Heinrich Handel-Mazzetti: [A Botanical Pioneer In South West China] - chapter 7

[p.33:]

Chapter 7. Lijiang and Mount Yulong Shan

Vicissitudes—the Naxi village of Nguluke — geology and summer flowers of the mountain — weather conditions — heaths and meadows of the glacial lake basin and the glacier stream Baishu — splendid primulas in the marshy meadows of Ganhaizi

Lijiang lies in a picturesque site at 2475m at the foot of two hills forming the end of a mountain range which projects from the north into the broad basin of a former lake. The outermost hill is quite low and is cut off by a stream diverted through the town; on it stands a glittering temple between tall cypresses. The basin is bordered by mountain chains running north and south. To the south west a lateral ridge bnnches off from a pointed peak, Weibi Shan, making a transverse barrier across the basin and forcing the western stream, which forms a lake there in times of flood, to swing round into the eastern part, wiere it joins the stream just mentioned and another coming from the north east, and turns towards the south. "To fhe north we could at Tirst see nothing but black clouds, through which we now and then caught glimpses of rock slabs and an icy summit, so high above us that from the lower parts of the town it towered far above the quite considerable northern hill. Not until some days had gone by did we see the "Pik" — in Chinese Yulong Shan (Precious Dragon Mountain) [note #31: Usually rendered as Jade Dragon Mountain.] — in its full splendour, and then only in the early mornings. It is chiefly this mountain that has made Yunnan famous among botanists and gardeners. First visited by the French missionary Delavay in 1884 and subsequently explored by Forrest, whose expeditions have extended over many seasons, it has yielded hundreds of species new to science and has provided some magnificent high alpine plants for our gardens. I did not intend to stay there long, but I wished to get a general idea of the flora and in particular to collect the cryptogams, which had never been studied at all; Schneider wanted to spend the midsummer weeks there.

First, though, we had to remain in the town for ten days, and endure various annoyances. In response to a letter sent in advance, the district official had procured accommodation for us in a fair-sized private house, but two days later we had to move into another as our coolies were too many for the proprietor. The population of Lijiang consists chiefly of Naxi (Na-khi), who are the local tribes of Moso [note #32: Moso are now regarded as a local tribe of Naxi. Most of the Naxi live in or around Lijiang (S.G.H.).], together with Chinese and Minjia, and also numerous representatives of other mountain peoples residing in the town. At the same time as our arrival the Austrian explorer A.K. Gebauer [note #33: Having set out from Burma on 12 January 1914, Anton K. Gebauer travelled via Tengyueh (Tengchong) up the river Schweli (Longchuan Jiang), crossed the divide to the Salween and went upstream to Tschenka. From there he crossed to the Mekong, reaching Weihsi on 27 March. All the passes were blocked by snow, but in the early summer he travelled to Atendze (Deqen), hoping to enter Tibet, out he was stopped by the Chinese. On the return journey he passed through the Liso village of Aiualo, later visited by Handel-Mazzetti, and crossed the Lenago pass to the Yangtze, arriving in Lijiang where he met Handel-Mazzetti. Leaving on 7 July, he returned via Tali to Tengyueh, where the British consul told him that war had been declared. After reentering Burma he was interned and spent the rest of the war in India, being repatriated in 1919. He brought back all his collections, including route surveys from Tschautou on the Schweli to Lijiang. Anzeiger der kaiserlichen Akad. in Wien, 1914, 51,101, 307 and 338. Anzeiger der Akad. der Wissenschaften in Wien, 1920, 57, 11.] entered Lijiang from the opposite direction. He had unfortunately been unable to cany out his plans, as the official in Deqen (Atuntze) — a man notorious for his hostility to foreigners — had prevented him from travelling to the Salween, and the official in Luchang had previously made it impossible for him to go any further upstream along that river. The information he gave me and his stroke of bad luck later became the inspiration for my most ambitious enterprise, a journey of which I still had no inkling at that time. We spent some pleasant hours together and replenished deficiencies in one another's kit Then, one dark evening, a telegram from Kunming brought news of the assassination at Sarajevo on 28th June 1914. I knew Archduke Ferdinand tolerably well, having once escorted him for a day and half [note #34: See biographical memoir, page ix.], and it was certainly a good thing that he never came to power, but the circumstances of his murder were such that we could not but fear for the consequences. I met Pere E. Monbeig, who was journeying to Kangding (Tatsien-lu) to make enquiries into the death of his brother, also well known as a botanist, who had recently been murdered on the road from there to Batang. He too gave me useful information and, two years afterwards, reminded me that I had predicted the outbreak of war to him at pur meeting. Li had a fight with a couple of soldiers over a woman and got badly beaten up.

