Heinrich Handel-Mazzetti: [A Botanical Pioneer In South West China] - chapter 6


Chapter 6. To Lijiang via Yongning and Yongsheng

We take charge of the caravan — subalpine pastures and scrub — rainstorms — the lake and lama monastery at Yongning — landscapes resplendent with flowers — leeches — a fieeble-minded prince— a chain-bridge over the Jinsha Jiang

Having sent off a few crates of herbarium specimens by a packhorse caravan to Huili, we left Yanyuan on 11 June to explore the mountain ranges of western Yunnan. Our route led along the river to Meiyu and then north-westwards over an infilled basin and a low ridge, the latter being the southwest extension of Mount Zala Shan. We then came to a sizeable mountain chain, tectonically perhaps the southward continuation of Mount Lama Shan; it juts out to the south from the Litang watershed and forms the western boundary of the Yanyuan basin. That morning the mafu once again delayed our departure for several hours by various underhand tricks, and that evening, disregarding our wishes, he sought quarters in another village, so that we had to send a messenger in darkness and rain to fetch him to the place where we had arranged to stop — the village of Duer Liangzi, which was on the east side of the low ridge just mentioned. Back in Yanyuan he had held his tongue, as he knew that we could easily call on the magistrate's help, but now hfe' again started to ask for more money. This time it was more than we could stand. Schneider took charge of the caravan and assumed direct responsibility. He got rid of the rascally mafu by leaving him two pack animals, and kept the remaining twenty three for us. All the same we had been cheated out of at least six hundred dollars, and we also had to settle debts amounting to fifty dollars which the mafu had run up with our coolies and other people. Two of his men defected with him, and he declared that he was returning to Kunming to claim the money from his employer — though it appeared that no such person actually existed. The next valley was a broad depression which reached its lowest level along its western ridge. It was filled with karst heath, normally arid, but now putting out green foliage and flowers of many colours. Some of the plants, though dressed in sober hues, were of arresting form, for example Arisaema consanguineum, the thirteen points of its palmate leaf, like the spadix projecting from its green spathe, being drawn out into slender threads 5cm long. A few deeply incised U-shaped valleys ran down from the north, but not much of the water in their streams reached the plain. On the western side was the Xifan village of Hungga at an altitude of 2900m, somewhat above the outer margin of the Yanyuan basin. In this terrain erosion produced grotesque appearances, both large-scale and small. I measured a runoff channel hidden among the bushes: over a long stretch it was 7m deep yet at ground level only 30cm wide.

The route which we took on 13th June led at first steeply up the mountainside — far too steeply for our caravan's comfort. Among the bushes were the tall stems and yellow blossoms of Incarvillea lutes, on bare patches of sandstone we saw the prostrate Potentilla ambigua with its large yellow flowers. Before long we entered the mixed forests of the temperate zone, which were now approaching their annual climax. Thanks to the abundance of conifers and the fresh green leaves of deciduous trees unfolding in the spring, these mixed woodlands were much less monotonous than primeval tropical forests. All shades of green were represented, and the trees varied enormously in habit and leaf size. There were small-leaved cherries and maples with large, lobate leaves (Acer multiserratum), both having wide crowns; silver-leaved poplars with foliage in loose narrow crowns and various species of Sorbus, notably S. vilmorinii with pinnate leaves. Soft-needled firs (Picea likiangensis and others) overtopped the broad-leaved trees, while the sombre green yews and Tsuga had broad spreading branches. The thick leathery leaves of the tree rhododendrons and the evergreen oaks made a dark background which showed off the flowering shrubs to perfection. Many of the finest plants which adorn our gardens come from the forests of Western China, including paeonies, various species of Philadelphus and Deutzia with white and pink blooms, Syringa yunnanensis, the splendid pink Neillia longiracemosa and several kinds of Berberis. Scrambling over the shrubs was Clematis montana with sheets of white flowers. Thick pads of dark lichens (Lobaria, etc.) covered every projecting branch, especially the dead wood, and pale beardmoss (Usnea longissima) hung down from the trees in strands several metres long. Further into the dark interior of the forest green and golden mosses festooned the twigs and branches, while other species formed thick cushions on the trunks. However, the most striking feature of the montane forests of southwest China is the bamboo. It formed the understorey everywhere, and at the forest margins and in clearings it grew in many-stalked clumps 3 to 4m tall with short twigs bearing whorls of pale green leaves. In the shade were plants such as Smilacina lichiangensis, not unlike lily of the valley, Lappula dielsii with forget-me-not flowers, and Arisaema lobatum. Flourishing on open pastures outside the forest was Neillia gracilis, a dwarf shrub with racemes of single pink flowers, and Triosteum hirsutum, a lush herbaceous plant with evil-smelling glands.

