Heinrich Handel-Mazzetti: [A Botanical Pioneer In South West China] - chapter 17

[p.79:]

Chapter 17. Over the Zhongdian Uplands to the Mekong (Lancang Jiang)

Arid mountains and gorges — a detour via Anangu — scrub-covered tablelands — a vista of unknown territory — yak herds —anatural bridge —Zhongdian —with a Tibetan guard over the Nguka-la —die ferry over die Yangzi at Jizong — tinough unknown country to die Mekong (Lancang Jiang)

The Chinese proprietor of the ferry below Sanjia-1 ngkou at first demanded an outrageous price, but a jab in the ribs made him see reason and we crossed the Shuiluo He (Duo Qu) close to its confluence with the Yangzi, though the horses had to swim. Two men entered the water with inflated goatskins under their chests and helped those horses that were unaccustomed to swimming. Taking several at a time, they fastened their halters to a rope stretched between them and held their nostrils above water. All this took so long that having reached the far side we decided to take our midday break. This proved a lucky decision, for growing there was a twiner, Ipomoea cairica, scrambling up into the trees. As soon as the stem was cut the deep pink flowers, over 6 cm across, and the delicate palmate leaves began to droop in the blazing sun. However, I swiftly unpacked a bundle of paper and put specimens in the press before they had time to wilt I did the same with several species of Ceropegia, namely C profundonim, a new species with narrow leaves and large deep violet-brown flowers, the common C monticola and the small green-flowered C. muliensis, which I found here only. The liana Iltigera mollissima, which also grew here, belongs to an otherwise purely tropitcal family.

Our palth took us along a narrow, steep-sided hogsback riidge interposed between a stretch of the Jinshj Jiang running from west to east and a tributary, flowing parallel to it, coming from Apa-la near Zhonjidian ito join the Shuiluo He barely 3 km north of tk ferry.'The narrow track along its southern face had to be repaired in some places before the caravan could pass, and if anything had rolled down the hillsice it wrould have vanished for ever in the raging torrent. I climbed a little knoll on the ridge, near the saddle which carried the track to Chuazi, the last Pumi village, so as to take bearings on the principal features of tthe country around the bend in the Jinsha Jiang. Opposite the village, beyond the tributary, there was ai vertical crest of hard limestone running north and south, outcropping as a narrow wall between the: strata of soft slate which extended in the depths of the gorge. A much broader mass of limestone not fair to the west of it formed Mount Lama-cuo (3300mi on Handel-Mazzetti's map), which was the continuation of the hogsback we had just tra-versec, though more than tour times higher. The erosiv; action of the tributary to the north of this mounain had scooped out a ravine with vertical walls. Next day our path led up its southern side, a climb of almost 1500m, at some points dizzyingly exposed. Though constructed with some care it was exceedingly narrow and the four roadmen sent ahead of us with heavy iron pickaxes were kept busy chipping away projecting rocks at the corners to make the path practicable for my not unduly bulky loads.

Growing on the rocks were some interesting xerophytes: Pertya phylicoides, a composite with needle-shaped leaves; two silver-leaved shrubs, Clematis delavayi and Desmodium cinerascens, Plectanthws oresbius, a little shrublet with a covering of ashen grey felt; Wikstroemia androsaemifolia, a robust plant with yellow flowers (a new species); the delicate Primula aromatica with elongated flowers, and near the top of the ridge a slender pyramidal cypress, Cupressus duclouxiana. The track crossed a col on-the crest, which was over 3300m high, and went down its south side. From there we had a striking view of the basin of the Jinsha Jiang as it flowed from the west-southwest, and beyond it the Xuechou massif. One can only picture the depths of the canyon through which the torrent surges, squeezed against the foundations of the steep mountain mass of Kudu. Far above the waters are the margins of the gorge, running side by side along the western border of the shallow basin and marking the line where the river has cut its channel through the rocks. Further on the basin rose up into the Xuechou Shan — Dier Shan range, while far below, deep in the bowels of the mountains, the brown waters struggled to escape from the last narrow ravine, at a point marked by a sharp crest with two summits. The huge scale of the landscape, coupled with the simplicity of its form, left a profound impression, although here it was very dry and barren, the sparseness of its covering of scrub being probably natural and not the outcome of man's destructiveness. Compass bearings and photographs enabled me to make a major correction to the course of the river as depicted on Ryder's map. The track now turned back into a valley occupied by Naxi tribesmen.