We received valuable help from the Evangelical missionary A. Kok, a Dutch citizen who was working here with his wife and two deaconesses. During Kok's negotiations with the general, who finally gave orders that the two soldiers should receive a severe flogging, the latter mentioned that President Yuan Shikai had just imposed a ban on cartography by Europeans and he accordingly attached a soldier to our household as a spy. His presence naturally did not deter me from carrying on my surveying and map making; all it achieved was that I worked more surreptitiously than before and took great pains not to comply with the general's request to notify him of my departure in advance, [p.34:] made so that he could provide me with letters of recommendation to the district officials and supply an escort of soldiers. In subsequent years this ban was reimposed several times. Tang Rirao, the dujun (military governor) of Yunnan, gave special effort to its enforcement; the Chinese newspapers were often full of it, and officials were strictly forbidden to give any of the available maps to Europeans. Instead of being glad that we were doing the basic work for a survey of their country — the results of which we would certainly not have withheld — they forbade it. And what have they themselves achieved in the field of modern cartography? China is probably the only country which has never carried out a comprehensive survey. There are good large-scale plans of the immediate environs of some of the larger towns, including Kunming, in crude lithographic reproductions. All other maps are hopelessly distorted, being compiled partly from reports of district officials and partly from reprints of European maps, and as any notion of copyright is quite unknown in China there is no legal remedy. Wherever they have inserted additional material it is full of crass errors; for example in a new map published by the Yunnan militia — based on Davies' map — the blank area between Weixi and Jianchuan is filled in, but the river from Weixi is shown as if. it joined the Yangbi Jian — an impossibility sometimes depicted on our ancient maps, which make a river divide into two branches only to rejoin several hundred kilometres further on. It is therefore the duty of every serious explorer to put the matter right, provided he has the necessary skills and a little time to spare, whether the rulers like it or not. We were very glad to move on 15th July on Fomest's recommendation to the village where he lived, Nguluke or in Chinese Xue shan cun [note #35: Both names mean "Snow Mountain Village". At 2860m, it is the highest village on the slope above Lijiang.], 15 km away to the north-north-west at the foot of the mountain range.

So gentle was the upward trend of the plain, which consists chiefly of talus with patches of tufa here and there, and so enthralling was the ride along the gravel track, that it came as a surprise to find that the ascent from Lijiang to the highest parts of Nguluke was as much as 350m. The village consisted of substantial half-timbered farmsteads each built round a square farmyard, generally with living accommodation on one or two sides and stables and byres for horses, cattle, goats and sheep on the others. Above these were storerooms, topped with wooden racks for drying corn. The farms lay in three groups on the gently sloping talus fan which begins to iiscend not far beyond the large market village of Baisha. Forrest [note #36: George Forrest, 1873-1932, botanist and plant collector. Worked in Yunnan from 1904 until his death. J. Macqueen Cowan, The Journeys and Plant Introductions of George Forrest, Oxford, 1952.] had rented a house, and during our stay in the village he was most friendly and helpful, drawing on his deep fund of experience to give me much good advice for my future journeys. The spacious lofts, though often smoke-filled and — to me a more serious drawback — open to the rain, were always my favourite quarters in Naxi houses. The men dressed more or less in Chinese style, though they usually wore a felt hat and when out in the fields a thick felt weather-cape exactly like the shepherds in the Tyrol, and the women were clad in pleated skirts reaching down to the ground. They often had a small temple but in the village there was no real priest or medicine-man, though in many houses one found their books of magic spells written in strange hieroglyphic script [note #37: See Bacot, Les Mosos (Handel-Mazzetti's note).] . It was by no means obsolete and was still in use at that time; in 1915 Kok showed me an account for the construction of a house which a carpenter had written out for him in Naxi characters. The men were keen hunters; some of them had homemade matchlocks, but most of them used crossbows with arrows tipped with poison from the root tubers of various species of monkshood, notably the twining Aconitum delavayi. A little scratch was said to Idll a bear at once; the huntsman then ran up and cut out the flesh round the wound before the poison made the whole carcase uneatable. At that time the people of Nguluke consisted largely of "botanists"; since 1904 Forrest had employed them as collectors, znd many of them, owing to their intelligence, keenness, feeling for form and not least their knowledge of he mountains, had given outstanding service. Trey knew the various plants which grew there better tf an he did himself, so he told us. So it came about that every evening of my nine day stay in the village they thronged into my house, invited or uninvited, spread out their spoils in the yard and held what developed into a regular plant market