As we climbed up we soon emerged from this zone on to one of the crests of the Da'erbi ridge, which is about 3km across and consists of stratified limestone in horizontal contorted beds. Here we had a splendid view of the Yanyuan basin and paused to take photographs. Its river curved round to the south and ran far into the distance before cutting through our mountain range, which reached its highest point in the north where it joined the main chain. In the troughs formed by the buckling of the limestone strata vegetable mould had accumulated to great depths, and here, at an altitude of 3775m, bordered by willow bushes and clumps of pine with rhododendrons and bamboo, were tracts of turf resplendent with low growing herbaceous plants in full flower. The vegetable mould was derived from the thick roots of numerous different species and from the weathering of the outer leaves which [p.29:] sheathed their buds, and despite its dryness it yielded beneath the foot Growing between the blue mounds of Rhododendron hippophaeoides was Incarvillea gtandiflora with its asymmetrical pink trumpets; beside it were the yellow bells of Nomocharis lophophora [note #26: Now Lilium lophophorum.], which might be taken for a Fritillaria, the purple blooms of Salvia pinetorum and the tubular purple flowers of Morina delavayi in dense inflorescences. Held up on erect stems were the stars of Anemone demissa, A. coelestina and A. rupestris, together with the orange yellow blossoms of Taraxacum eriopodum. Mandragora caulescens, its short juicy stem topped with a tuft of hairy leaves, put out long leafless spreading flower stalks, often in great numbers, ending in large nodding brownish-green bells. The thickets consisted of willows (Salix tenella) with Sorbus vilmorinii, Lonicera chlamydata, Sibiraea leyigata and other shrubs. At their margins and extending deep into their shadowy interiors we saw tall slender plants of Meconopsis fortestii with blue flowers, a Corydalis (probably C cheirifolia) and the yellow C. yunnanensis, more than one species of Smilacina with racemes of whitish or greenish brown flowers, and Arisaema wilsonii with purplish brown spathes up to 25cm long.

As we began the steep descent on the west side another superb vista opened before our eyes. The rain had cleared the air, and only in the distance were the mountains high enough to tower into the clouds. We gazed down over a broad depression filled with wooded crests all of much the same height (3500m) and all built of almost horizontally stratified rock. Between the crests were patches of uneven ground, and the landscape was incised by the nanrow valleys of the rivers which traversed it. Further away was a range projecting above the forests and breaking off sharply at a comer to the south of Yongning [note #27: There were two towns called Yonging, this one in Yunnan and another in Sichuan (SGH).]. Behind that was a tall mountain of paler rock, probably Xuechou Shan in the Yangzi loop, and to the left there were a few dark spires belonging to the gigantic Lijiang range, the rest of which was hidden in mist. Before us to the west an impressive rounded summit rose out of the depths; it was Shizu Shan and marked the position of Yongning, our next destination. Passing two Yi villages, the track plunged steeply downwards, losing 1200m in altitude, and then ran almost level for some distance between screes and huge boulders along a dried up stream bed, where it would have been suicidal to have stayed on horseback, Pelted by thunderstorms, we arrived as dusk was falling at the lonely ferry which crossed the Wolue He where it flowed between cliffs of limestone and tufa in a gorge with an attractive shrub vegetation. The Yanyuan river (Yanjing He) joins the Wolue He here, and they run northwards together to the Litang. I felt ready to take an oath that this would be the last journey I would undertake in remote mountains, little dreaming how well I was to get to know the rainy season in the Yunnan Alps. About an hour later we reached the village of Wolue He, too tired to do anything except stack our rich booty in separate piles, one for each place. If one omits this task, the memory will be clouded by even a single night's sleep. The work of arranging the specimens between paper we left, however, until the following morning.

