliferous sandstones which skirt a great part of the western Himalaya.

One or two pine-trees, and some bushes of Euphorbia pentagona, were almost the only features in the vegetation which distinguished this open valley from the plains of India. On shady rocks along the stream three or four ferns were common; the oleander also grew near water; a dwarf date-palm occupied drier spots; and I saw a few trees of Cassia fistula. Crossing a broad shallow river which flowed to the eastward at the southern boundary of this dhun, in a depression faced by cliffs of sandstone, I entered among low hills covered with scattered trees of Pinus longifolia. This plant appears to grow luxuriantly on hot dry hills; the trees did not attain a great size, but appeared vigorous and healthy, with thick trunks and gnarled branches, exactly like the Scotch fir, except in the great length of the leaves, which are pendulous from the ends of the branches.

On the 18th, I crossed a sandstone range, in which the strata exhibited an anticlinal axis, dipping towards the plain on both sides. The ascent was easy, and the summit was not above the limit of tropical vegetation, as a banyan-tree grew on the top. The descent was much steeper and considerably longer, the valley to the south being a good deal lower. The road was good, being in the steeper parts paved with large flat stones, while in the more rocky parts the sandstone was cut into steps. A flat and well cultivated valley lay to the south of this range, in the centre of which flowed a river, in a wide channel several hundred feet below the level of the plain: it was very shallow, and was crossed by stepping-stones. Another hilly tract followed, covered with straggling bush jungle, and on the upper part with pine-forest: this was also of sandstone, very soft, and excavated by the various little streams which traversed it, into narrow and deep ravines. Even foot-paths, by constant use, were sunk four or five feet deep in the soft rock. The dip of this range was gentle, towards the plains of India.

I encamped on the 18th at Seda, under the shade of a superb banyan-tree, in a hollow in this sandstone range, and next day continued my journey to Jamu. Emerging from the hills after a mile or two, I entered a third valley, and followed the course of the little stream by which it was watered, to its junction with the Tawi, along which I travelled about four miles; to the town of Jamu, which is built on the outermost range of hills, at the point where the river Tawi finally quits the mountains. These hills rise very gently from the plains, their southern slope forming a long inclined plane, densely covered with a jungle of low thorny trees. The same sort of jungle usually skirts their base to a distance of two or three miles, or as far as the alluvial soil of the level country which lies beyond is covered with stones and shingle. It is principally composed of Acacia modesta and Catechu, and of two species of Zizyphus. The northern or inner face of this range of hills is very steep, often quite precipitous; and where they overhang the Tawi, they terminate abruptly in a line of cliffs facing the river. A similar range, but a good deal lower, descends from the eastward towards Jamu, and, like the other, presents a series of vertical cliffs covered with brushwood towards the river. The town occupies the gentle slope which faces the plains; it is a straggling and dirty place, but with some very good houses. The principal building is the residence of Maharaja Gulab Sing; at the time of my visit occupied by his eldest son. It is situated on the edge of the cliff, overhanging the river, and commands a fine view of the open valley of the Tawi below, and of the mountain ranges to the north and east, the more distant of which were still tipped with snow.

The outermost range of hills, which does not rise to any great elevation, consists entirely of loose conglomerate coarsely stratified, the beds dipping very gently towards the plains. The boulders of which it is composed are waterworn, and very various in composition, but all referable to the interior ranges; a few thin beds of sand and of a clay resembling pipe-clay, are interposed between the strata of conglomerate.

The very curious country through which I had been travelling since the 16th, had so much the appearance of a succession of valleys parallel to the plains, and separated by long ranges of hills, that it was difficult to avoid taking up that impression, which, notwithstanding, I believe to be an erroneous one. The gentle slope of the different tributaries which join the Tawi from the right and left, tends to keep out of sight the longitudinal ranges parallel to that river, from which the lateral ramifications proceed. When we obtain a detailed survey of the district, it will be found that the lateral valleys on each side of the Tawi do not correspond in direction, and are not quite opposite to one another, and that the apparent uniformity is caused by the great width of their valleys, when compared with the elevation of the bounding ranges. The Sewalik sandstone here attains a width of at least thirty miles, which is very much more than is found further west.