Yulong Shan rises at first gradually out of ihe plains. Wooded spurs project eastwards a f:w kilometres from the main ridge. Between them were old talus fans covered with heath vegetation, though above 3300m these merge into troughs filled with luxuriant meadows. Only to the south of Ngulute, where the height of the range is considerably less, does it plunge almost directly into the plain. At ihe foot of the mountain there is little solid rock exposed. What there is consists of limestone in which I later found fossils —poorly preserved cords of as yet undetermined age. This rock is almost everywhere covered with talus and conglomerate. In the alluvial cone of Nguluke there are substantid rounded blocks of volcanic stone of diabase tyjie, resembling lava in structure and containing inclusions of limestone with crinoids. These cone from high up the mountainside, a considerable put of which, including the crest itself, consists of cbrk diabase. To judge from the inclusions, this may have originated in a very ancient eruption; on the summit of Unliipe (4950m), which looks straight down on to the village, the diabase abuts in a sharp vertical lire against the pale limestone lying to the north. Mcst of the houses were built of this volcanic stone in various shapes and forms, commonly spheroid il masses of radial structure. To the right, nearer thin Unliipe, stands Hosayigo, a white trapezium risiiig steeply from its scree slopes to a peak at 4400m. To the north it continues as Saba, a rugged crown with countless notches and recesses, while the main ridje, a jagged crest climbing higher and higher curves in an arc towards the west. Although its exact [p.35:] relationships are not discernible from below, the main summit of Yulong Shan, a superb ice pyramid 5450m in height, known to Europeans as the Lijiang peak and to the Naxi as Satseto, is directly visible between Hosayigo and Unliipe. A gigantic rock pinnacle, Chaloko, projects outwards from it to the east in the same line as Hosayigo-sabe.