Next evening, after a steep climb up paths on slippery clay, we reached Fumadi, another Chinese village, though the surrounding population was Yi. Riding onwards over a broad ridge covered with spruce forest, at an altitude of 3300m, we were left in no doubt that the rainy season had begun. The ridge was built of limestone and sandstone, and the rocks were so hard that the streams had barely cut into them. A short distance further on the meadow flora at this fairly lowlying spot had now developed to a stage where the blue Sttvbilanthes versicolor was flowering in abundance though the plants were still quite low. With it were the blue flowers of Iris bulleyana together with a small golden brown orchid, Oreorchis erythrochrysea, a new species. On the descent we passed through almost pure woods of Pinus armandii, a tall pine with bluish green needles and cones 20cm in length and 12cm in diameter. Beneath it was bamboo undergrowth. Further down we encountered a fair-sized grove of Mahonia alexandri, each trunk rising to about 3m before branching, and carrying large spiny pinnate leaves and multiple racemes of yellow flowers. Two streams, coming from north and northwest, joined here. The scattered Naxi village of Gaitiu lay in a cultivated though rather dry basin. We soon left the stream which came from Yongning, curving round in a wide branch to the north, and ascended gradually up a tributary which was the outflow from the Chuosuo (Lugu Hu) lake. Here on the sandstone, and later even more frequently, we found some distinctive woods composed of aspens, Pinus yunnanensis and the large-leaved Quercus dentata. The trunks of these oaks were often completely clothed with Drynaria delavayi, a fern which looked as if it were an enlarged version of our polypody with broader fronds. However, beneath these fronds it bore faintly lobed epiphytic leaves, already brown and withered, which closely covered the tree trunks, their function being to collect water and nutrients for the thick creeping rhizomes. Before long the valley levelled out and its floor became broad and flat, filled with black marshy soil which supported rich green meadows. The slender yellow Ms fomestii grew in large clumps among tall red primulas now coming into flower. This plant - Primula beesiana — is one of a section found in southwest China. All its members have tall scapes with several superimposed whorls of flowers, in some species flat and open, in others bellshaped, ranging in hue from pink to carmine. The wide valley skirted round the left of the lake, and on a lateral stream we saw the yamen of the Naxi prince of Chuosuo, a timber farmhouse where the old gentleman, to whom we had an introduction from the prince of Guabi, gave us a friendly reception.