Here, and on the following day where it ran across more or less undisturbed strata of limestone and sandstone, I found some interesting plants, notably the orchid Pecteilis susannae, an erect plant with fleshy or waxy white flowers having a spur 15 cm long and resembling those of the commonly cultivated tropical orchid Angraecum sesquipeda. Another find was Dobinea delavayi [note # : Dobinea delavayi has certain remarkable features. Unlike most members of its family (Anacardiaceae) it is not a woody plant but a herbaceous perennial. The female flowers are carried in catkins resembling hops, while the white bracts of the male flowers are shed at a very early stage. The fruits are located in the middle of the bracts, or in other words the stalk of the fruit is fused with the bract. The thick woody rootstock, characteristic of a steppe plant, has an odour of pine resin. (Handel-Mazzetti's notes, originally part of the text).], a herbaceous plant a metre tall belonging to the Anacardiaceae. It was not uncommon in habitats favoured by the orchid. The people collected the leaves of Gerbera nivea, which are rather less than 20 cm long, sepa- [p.80:] rated the leaf stalk and central vein and used them to make fabrics.

Bandits were said to frequent the Apa-la, and the lama at Yongning had advised me to send men forward from the last village to reconnoitre and to take an armed escort with me, as the Chinese soldiers, if they encountered bandits, would promptly throw away their weapons and run off, or might even join forces with them against the master they had been ordered to guard. Partly for this reason, but chiefly because the route over the Apa-la had previously been surveyed by Ryder though there was still some unexplored territory between it and the area I had mapped in 1914, I branched off to the left from Laba and travelled southwards up the longest branch of the valley, climbing high up on to its western side and passing the Yi hamlet of Huajiaoping. Towering up on the opposite side, the arid limestone massif of Kudu presented its back to us, here known as Aka-e'ndyo. Besides a boy who came with us from Laba, we had as our guide a Tibetan trader who dealt in drugs and medicines and who had taken the same route as ourselves from Muli, having failed in his attempt to get to Kangding (Tatsienlu). The narrow track led down to the stream and then steeply uphill beyond it. Lao Li's pony, as feeble a specimen of horseflesh as his rider was of mankind, started to flag, and suddenly, with a scream from Lao Li, both of them disappeared over the side of the track behind me. Lao Li landed on his feet a few metres down the slope, but the pony tumbled down to the soft marshy valley bottom, damage being confined to the shattered saddle. Lao Li had to walk all the way to our campsite at the next spring; though this was irksome the exercise was no doubt beneficial.

After a short ascent the route led along the hillside and then across the flat Gitiidii pass (Mach-angba in Chinese) at 3425m. The meadows of the pass took an hour to traverse and displayed the superb flora which I had seen on the Zhongdian plateau in 1914. Here and there were marshy patches with willows. As we went down into the Bapaji valley a splendid vista unfolded through an opening in the pine forest; far away to the south, but perfectly clear, was Mount Haba Shan, a soaring snowcapped dome looking like the Order seen from the north, with the range running southwards from Bede to join it, including the peak which I had climbed in June; closer to me was Shusuzu, an old friend from the previous year, and the reddish sandstone mountains between it and the Piepen summits. Beyond the river Bapaji at the foot of this range, a narrow strip of pine wood marked the edge of the river gorge; from there it was only a short day's march to the terraces at Bede. I crossed the Bapaji, turned to the right, climbed to the edge of the gorge and arrived at Anangu (Ananchang in Chinese), a flourishing mining town of some size. Though built in Tibetan style, its mixture of peoples included Chinese, among whom were several officials, and the chief of police kindly invited me to tea. He gave me two policemen for the journey to Zhongdian and since thieves were numerous in Anangu — as must always be expected in industrial towns — he sent another for that night. This man performed his duties as watchman by lying down outside my door and snoring with all his might Indeed, he did not awake until got up, took a long strap and endeavoured to chaise I out a cock which had been hiding just above my bed | and long before daybreak had put an end to i slumbers.