The very next morning I rode through steep pine forest up the path visible from the village. Before I had gone far I saw the elegantly shaped white flower heads of Pterocephalus hookeri and the red blossoms of Nardostachys grandiflora, a genus related to our valerian. Spread out beyond the steep crest between dark fir-woods was a meadow called Ndwolo at about 3500m, covered with an unbroken sheet of flowers, the most conspicuous being Strobilanthes versicolor, with large blossoms, usually blue in colour, not only at the top of its one metre stalks but in panicles on all parts of the plant With it was the single-flowered Trollius yunnanensis and, in drier places where the soil was not so deep, the tall Cimicifuga foetida and Veratrum stenophyllum together with the yellow Draba yunnanensis, fully comparable in size with our Biscutella, and others. Scattered here and there, their many stems sticking up like broom and their twigs spreading into an umbrella shape, were various shrubs, especially willows, lilacs, spiraeas (S.arcuata), Philadelphus delavayi and Berberis dictyophylla, their branches covered with dark grey mosses (Orthotrichum hookeri, Hypnum hamatum) and with grey and brown lichens (Parmelia cirrhata, Oropogon loxensis, Leptogium menziesii) in thick crusts and balls. But the tall vegetation harboured the first leeches and from there onwards, far up the mountainside, they lurked in their thousands to prey upon the unsuspecting wanderer. We went on up a steep slope through a strip of forest where lichens were even more abundant and came upon a rocky stream bed where the white heads of Juncus allioides dangled in the spray on their thread-like stalks, and the white flowers of Saxifraga chinensis, their petals elongated into ribbons, hung down in one-sided pairs from scapes dotted with red glands. We went up over stony turf past bamboo thickets beneath which a low-growing bramble (Rubus lutescens) had opened its large yellow flowers, accompanied by red and yellow Pedicularis in growing abundance. At 3700m there was another level spot, a boulder-filled corrie completely overgrown with lush sappy vegetation — tall yellow Ugularias, Salvia flava with yellow flowers spotted with violet, giant Potentillas, blue aconites and delphiniums, the pinkish-violet Megacarpaea delavayi and the mightiest of all louseworts, Pedicularis superba, its quadrangular stems as thick as a man's finger and nearly a metre in height, its leaf whorls fused into sac-like sheaths at their bases, and its numerous flowers, 4cm across, of a delicate translucent pink with a violet tinge like our Orchis purpurea. The last traces of the path faded out and at 4125m we passed the last few scattered firs, yet the higher we climbed the more glorious was die alpine flora. The steeply sloping meadows were ablaze with brightly coloured flowers vying with one another in size and splendour between clumps of grass, chiefly fescues, sedges and Kobresia species. There were Cremanthodiums (C. nobile and C. campanulatum), some with yellow flowers like those of Arnica and others brownish-black, the gland-studded Saxifraga nutans with racemes of yellow flowers and many others of its genus, primulas in enormous numbers, crucifers such as the graceful Dipoma iberideum in various hues, Potentillas, dainty Pamassias, the low growing Dracocephalum speciosum, the forget-me-not-like Microula hirsuta, innumerable orchids such as Orchis spathulata and the vanilla-scented Gymnadenia delavayi, the yellow-flowered Cyananthus macrocalyx, edelweiss (Leontopodium calocephalum) with wide flat stars, but yet again the finest of all were the red louseworts with their fantastically shaped flowers — Pedicularis delavayi with long narrow corolla tubes and P. elwesii with large heads — nearly 10 cm across — of densely packed blooms on almost leafless stems. Long white screes, never entirely at rest, extended down the mountainside and they too had. their own peculiar vegetation. Pleurosperum foetens and Trachydium hispidum, umbellifers with huge taproots, pressed their elegant rosettes of leaves tightly against the stones, together with Meconopsis rudis carrying large nodding blue flowers on tall stems covered with prickly bristles; Eriophyton wallichianum, looking like a deadnettle clothed in white wool; several species of larkspur, yellow, blue and violet, and Hemilophia pulchella, a small crucifer with pink and white flowers. Anemone wpicola, the undersides of its leaves suffused with violet, pushed its thread-like rootstocks up between the stones to unfold in loose mats on the surface. At 4250m, in depressions where the snow had lingered, we encountered yet more primulas of the utmost beauty: P. pseudosikkimensis with large pendent sulphur-yellow bells and P. pinnatifida with compact spikes of small blue flowers. With them were a low-growing dark violet sage (Salvia evansiana), the tall pink Allium victorialis and others. All these alpine flowers gave off marvellous perfumes — a scent of honey from the blossoms and a spicy odour from the glandular hairs. The perfume from the dried specimens still pervades the herbarium parcels ten years afterwards. Our own alpine flora cannot match the splendours of this range and my later travels in Yurnian convinced me that Yulong Shan is the nonpareil, the finest of all floral mountains, unrivalled in its glory both scenic and botanic. I cannot imagine anything which could possibly excel it The work of Delavay, Forrest and Schneider (I do not include my own, as mine was only a short exploratory visit without any attempt at thorough study) has already yielded some five thousand species of flowering plants from its highest slopes down to the river at its foot and these, coming from an area only fifty kilometres square, comprise almost as many species as the flora of the entire Balkan peninsula. On that day, however, my enjoyment was rudely interrupted; it began to hail and then to pour. The rain ran in at my neck and out at my boots; that was only the first of many days on which I got wet through.

But even when the sky was blue, the village basked in warm sunshine and atmospheric humidity fell to 35 per cent, Jade Dragon Mountain was seldom without its cloud cap. Nearly every day, seemingly from nowhere, a cloud would gather round the peak and a violent thunderstorm would ensue; Forrest once counted sixty lightning flashes in [p.36:] a minute. This heavy precipitation, invariably falling as snow, produces intense glaciation, and together with the steepness and instability of the slopes this is the reason why the treeline and the high alpine floral zone are at far lower altitudes here than in most other mountain ranges of Yunnan and southwest Sichuan, which though similar in situation are less isolated and do not have such lofty peaks. The unforgettable riches of this high alpine flora, brought together and preserved on this range in a concentrated or "potentiated" assembly, are presumably explained by the moistness of the mountain climate in conjunction with the geological diversity of the rocks and the open, free-draining soils. In the rainy season low cloud and drizzle are frequent in all parts of Yunnan, and the Lijiang range does not shelter any of the surrounding country — with the exception of the depths of the Yangzi canyon — from mis precipitation. Even on freely permeable gravel soils it is enough to allow the local equivalent of heath vegetation to flourish luxuriantly. Such vegetation is seen at its finest along the track leading northwards at the eastern extremity of the range towards Baishui (Whitewater).