Next day (17th June) we watched the rain from our rooms, which had been papered with sheets from an English newspaper — chiefly pictorial advertisements — and dried our pressed plants over a fire in the covered gallery outside the house, while a large brown pool in the yard threatened to flood into it On 18th June the sun came out again and we went out on Lake Lugu Hu in dugout canoes. Set among forested hills, its waters were deep blue, and [p.30:] the stones and even the bubbles coming up from the bottom shared the same hue. Festoons of algae swathed the submerged limestone, and my task was to collect them for preserving in formalin, and to fish for plankton. Floating on the water in large masses were the handsome white flowers of the new Bootia crispa, each on a slender stalk, and when we explored the boggy meadows we found much of interest From the meadow itself, where the ground was fairly firm — although my horse sank up to the knees before I realised what was happening — down to the water's edge there were three distinct zones: first a ring of Iris fomestii with Anemone rivularis, an Artemisia, Erigeton praecox, Potentilla griffithii, Poterium filiforme, a Festuca, a little violet and our own bird's-foot trefoil; then on the black soil masses of silverweed, shepherd's purse, plantains, three small knotweeds (Polygonuni), sedges, Lysimachia parvifolia, Gnaphalium multiceps, Salvia plebeia, a little gentian and Pedicularis densispica — a plant community much the same as that on the banks between the ricefields; and lastly, on the wet mud, Rorippa islandica (R. palustris), Senecio oiyzetorum, Chenopodium spp. and Bothriospemum tenellum. Growing in the dykes were Aconis calamus, marsh marigold, water crowfoot, arrowhead, water plantain and Hippuris. At the northern side of the lake sandy clay-slate, alternating with bands of light brown marl and shale, dipped gently northwards, though interrupted by limestone reefs, and on Shizu Shan — a mountain of impressive shape — overlain by massive beds of limestone. Next day we took a track which at first went up past some Xifan villages on to the mountain; then it curved off to the left over a little saddle into a valley with a dry streambed, leading gradually downhill into the Yongning basin. The boundary between Sichuan and Yunnan ran through the lake, and we were now in the latter. Low spurs projected into the basin, and abundant springs rose at their extremities; here again we found boggy meadows with a rich flora, notably a new smooth-leaved dock, Rumex yungningensis. The local ruler, a friend of the tusi of Guabi, had sent two mounted men to meet us and we soon arrived at his village, a long line of large log-houses at the foot of the mountain, cut in two by the river which broke through from the south and was crossed by a large Chinese bridge. The paths were deep in mud, and it was another halfhour before we reached the large monastery, situated just on the far border of the plain.

We were received by the abbot of Yongning, a man of giant stature dressed in the red habit of a monk, his enormous skull smooth shaven (Fig.6). He was of Naxi origin but completely Tibetanized. His office was hereditary; two of his sons were being trained as lamas. The monastery was built in the form of a quadrangle with the temple in the middle. In front of it was a courtyard with beautifully decorated stone fountains, but the massive masonry of the temple and its golden pinnacle towered above all the rest. Its huge door was concealed by curtains, and dim light penetrated into the interior through a few small windows high up and through the gaps in the brushwood which filled the space beneath the roof. The great image of Buddha in the middle was new and there was nothing else of much value. The lamas lived in cells in the outer buildings of the monastery, which were girdled by wooden galleries at first floor level. They were said to number some two hundred, but only twenty were present at the time, the rest being away on summer holiday, collecting money or on other errands. We were housed in the main building next to the abbot's quarters. The balcony in front of our rooms was adorned with a large wall painting of Lhasa, where the abbot had studied for ten years. The lamasery was situated at the foot of a steep, more or less freestanding hill some 550m in height, consisting of sandstone with limestone outcrops. Its name was He'er and I climbed it next day. Gear weather enabled me to survey the entire surroundings. Everything was green: the meadows and the fields of oats and barley down in the basin at 2725m, through which the silver thread of the river wound in broad meanders; the heath pastures, an open plant community with few grasses but rich in sedges and low growing herbs, which covered Mount He'er and all the low rounded tongues or spurs projecting into the plain on its distant side; and above them the forests, in which the fir zone stood out sharply as a blackish line at the very top. There were a few cliffs and peaks of pale grey limestone, notably the south-facing precipice of Shizi Shan to the east and the three rounded summits of Mount Waha towering above the forest to the south. To the north east a dark chain of pointed spires separated the whole of the sunken woodland area from the Litang river valley, in which Muli lay, and to the north, gleaming under fresh snow, was Gonshiga (4900m), a steep summit beyond the forests, on which I was destined to set foot next year. The wooden huts of the Naxi and Xifan villages blended inconspicuously into their surroundings at the edges of the plain. At midday, after my return, I was suddenly attacked by dysentery with fever and faintness, which prevented me from climbing Mount Waha but which soon disappeared without any after-effects. The Naxi [note #28:The Chinese call them Moso, a name which is fainltly contemptuous and is disliked by the tribal people (H-M's note).] people of this district constituted a separate tribe and called themselves Liidi. Their language was closer to that spoken in Guabi than that of Lijiang, and their understanding of the latter was far from perfect. Besides the abbot they had a temporal head who resided at the other end of the village, and still further away there was a Chinese official with a few soldiers, who was said to find life here not particularly agreeable, though he was to all appearances on good terms with the lama. We visited them both. I measured another baseline on the plain for my map. Schneider bartered a rifle, a clock and some other items with the lama in exchange for ethnological objects, while I was given a little flat-nosed dog for the promise of an Autochrome photograph. On 23rd June we set off once more, to journey along largely unknown roads via Yongsheng to Lijiang. [p.31:]