CHAPTER XI.

Leave Jamu to return to Tibet—Lake of Sirohi Sar—Vegetation of lower hills—Dodonæa—Ramnagar—Garta—Dadu, on a tributary of the Chenab—Camp at 10,000 feet—Badarwar—Padri pass—Descend a tributary of the Ravi—and ascend another towards the north—Sach Joth, or pass—Snow-beds—Camp in Chenab valley.

On my arrival in Kashmir, I had forwarded an application to the Indian Government, requesting permission to return to Tibet, for the purpose of visiting the mountains north of Nubra, which, from the advanced state of the season, I had been unable to do the previous year. Soon after reaching Jamu, I received intimation that the Governor-General, Lord Dalhousie, had been pleased to accede to my request. I had already determined, if permitted to return to Le, to take the route by Zanskar, which, though much frequented by the natives of the country, was quite unknown to European travellers; but as the season was far advanced, I chose a road through the higher hills, instead of taking that leading directly to Chamba, which would have obliged me to travel for at least a week through the hot valleys of the outer ranges.

I left Jamu on the morning of the 23rd of May. After crossing the Tawi by a ferry immediately below the town, my road lay for three miles up the left bank of that river, along an open sandy plain, only very partially cultivated. I then turned to the right, and entered the low hills which skirted the plain on that side. The road generally followed the course of the ravines, which have been excavated out of the soft sandstone by the numerous tributaries which descend to join the Tawi. These streams are all of small size, with gravelly or sandy beds, and are separated by low ridges of some breadth, faced generally by perpendicular cliffs. An undulating country of this nature occupies the whole of the space which intervenes between the outer range of hills and that next to it, a distance, by the road along which I travelled, of about twelve miles. This second range is a branch given off by an axis, whose direction is nearly east and west. The road ascended to it by a very steep rocky path, after surmounting which I found myself on a considerable tract of nearly level ground, partly occupied by a pretty little lake, with grassy banks. On the banks of this lake, which is called Sirohi Sar, and is rather less than half a mile in length, I encamped on the 24th of May, in a grove of very fine mango-trees. The depth of the lake did not appear great, its margins being for a considerable distance very shallow, and producing an abundance of reeds and water-plants, among which the sacred Nelumbium, with its gay flowers, was conspicuous. The elevation of the lake, as deduced from the boiling-point of water, I found to be 2200 feet. It occupies a depression in the top of the ridge, being surmounted on both sides by low ranges of hills, rising only to the height of a few hundred feet. At the east end, a low flat plain, interrupted only by a few regular rocky knolls, seemed to indicate that the size of the lake had formerly been more considerable than at present.

The vegetation of the country between Jamu and Sirohi Sar was entirely of a tropical character. The rocky hills were in many places covered with thinly scattered pines, all of small size, and generally with much-contorted trunks, but apparently healthy and vigorous. In the cultivated grounds the plants were identical with those of the plains, but, as is usual in all hilly countries, the barren tracts produced a flora of a different character. Nerium odorum was abundant on the banks of streams, and I met with Cassia fistula, Punica, species of Rhus and Casearia, as well as the curious Euphorbia pentagona, and now and then the beautiful Bauhinia Vahlii. Acacia modesta and a Zizyphus were the most common trees. The lake produced a great variety of water-plants, but except an Alisma and Dysophylla, both of which were new to me, the species seemed all natives of the plains.

On the 25th of May, I proceeded along the side of the ridge in an easterly direction, passing several small flat-bottomed depressions, apparently the sites at a former period of small lakes, similar to that from which I had commenced my march. The road was rocky and rugged, and gradually rose several hundred feet to the crest of the ridge. Pine-trees were generally plentiful. On reaching the top, shortly after daybreak, a fine wide undulating valley was seen below, bounded on the north at the distance of about ten miles by a third range of mountains, and traversed by several streams, which had excavated for themselves deep perpendicular-sided ravines in the sandstone strata. All these streams had a westerly course to join the Tawi, which, issuing from a deep valley behind the third range, crossed the open plain in a south-west direction.