Next morning (16th August) I set off northwestwards from Anangu, taking the main route towards Zhongdian, the district centre. High up on the hillside to the left were the spoil tips from (the silver and gold mines which had attracted the Chinese into these mountains. Soon we descended again to the stream, which flowed beneath a dense co'ver of large bushes of buckthorn (Rhamnus virgatus) :and spindle-tree (Euonymus lichiangensis), the laltter having four-angled twigs with longitudinal winigs, and very narrow leaves and seedpod valves. This stream, the main affluent of the Bapaji, came from Piepen to the west-northwest. The track left it and climbed up the left side of the valley. The sandstone banks on either side of the track were coated with green — the sporelings of various small hairmoss«s. We came out on to a broad ridge clothed with patches of woodland interspersed by flower-fillled pastures. Streams ran down on three sides — towairds Bede, Laba and Zhongdian. The pass was called Patii-La (3925m); though it did not offer much iof a view I was able to get my bearings without difficulty. After descending for 375m we came Ito a branch of the Zhongjiang He and halted for our midday rest at a spot where the river broadened a little. This branch ran due north from Piepen and encircled the whole mountain. While exploring a little bog nearby I noticed for the first time thiat a tiny plant growing on the bare mud — a plant wlhich I had seen once or twice before but had dismissal as nothing more than a dense clump of seedlings — actually bore narrow fruitlets under its tufts; of spathulate leaves with forward-pointing teeth, and these fruitlets arose from still tinier green flowers. Not until long afterwards did I realise that it was Circaeaster agrestis, one of the few flowering pllants whose affinities and taxonomic status are still toitally obscure [note # 93: Later assigned to a family of its own: Circaeasteraceae.]. The second tributary of this branch of the Zhongjiang He flowed down an equally gentle gradient between the conies at the foot of the ssteep face of Piepen. From its valley — continuing in the same direction — the path climbed up to the main pass at Shulakaza, which lay on a broad plateau. A ridge extending from it to the right looked as if it might give a good view of the country. I climbed it and was not disappointed. The larches and oaks which grew there did not prevent me from talking some useful photographs. Looking southeast ailong the line of the Bapaji I had a clear view of Xuechou Shan. To the left and somewhat closer was the mighty bastion of Kudu, some 4700m high, its banded strata dipping slightly northeast To the right were the splendid peaks of Piepen, the other side of which had been the site of my final explorations in 1914. Its pale grey limestone pinnacles towered up from scree slopes and forests of dark green fir, sharply silhouetted against black thunderclouds, but [p.81:] before long they were enveloped by white mist. Towards the north the Piepen range diminished in height, its last brown sandstone ridges projecting only a little above the undulating plateau which extended — over 4100m in altitude — between us and Zhongdian. Just beyond the pass I counted eight little tarns, gleaming like mirrors, but otherwise most of the terrain was covered with brown holly-leaved oak scrub. Its predominance, taken in conjunction with the unusual assortment of plants in the meadows, irdicates that the Zhongdian district is really an outlier of the floristic region which borders on the high altitude deserts of Tibet. A similar area to the north of Kangding (Tatsienlu) is characterized by a band of holly-leaved oak extending over an altitudin-al zone of 700m. However, it was the view to the north, over unknown country towards Xiangcheng, which gripped my attention. It was an inextricable tangle of minor peaks and folds, apparently for the most part consisting of sandstone, only a little higher than the plateau itself. This plateau, almost totally unforested, was furrowed by a few valleys running from west to east, belonging to the Shuiluo He river system. The landscape, as I saw it on that day, conformed exactly to a description of the Yaragong district between Batang and Xiangcheng as portrayed to me some time later by one of the missionaries. A survey or photograph from another spot might have enabled me to construct some sort of map, but such an opportunity was not granted. To the left the continuation of this mountainous tract, which probably does not exceed 5000m, was hidden behind Lopipema, a mountain situated not far to the northeast of Zhongdian. To the right, as I was descending from my crest, Gongga, a snow-capped mountain of at least 5500m near Gongling, emerged from the clouds for just long enough to enable me to take a bearing; it seemed to lie to the west of and outside the high rocky range which I had seen from the crest south of Muli.

Zilu was a small grazing ground for yaks situated on a stream flowing in the fourth direction, i.e., north-westwards to the Zhongjian He. That evening numerous herds of yaks, sheep and goats, each several hundred strong, belonging to the Zhongdian lanas, ambled slowly past my camp over the emer-alc green meadows, brilliantly lit by the setting sun. Ariong them were numerous hybrids between yaks and domestic cattle, representing a complete range of intermediate types. Though female yaks are not particularly good-looking, yak bulls are often of splendid appearance. Some of them were used as pack animals. They were extremely nervous, and wren I encountered them they would often turn aside from the track in fright and clamber up or doivn the steep slopes, but they were thoroughly do:ile withal. The route continued straight on through evergreen oak scrub with sparse conifers ov;r low ridges, then through mixed forest and deciduous bush along the valleyside, gradually descending to the Zhongjiang He. The little river wound its way quietly over its gravel bed between the lush meadows and arable fields of the wide valley floor; further up it disappeared behind Mount Lcpipema, about 4500m in height It was a mountain of soft outlines and rounded contours; its brown sandstone soil was bare and exposed everywhere and its slopes had remarkably little forest cover — far less than would be expected in this district — probably because it supplied wood for Zhongdian and the numerous Tibetan villages. Their substantial, gable-roofed houses, distempered in red or brown, were scattered along the valley.