The track ran southwards on to a low elevation to the east, bypassing the river valley which entered the plain through a little curving gorge. The vegetation was now hastening with giant strides towards its climax, or, to put it more correctly, in attuning their climax the tallest plants overwhelmed and hid their shorter neighbours in a continuous display. The whole forest was lit up by the low trees of Rhododendron decorum, covered with heads of large beautifully shaped white flowers (Fig.7). The low growing anemones had withered and were now overshadowed by taller herbaceous plants 20cm or more in height, including a tall dark red sage (Salvia castanea), the scabious relative Morina delavayi with prickly hairs on its leaves and carmine-red flowers several centimetres long, several species of lady's slipper, notably the dull purple Cypripedilum himalaicum with flowers almost the size of a man's fisi, and Leontopodium dedekensii In little hollows and dolines and between the bushes tall-growing herbaceous plants were shooting up and forming a third tier or storey, among them several species of Ugularia with large ovate or kidney-shaped leaves arol long panicles or racemes of golden yellow flowers (here chiefly L, veZfea) and our native Chamaenerion angustifolium When we reached the river once more it was splashing down tiny waterfalls over outcrops of almost horizontally bedded limestone into blue pools in a delightful forest valley flanked by a low range to the east and the much higher slopes of the Waha chain to the west Before long we came upon some boggy meadows separated from the river by a line of sea-buckthorn trees (Hippophae) and Euonymus lichiangensis with winged stems. The meadow flora was superb: red primulas and louseworts, the yellow Iris fonestii, the blue /. bulleyana and Strobilanthes versicolor, Euphorbia nematocypha, large and bushy, white anemones and Aletris. In these delightful surroundings, situated as picturesquely as a village in the Alps, was Mudijin, the Xifan village where we were to spend the night The valley then narrowed into a gorge filled with forest, bamboo and magnificent flowers. Flourishing mainly under the bamboos was the gorgeous Nomocharis mairei, its flat white flowers flecked with purple and its inner petals delicately toothed. But now in the rainy season we encountered some of the annoyances of this zone: tiny biting flies (! Simuliuni) and a species of leech with longitudinal brown and green stripes which lay in wait on herbaceous plants or bushes, stretched itself out, as thin as a hair and as long as a man's finger, towards the passer-by, and in an instant attached itself to one's clothing or the horse in its quest for blood. Its bite was quite painless and was only just perceptible if one was watching it closely. Yet the wound went on bleeding for almost half an hour, so that everything was smeared with blood — indeed, that was the most revolting part of the business. That afternoon we emerged from the Yongning valley system over a little col at 3450m clothed with birch trees with an undergrowth of willow. From here we had a fine view of the southern summit of the western range, a summit which we had previously identified from afar and which Gervais-Courtellemont had christened "Pic Le Guilcher", though quite unnecessarily, as its name — as was reported to me from the other side — was Alo [note #29: Alo is Mianmian Shan, north of Yongsheng and northwest of Langshu Zhuo.] , or in Chinese Yue Shan. During the gradual descent, peering between the pines, we had our first glimpse of the snows and glaciers of Mount Yulong Shan (Yulong Xue Shan). Emerging for a few moments between the clouds, still far away beyond the unfathomable depths of the Yangzi gorge, it seemed to greet us from on high. Threading our way between deep funnels in the karst, we arrived that evening at Piyi, situated in a basin which drains eastwards into the main arm of the Woluo He. Our direction was still in the main south-south-east, and we went on across another fair-sized basin and past two side valleys higher up. Towards evening we reached the Woluo He itself, below a large village called Baodu, running north-east down a narrow valley. All round there were villages where Chinese and Naxi lived side by side. The former must have migrated from far away, for they had built large bucket-wheels in the river, of a kind which I did not see again until my later travels in Hunan, Ninglang, where a district official was stationed, was located to the south of Baodu near the end of an elongated basin where rice was grown and coal was mined high up in the mountains on its eastern side. The river narrowed again and made a seemingly paradoxical bend, the angle of which received a substantial tributary from the east We soon came to the river again and went up its east bank to Xinyingpan, where a nephew of the tusi of Guabi was the ruler.