On the 18th July I rode out in that direction. At the margin of the former lake bed, which continues northwards for 5 km from Nguluke, thickets of holly-leaved oak were crowded between blocks of conglomerate left by the receding glacier. At that time of year there were relatively few plants in flower on the stony soil, but in spring there would be Incarvillea grandiflora and in autumn a narrow-leaved sky blue bellflower, Cyananthus argenteus, lying prostrate on the ground. At the edge of the forest we came upon the caravan route from Lijiang to Dagu and Yongning, and soon reached the chasm with its underground watercourse which drained the glacier in ice age times and which now emerges at the plain and flows along its eastern boundary to the town. Coniferous woodland predominated, with much spruce and Pinus tabulaeformis, though with an undergrowth of holly-leaved oak in many places. In shady spots there were dense stands of luxuriant perennials up to a metre in height, including Strobilanthes versicolor, Rodgersia pinnate and a tall meadow-rue, Thalictrum delavayi, carrying broad panicles of countless mauve flowers with quite large petals. Old moraines, now tree-covered, ran down from our left, and the trail climbed over them to its highest level (3110m). Beyond it stretched a beautiful green meadow with sinuous indented margins; just above it began the forest, its lower boundary defined by a precisely horizontal line. The subsoil consisted of pebbles cemented together into conglomerate and it was obvious at first sight that the meadow had once been a lake. Indeed, its Chinese name [note #38: The Naxi name is Gaba.] Ganhaizi (dry lake) seems almost to imply some geological insight, though it is probably derived from the fact that it sometimes fills when the snow melts. Owing to the permeability of the soil it has remained treeless, and in some places it was so dry that only a few herbaceous species flourished, notably Anaphalis chlamydophylla, which looks like one of our everlastings and covered large stretches with its straw-textured silver-white flowers, and Leontopodium dedekensii Everywhere else, however, there were flowers of all colours, even among the heathy undergrowth in the surrounding pinewoods. Among the most striking were the dainty little Gueldenstaedtia yunnanensis with sky blue flowers, the pink-flowered Triplostegia glanduiifera, related to our valerian, the yellow Viola delavayi, the white Androsace erecta, Drosera peltate, an insectivorous plant with indented halfmoon-shaped leaves covered with red glands, Swertias in vjirious colours, the prostrate creeping Lysimachia congestiflora with yellow flowers, the tall Pediaularis polyphylloides, a coarsely hairy plant with creeping purple-red stems extending widely, each stem bearing rows of red flowers, the erect P. integrifolia with entire leaves and the taller, stiff, much branched P. tenuisecta with finely cut leaves; the tall herbaceous Scutellaria likiangensis, salvias, moiinas, asters, various-species of Saussurea, resembling our knapweed but with leaves like those of our silverweed covering the ground and long stiffly erect flower spikes; Spencena ramalana, a plant as beautiful as it is interesting; the tall bristly blue Onosma paniculatum, Roscoea cautleioides with orchid-like flowers in shades of violet, pink and yellow and various true orchids including Lady's slippers and Habenaria glaucifolia with fleshy liaves tightly pressed against the ground and liarge, curiously shaped green flowers with rolled sepal tips; and also forming rosettes, but in this instance up to half a metre in diameter, the composites Jurinea forrestii and /. berardioidea with sessile flowerheads, the first having several in a cluster and the second having individual flowerheads between the leaves. Slender sedges and rushes heightened the similarity to a meadow, among them the lax-flowered Cyperis sieberianus and the narrow-leaved Fimbmtylis diphylla, together with Juncus allioides with its white flower heads. Satseto emerged from a backdrop of dramatically sunlit clouds. A sharp crest presented its eastern flank towards us, its upper rocks plastered with snow and fluted with snowchutes, its lower part an ice slope, and then swung round to the right to end in a monk's cowl dome. A colossal randkluft (marginal cleft) ran along its foot, and the neve basin which fed the glacier was barely visible from below, but the icefall of the glacier's sinout hung down the slope as far as the highest trees. It looked as if it might be possible to climb the ice wall from the glacier on to the crest, but I do not know in what season the weather conditions might enable even a first rate mountaineer to attempt it. Nothing disturbed the majestic calm of this lonely uninhabited place; the only sound was the murmuring of the glacier stream heard from afur as it plunged over the precipice and hurried to join the Yangzi, running through a gorge which it had cut to a depth of 125m not far ahead of us. The path led down through the woods to a stone bridge. The milk white brook flowed over its bed of sand and gravel between dense thickets of willow, much of it covered by moss, while birches and various coniifers including larches clothed the slopes which descended steeply down to the valley floor, rather narrow but perfectly level. In this spot, untouched by man, beneath the towering crags and snowy heights of the mountain range, every step revealed a new aspect, each more magnificent than the last [p.37:]