I reached the river at midday, at what proved to be a fascinating spot The valley narrowed or, more correctly, the river had carved out a small ravine in the left side of its broad channel. Qose to the river bank, welling up in a recess in a bed of stratified limestone, was a hot sulphur spring, accessible only by wading through the stream. Evidence of its previous activity was to be seen in the sinter deposits laid down like a flight of steps above the present-day outlet, where they formed a marsh with a thick growth of grasses and sedges and an abundance of orchids. Just downstream from the spring there was a natural bridge, about 20m in height and the same in width, beneath which the river flowed through an underground cave. The luxuriant vegetation growing above the spring owed its existence perhaps to the supply of sulphur and the abundant moisture, and possibly also _o to the frequent presence of man. The hot. water (43°Q flowed swiftly through a manmade basin of stone, and any attempt to fish for plankton would probably not have yielded much of a catch. I tried to photograph the natural bridge and the course of the river downstream. It made a sharp bend and then, leaving its original channel, cut through some outliers of the plateau from which we had just descended. However, when I had pulled the cover halfway out the whole plateholder came with it, and I had so few plates left that I was unwilling to use another. The streams running down from Lopipema united in a channel in the valley bottom, separated from the river by transverse limestone ridges. Dry heathy meadow vegetation clothed the elevated sections between the meanders of the stream, which drained into the Zhongdian lake. A peculiar louse-wort, Pedicularis tricolor (a new species) was flowering on them in abundance; it is a low-growing, mat-forming plant with long tubular flowers having large, pale yellow, white-edged corolla lips and a reddish beak. Ahead of us gleamed the golden domes above the white walls of the two large lama temples. It had been raining intermittently and the track was indescribably muddy. It bore off to the left, and two and a half hours later we were in Zhongdian.

Situated at 3400m, it was a chilly place. Its streets were filled with a mixture of natural mud with Chinese and Tibetan filth, but I was comfortably accommodated in a large room in what had once been a Tibetan temple. The official at Yong-ning had not sent any letter of recommendation; the police officer asked to see my passport, felt my baggage for weapons and withdrew, satisfied but somewhat embarrassed, after realising that the suspected rifles were merely tentpoles. One of the chief worries of the Chinese was that Europeans might bring weapons for the "Manzi", as indeed happens on a large scale via Tibet. The district official invited me to a meal at which he staged various little happenings with the aim of blackening [p.82:] Tibet and the Tibetans in my eyes. A Chinese burst in, greatly excited, and told us that two of his countrymen had been beaten up by Tibetans while they were on a mountain digging for roots — simply because the Tibetans wanted to keep the trade to themselves. He told me that the fish sold in Zho-ngdian — 1 had bought a large carp — was unfit to eat, since the Tibetans cut up their dead and threw them into the streams, or left them out for the vultures to consume. The Tibetans practised polyandry, he said, and smeared their bodies with butter, which was the reason why they stank.

"So far as we're concerned, your polygamy is just the same thing, and your people stink of garlic, which is much worse!" was my rejoinder.

"Are you perhaps thinking of going on to the Salween (Nu Jiang)?" asked the official.

Though the inhabitants detested him, he was certainly not hostile to Europeans and was most friendly towards me.

"That hadn't occurred to me", I replied, "only as far as the Mekong (Lancang Jiang)", well aware that I must not arouse even the slightest suspicion. There was a telegraph in Zhongdian; one message from him would be enough to wreck all my plans, as had happened to the unfortunate Gebauer in 1914. 1 learnt something of the situation in the world outside China, but news trickled through very slowly. I handed in my telegram at 9am. The telegraphist worked out the number of words and I paid for it. Then at 4pm he sent a chit with a message in English: "Please is it one words or two wds...?" and I finally received an answer four days later, though the distance was no greater than I could have travelled in four days at home. I had to go to the bank and draw money remitted by Kok, have my shoes mended and have a new pair of trousers made. All this took time. The ponies needed a complete rest. My grey had begun to limp, and had a sore on the shoulder where the harness had rubbed; this had been caused by the loss of the crupper, which had broken and been lost in a pool of mud.