We could not get much sense out of the fellow, who was obviously feeble-minded, but we stayed with him for two days, using the first (27th June) to climb the mountain above us. The second day we were kept indoors by wretched weather — hail mixed with rain, overmuch of which found its way into my lodging in the temple. A track led past a small lamasery below Hongguwo summit, which seemed from below to offer the best vantage point, and then over the ridge accompanying the valley to the east Nowhere was there any trace of a mountain path branching off from it, but somehow I had to reach the summit I therefore climbed boldly into the long grass, leading my horse and threading my way between fallen tree trunks. When I emerged into the open again and mounted my horse, I was horrified to find that the pouch containing my aneroid was missing. However, my coolie had picked it up, though without saying a word, and now handed it back to me. Fortunately the instrument was undamaged and even its setting was unaffected. The summit at 3475m was covered with pine scrub and proved totally unsuited for my purpose as it offered no view to the east All that I could see were two high mountains near Yanyuan, and I used them to fix its position. As there was no time to climb another summit, I had to be content with the western half of the panorama. A few dark rocky spires belonging to Mount Yulong Shan had been visible for a short time from much lower down, but now the whole mountain was shrouded with cloud, and as far as Alo all that could be seen was mountain landscape of the same appearance. It was somewhat [p.32:] lower than the summit where we stood, and consisted of sloping plateaus, rounded ridges and here and there a level triangle, all built of more or less horizontal alternating strata of limestone and sandstone, with a filling of multicoloured marls in a few depressions. This territory, from Da'erbi to Yongning in the west, and as later became apparent, far beyond Jinsha Jiang to the south, is, superficially at least, the geologically least disturbed part of the Yunnan highlands. That night there was some shouting caused by the capture of a thief who had entered the yamen and taken a few harnesses and other items from our boxes. Next morning another was caught in the act of making off with Schneider's raincoat, so that in the end we had to insist that the "prince" should give us a written undertaking to assume responsibility for everything which might be stolen from us by his "law-abiding" people. The plant life had nothing special to offer apart from some fine trees of Comus capitate, now in full bloom, a mass of sulphur yellow, this effect being produced by the four bracts which surround each of the tiny flowers. We also found a tall slender overhanging shrub with pink flowers — Campyhtropis polyantha.