Two days later I went up the mountain again, at first takirg the same path as on the first occasion, but then straight up the steep grassy slopes to the highest corrie below the Unliipe crest. Once more I revelled n the glories of the alpine flora, but yet again my enjoyment was rudely cut short. It began to rain ard then to pour. My porters thought it was unsafe to continue on die slippery turf in their straw sandals and in thick mist without a guide it seemed unwise tc venture further in unreconnoitred terrain. A little further down I therefore took a steep track leading southwards, so as to reach the volcanic rock which forms the southern part of the range. Here too there was a rich flora, including clumps of Uoydia oxycarpa with yellow flowers like lilies, the white Allium yunnanense, the dusky blue Codonopsis bulkyana and many others. The bright yellow lichen Acarospota discumns (a new species) with its radiating storage branches made a splendid contrast against the black rock. I climbed up over a little col and down the next gully. I had sent back my leather gaiters with the horse so as to make climbing easier, but my waterproof cape brushed against the dripping leaves ard was soon full of leeches. When I got home I found eight of them on my legs. They had wormed their way through my stockings and, distended with blood, had been unable to get out again.

On 22nd July I at last set out in search of the plant which in recent times has perhaps attracted more attention than any other flower from Lijiang, namely Primula littoniana, though as later transpired it had a prior name — P. vialii — by which it must now be known (Fig.8). Here it grows only in a few spots — in a bog on the range to the east opposite Nguluke and at the second "dried up lake" near Ganhaizi, on the direct route from Lijiang to Zhongdiai. This last place was reached on the downhill slope facing south-west, where the small Primula yunnanensis with one to three flowers grew in masses by the wayside, especially on sandstone, together with the low-growing Rhododendron sinolepidotum, its large deep-red flowers contrasting vividly against the white limestone. Lying in a boggy maadow in a depression on the ridge was a clear pond called Haleko. Beyond it, 200 m further down, the path led into the large depression of Ganhaizi. Not far off was a Naxi village divided into three groups of houses. Even on the mountainside, just as in the immediate neighbourhood of Nguluke, the spring-fed bogs were gay with flowers. Among them it was surprising to see the blue Cynpglossum amoenum growing luxuriantly, the yellowish grass-of-Parnassus Parnassia wightiana, a dark Phlomis (P. atropurpurea) and a tall bog edelweiss (Leontopodlum calocephalum var. uliginosuni). The limestone conglomerate was adorned by Didissandra sericea with thick whitish woolly rosettes of deeply furrowed leaves and dainty nodding umbels of blue flowers with variable markings. But the highlight of the day was the boggy meadow itself, which lay at an altitude of 3130m and drained into a lake with no outlet. Our search did not take long: beside the little stream, which had cut through the black soil down to the gravel bottom, we soon found our primula — indeed its colours guided us to it. It is hard to say whether it is the most beautiful of all primulas, but it is certainly one of the most imposing. Its stout scape reaches a height of 70 cm, and is topped by a spike 15 cm long and 3.5 cm across. The upper part of the spike is coloured scarlet by the unopened calyces and the lower part is a cylinder covered by the violet-purple flowers, somewhat deflexed. There was no need to pick and choose the subject for a colour photograph. Wherever I set up the camera there was a gorgeous background — a rivulet, the meadow itself in all its hues, the narrow band of the lake, and the dark pinewoods, below which the flower meadows, so thickly covered with 'Strobilanthes that from a distance their colour was blue-grey, sloped down to the lake floor. A herd of shaggy black yaks grazing in the distance would have reminded even the nonbotanist that he was not in some pretty corner of our own Alps, but quite near Tibet. And yet, only two years later, when I came back from the Nu Jiang to Lijiang, the forests seemed to have dwindled to a shadow of their former selves.

On 25th July I returned to the town, determined to dry and despatch the vast amount of herbarium material which I had collected — indeed I did not greatly care if some of the packages got scorched — and to prepare for my next journey, again into little known country. Unfortunately the caravan leader who conveyed the tins to Kunming must have dropped one of them into the water, or perhaps the rain seeped in through a badly soldered joint Two years later, when I opened it, I found its contents spoilt and rotten, but still I was glad that most of the material came from the Lijiang area, which had already been thoroughly explored, though alas other dried plants collected near Wali and on the journey from Yanyuan to Lijiang were also spoilt.

[chapter 8:]