I sent off my specimens in wooden crates, but one of them had no tin lining and the colour photographs taken at Muli and Shulakaza — remarkably successful considering the age of the plates — were spoilt by a green haze spreading in from the edges. All this kept me busy from 18th to 23rd August and because of wet weather I was able to make only one excursion — to the lake near the great lamasery. My guide or sponsor from Yongning took me there, though at first he misunderstood me and thought I wanted to visit the larger and more distant lake which Szechenyi [note # 94: Széchényi, Béla. Im fernen Osten — Reisen des Grafen Széchényi in Indien, Japan, China, Tibet und Birma, 1877-1880.] calls Omintsoka, and which receives the waters from the Zhongdian stream and its tributaries. The little lake was a miserable puddle, but on some limestone blocks nearby I collected Saxifraga candelabrum, a species with numerous glands, and in the heathy meadow I found Saussurea romuleifolia with grass-like leaves, Microula pustulosa and others; and on solid mud Aongstmemiopsis julacea, otherwise known only from Java. From there I sketched and took compass bearings on the few conspicuous landmarks which were to be seen on the low forested slopes on either side of the broad valley channel running south towards Xiao Zhongdian, so as to provide a view from the north of the terrain which I had photographed from the south in H914.1 was told that a band, some twenty strong, of the dreaded brigands from Qagcheng (Xiangcheng in Chinese) had been seen two hours distant from the town, and that all the one hundred soldiers of the Chinese garrison had been sent in pursuit of them, though they had gone in the wrong direction by mistake. The official therefore had no men to spare and as he had not much confidence in his soldiers he sent a message to the headman of the last villlage at the foot of the mountain along my route ordering him to supply a party of Tibetans to escort me: as far as Meti, I having insisted, despite his efforts to dissuade me, on sticking to my plans for travelling due west to the Yangzi and on to Weixi.

I set off on 24th August with two police soldiers and a picturesque troop of Tibetans — five armed with lances and swords and two with firearms. At first we travelled south along the route towards Xiao Zhongdian and Lijiang. It ran at the foot of a low limestone ridge in a valley occupied by a few villages but otherwise rather barren. Large tracts were tinted blue by Halenia elliptica, a gentian relative with spurred petals. The stream flowed towards me and on to the Zhongdian lake. Tlhen we turned off to the right and halted for our midday rest at the home village of my escort, a place called Beishaoge, where the men borrowed an ox to carry their kit. The western side of the valley wsas also extremely barren, consisting of a stratum <of arid limestone, steep but not very high, running in an almost straight line, with gaps from which ailluvial fans spread down towards the stream. Above it was a narrow shelf and above that the forest-clad ridge of the watershed between the Zhongdian depression and the Jinsha Jiang. Rising up from this, at the; same latitude as Zhongdian, was Zhere, a low domie with a truncated cone resting on its eastern side, reaching about 4500m. I thought I could see waterfalls on its slopes. When we reached Beishaoge the uppermost stratum had already faded out and our route climbed westwards across soft clay-slate. There was a. splendid view from a spot on the broad undulating ridge, which diverged from our route and ran southwards as far as a valley which cut into it to some depth. The view comprised the whole of the forme:r lake bed as far as Xiao Zhongdian, bordered by gravel terraces and separated from the Zhongdian bzisin by a low brown ridge, beyond which the Zhongjiamg He emerged from the transverse valley below the natural bridge. In the background were the eroded slopes of Piepen, unfortunately shrouded in rain, airid the highlands to the north of it. Close at hand the terrain was covered with forests of dark green firs, miany of them grey with beard lichens. In the clearings were waterlogged bogs, in which magnificent yakis were quietly grazing. Rhododendrons, willows and spiraeas flourished there, together with Swertia calfcina, a plant 75 cm tall with pure white flowers 4.5 cm [p.83:] across, and the new Saussurea uliginosa. A huge Tibetan mastiff, black and shaggy, made even more terrifying by a collar of red wool, nearly pulled up the long stake to which he was tethered outside a yak-herd's hut; as long as we remained in sight he went on barking and leaping up, his bloodred eyes bulging from their sockets in his rage.