Leaving the ridge, the road descended gradually to the plain, and after crossing a deep ravine, with precipitous walls, continued through a fine level country to the village of Thalaura, about a mile from the third range of hills. The sandstone frequently contained a few waterworn pebbles scattered through it; and a bed of coarse conglomerate, with an indurated matrix, capped the cliff above this ravine. Some strata of indurated clay and soft slate also alternated, but rarely, with the sandstone. The plain was well cultivated, being chiefly laid out in rice-fields; and the people were all busy ploughing, sowing rice, and harrowing with a log of wood, drawn by bullocks and kept down by the weight of a man.

On the earlier rocky part of the road, the vegetation was much the same as the day before. Dodonæa was common, as it is in most parts of this hilly tract, never, however, rising out of the tropical belt. I do not know how far to the eastward of Jamu this plant extends; but as it does not seem to occur to the east of the Sutlej, and probably stops much sooner[18], it appears to prefer a rather dry climate, and will, I think, be found limited to the drier portion of the Peninsula, from which it probably extends through Central India, and along the hilly country west of Sind. On the open plain the pines entirely disappeared, and the aspect of the vegetation was entirely that of the plains of India.

From Thalaura I marched, on the 26th of May, to Ramnagar, crossing the third range of hills, the ascent of which was at first very steep and rocky, over a made road, paved with large stones, in many places much out of repair. This range was also sandstone, dipping to the north at a gentle angle; some strata of indurated clay occurred between the beds of sandstone. These hills were precipitous to the south, and sloped gently towards the north, in the direction of the dip. The tree Euphorbia, which, with its stiff fleshy branches springing in verticils of five from the stem, forms a striking feature in the vegetation of the lower hills, was common on the ascent, and the yellow spinous Astragalus, which I had observed between the Chenab and Jamu a fortnight before, was frequent on both sides of this ridge; but even at the top, except one species of Indigofera, no plants indicating elevation were met with: on this account I omitted to determine the height of the range by the boiling-point of water, but comparing its elevation with that of Ramnagar, which was in sight, I estimated that it might be about 3600 feet. To the north lay another valley, considerably more rugged than that crossed the day before, and evidently much more highly inclined, as its eastern termination was not far distant. This valley was traversed by the principal branch of the Tawi, the source of which is in the mountains east of Ramnagar.

The descent from this range was very gradual, the road running obliquely to the eastward, among scattered pine-trees, over bare sandstone rocks, till it reached the bank of a small stream separated from the Tawi by a low range of hills. During the descent, a number of plants of Himalayan forms made their appearance, which had not occurred before: these were a berberry, Rubus flavus, and Myrsine bifaria. Olea cuspidata was seen lower down, and a species of alder grew in shady ravines along the edge of the stream. In the bottom of the valley, the mixture of the forms of the middle and lower zones was curious and interesting. Pinus longifolia occurred with Phœnix sylvestris, alder with Rondeletia and Rottlera, pear with Sissoo, and Fragaria Indica and Micromeria with Trichodesma and Solanum Jacquini. At the same time, it was evident that in this dry stony valley the tropical species, which formed the majority, were more at home than the stragglers which had descended from above.

After ascending for a short distance along the banks of the little stream, the road crossed it, and after a short steep ascent from the right bank, the remainder of the day's journey was nearly level, along the sides of hills, or over a high table-land to Ramnagar, a small town and fort, formerly the residence of Rajah Suchet Sing, since whose death the place has been rapidly falling to decay, most of the shops of its well-built bazaar being now empty. There were in the neighbourhood one or two large gardens, in which the trees and plants were nearly all Indian, Sissoo and Melia Azedarach being the most common. A single plane-tree was scarcely an exception; for though undoubtedly more at home at greater elevations, the plane (like the poplar and many of the fruit-trees of temperate climes) does not refuse to grow even in the plains, as is proved by the occurrence of a number of trees of it of considerable size and apparently healthy in gardens at Lahore.