We left Xinyingpan without regret, though not before I had measured another baseline. We followed the main east branch of the valley up to its origin and found that the simple picture of the route given by Jack [note #32: Jack, R. Logan, The Back Blocks of China, London, 1904.] was totally incorrect The source of the river Woluo He is situated here, just above 27° of latitude. Crossing a col at 3075m, we came to a tributary flowing southwards. It followed the same Hongguwo range until it joined a little river which crossed the forest-clad sandstone range from east to west, though its valley was not deeply incised. Running along a stream, the broad ancient cobbled roadway, now completely ruined, led upwards, overgrown by oaks thickly covered with Drynaria. The mottled begonia-like leaves of Parasenecio fotrestii flattened themselves against the mossy forest floor and Saxifraga chinensis was opening its flowers with their long deflexed tips. It was pitch dark when we reached our lodgings in Boluoti, the most southerly Xifan village. It lay beyond a second river, parallel to the other, which seemed to drain separately into the Jinsha Jiang. On 30th June, after our guide had slunk off into the forest in the rain, we at last made our way there, travelling south-westwards up and down across three side valleys running down to the river, and deviating round a higher group of peaks. From a crest we at last had a clear view far towards the south and west The Yongshang basin lay 500m below us; it was really a broad shelf bounded on the far side by a slightly raised edge. Crossing it were a few deeply incised streams which ultimately united and cut their way through the marginal ridge, before running down into the still larger Sanchaun Ba 550m lower down. Today there were again changes in the flora. Antiotrema dunnianum, a member of a new genus of Boraginaceae, grew in the undergrowth of the pine forests, and the splendid new Trigonotis heliotropifolia beside a brook. Much lower down we met some woody climbers: the dull purple Paederia tomentosa and Millettia dielsiana. Once again we made a long day's march, spending the night at Yongsheng on the further edge of the basin, at 2300m.

Yongsheng was a large town and a busy trade centre. We decided to take a rest day to catch up with the task of drying the pressed plants, a duty which had been somewhat neglected in the last two days. In the event it was done almost too thoroughly. Several hours after nightfall I noticed a pungent reek, more powerful than the ordinary smells of a Chinese lodging. I went to investigate and in the next room I found a stack of drying paper half a metre high, with the whole of one side smouldering; a coolie had set it on fire with his pipe. Pere Guil-baud, a French missionary who like so many of his confreres was an ardent collector of beetles, paid us a friendly visit. However, we could not afford a prolonged stay and on 2nd July we set out again in the rain, first downhill in wide loops beside the gorge of the stream with its red limestone (presumably ferruginous) to the densely populated Sanshuan Ba. Its stream debouched into the Doluoti river but we did not follow it any further. Instead, the track continued in a westerly direction and ascended into a small valley, finally leading down to the Jinsha Jian beyond the Naxi village of Dawan, a day's journey from Yongsheng. From the highest col we had a magnificent view of the snow peaks of the Lijiang range, but not until we had descended some way, to a level where the subtropical vegetation had now assumed a green colour, did we emerge from the mist again. Phyllanthus emblica was ripening its apple-shaped yellow-green fruits, which the Chinese eat despite their acidity. Here I recorded the highest temperature reading of the year (31.5°), and it was also the lowest altitude that we had reached for some time. At water level under the iron chainbridge at Zi Lijiang it was only 1440m. Sixteen chains side-by-side, with two to serve as guard rails, stretched between two rock platforms 20m high, spanning the Jinsha Jian where it narrowed to 80m. The planks were very insecurely fastened and the whole bridge swung so violently that not more than two horses were allowed on it at a time, a rule which the watchmen on the tall towers above the gateways at each end of the bridge enforced by signalling. On the far side the miserably ill-paved track climbed up again. A brook plunged out of a ravine in several waterfalls over red limestone rocks. Down the valley towards the south the walls of the river gorge continued almost vertically as far as one could see. Leading up a steep stairway on a rock face, the track offered magnificent vistas. We had to engage several extra men to help with the loads, but our ponies were unbelievably surefooted and not once did we have to dismount. One thousand metres above the river the hillside became less steep and we soon arrived at the village of Duinaoke. On 4th July we ascended gradually through woodland to the pass at 3125m and then down on steep cobblestones into the Lijiang basin. It was raining yet again, and although the hedges were gay with fragrant honeysuckle, abundant white roses and tall golden yellow St John's wort, the dirty town seemed to offer an unfriendly reception.

[chapter 7:]