We had not much further to climb to Nguka-la, a pass at 4125m lying between crests about 100m higher. Among the plants growing in the fir woods were Senecio pleoptews, 1.20m tall, with large pirmate leaves and heads of three florets grouped into large corymbs; giant monkshoods (Aconitum scaposum and A. franchetii); two large delphiniums, both probably new species; and the white Beesia calthifolia (Ranunculaceae). The whole district was dark and sombre, an impression created not merely by the woods, but also by the dull green rock, which was a volcanic tuff, and the leaves of the rhododendrons (R. recurvum), while the overcast sky seemed to deepen the gloom. Though the slopes on the eastern side had been gentle, the descent from the watershed to the Yangzi was quite the opposite. The route led down a continuous stairway, laid out in zigzags yet nevertheless exceedingly steep, built from angular blocks of hard igneous rock. I climbed down on foot, scorning my police escort's outrageous suggestion that I should remain in the saddle. Nothing but the clink of horseshoes and the calls of the mule drivers disturbed the quiet of the primeval forest, dark green and dripping with moisture. A flowering currant (Ribes acuminatuni) grew epiphytically high on the trees, the tiny Polypodium sikkimense in moss cushions on their trunks, and the new Pimpinella muscicola (now Acronema muscicoluni), nearly as minute, in the moss covering the rocks. The cleft through which we were descending became less steep, and a pure growth of holly-leaved oak replaced the mixed woodland — which was of a type which might almost have been classified as moisture-loving. Remaining on the northern side of the valley, the track rounded a rocky comer and came to the village of Meti at 3125m. The people living in this group of houses were apparently Naxi-Tibetan halfbreeds, though another hamlet lower down belonged to Lisu. We continued our descent along the valley floor; though narrow, it was not very steep. High up on the side of the valley was a limestone stratum projecting as a sharp crest towards the Jinsha Jiang. Along the stream there were some dank secluded comers with a rich growth of shrubs including the elm-like Euptelea pleiosperma, and further down, where conditions were subtropical, Alangium chinense, Chionanthus retusa and Machil-us ichangensis. The mountain beyond the river was much lower than the one we had just traversed, though at the bottom its slope was very steep. Making a double bend, the Yangzi came flowing towards us. We followed it downstream, and as we had already done a good day's march and it was starting to rain we sought shelter in a house in the little village of Tsondyo, a few kilometres further on. Water soon began to drip through the roof on to the mosses which I had spread out to dry and my other possessions. 1 tried to dodge it by shifting to another spot, but in the end I was forced to have one of the flysheets unpacked and spread over the shingle roof.

Next morning an hour's journey downstream through woods of pine and oak brought us to the ferry opposite a fair-sized village called Jizong. The river was a raging torrent, and the boat was of triangular shape so that it could be towed against the current. On reaching the middle, however, it spun round, and it took no fewer than fourteen oarsmen to propel it to the other bank. We landed some distance downstream, at a spot where an abundant spring welled up in the valley bottom, its outlet shaded by a grove of giant oaks and laurels. Ferns grew in profusion on the leafy soil under the trees. The stream had cut its way through a hollow tree-trunk, and the tree straddled the water as if on stilts. The trees were festooned with huge lianas (Vitis and Celastrus) dangling down to the ground and putting out roots. Some of them, weighted down by cushions of moss, dipped into the turbulent waters, which tugged at these natural ropes and made the tree-tops quiver. The task of ferrying the caravan took until noon, but, having found accommodation in the village headman's house, I spent the rest of the day sorting out my dried plants. Next day we resumed our journey along the large valley which debouches into the Yangzi valley at Jizong. Its winding course was the route to Kakatang, a village three days' march ahead, between Weixi and Xiao Weixi.

The valley ran at first north-westwards without any substantial gradient, breaking through three more or less vertical limestone beds. Infolded between mica-schist and, later, arkose sandstone, these ran parallel to the river further downstream. The valley bottom, in some stretches quite wide, was well fanned and there were several fair-sized Naxi villages. Their inhabitants were extremely friendly: when we were taking our midday rest near one village the people came out with presents for us. On both sides there were broad ridges about 500m above the valley floor. Leading over the northern side, branching off just before Shogo (Shyon-gung on Rock's map) where we spent the night, was the main road to Deqen (Atuntzu) via Benzilan (Bangdsera), a route travelled by several Europeans, but my path now took me into unknown terrain. The volume of the little river was enough to prove that our valley could not really be as short as it is sketchily depicted as being on the maps. Within Shogo the track forked; my route followed the branch coming from the south. The elevation of the margins of the plateau on either side above the valley floor became less and less, the valley became somewhat steeper and then narrower, while the slopes were clothed with evergreen broad-leaved trees with scattered spruces and Tsuga instead of the pines which predominated lower down. From 2500m upwards, growing singly alongside the stream, were some huge dark pyramidal conifers, at least 50m tall. One species was a giant fir, but the other, with flat, sharp-pointed leaves marked by two white lines beneath and large green fruits about 3 cm long, each containing a hard seed, was Tomeya fargesii, a relative of the yew. Among other interesting plants were Actinidia purpurea, an enormous liana; Schefflera stenomera, its palmately divided leaves having [p.84:] remarkably long narrow segments; the new Penta-panax larium, also belonging to the ivy family; and Pterocarya forrestii.

Shatyama (To-na-ko on Rock's map), where I spent the night before leaving the valley, consisted only of a few scattered Lisu huts, as did the nearby hamlets. The stream was still quite substantial and the villagers, so as to get across to their fields of buckwheat, had bridged it with a slender treetrunk, which had been barked, hacked more or less level along the top and provided with branches on both sides as handrails. It was far too slippery for my mountain boots, and I had to crawl across on hands and knees to botanise on the far side; I was glad not to have any spectators for this performance. Among other plants I found Struthiopiteris orientalis and an Angelica. The valley continued south-eastwards, coming, as later became cleiir, from the Lidiping plateau near Weixi. Its total length is hence 80 km and the stream in fact flows in a shallow longitudinal channel on the crest of the watershed between low wooded heights. My route branched off to the west, entered the forest and then emerged on to a col, Akelo, at 3150m. The canebrake was fiill of mud-filled holes between exposed roots, and several of the pack animals tumbled over. Growing there was a large Sphagnum moss (S. pseudocymbifolium) and, in shady channels, a new species of aconite with vivid green flowers (Aconitum chloranthuni). From our midday halt at the Lisu hamlet of Akelo (Ko-lo on Rock's map) I looked down the steep slope beyond and saw that the giant fir did not grow there. As I had not yet seen any of its cones I left Lao Li behind to procure some for me. From the material which he obtained I later identified it as Abies chensiensis. Below us lay the vale of Weixi, bounded on this side by the steep crest sloping down from Lidiping to Akelo and continuing north-westwards for some distance before rising up into a higher elevation near Xiao Weixi, a village on the Lancang Jiang (Mekong). On the far side it was bounded by the Weixi - Lancang Jiang watershed. The vaL was in the main a broad trough, divided across its width by spurs of sandstone, with some limestone here and there, projecting between lateral streams of varying lengths, with a medley of folds and minor peaks. We crossed the lateral valley on foot, turned right and climbed up to a flat boggy patch which was the source of the first stream flowing down the far side. Here we found a group of Chinese hamlets called Tima. My servant was delighted to see them, but I was somewhat less enthusiastic, for on entering the room which was offered to me the stench was such that I recoiled in disgust The villagers were at first reluctant to let me have the altar room under the roof, allegedly because the lady of the house slept next to it, though when I insisted they gave way. Next morning, passing through an ill-lit doorway, I bumped my forehead so violently against the absurdly low lintel that I let out an involuntary cry, staggered and fell almost senseless to the floor. The resulting occipital headache lasted until the following day. Descending the valley — in some places a ravine and in its lower part very steep — I reached Kakatang [note # 95: Ka-ka-t'ang-a "narrow, fifthy conglomeration of huts and pigsties" (Rock, p. 303).] on the "Da He", the "big river" of Weixi, and spent the night a few hours' journey downstream at Anadon [note # 96: "After a march of 30 li (from Ka-ka-t'ang) we pass the small Tibetan hamlet of A-nu-ndo, the Chinese A-nan-to......the principal crops are maize, millet, beans and tobacco; the leaves of the latter are tied to strings which are stretched under the eaves of the houses to dry." (Rock, p.304)]. Next day, 1st September, I sent Lao Li with the soldiers to Weixi to buy supplies. I told him to show the officials my letter of recommendation from Zhongdi-an and ask for a further letter to the Naxi tusi of Yezhe, but to refuse any offer of an escort.

[chapter 18:]