It was still night when everyone rose and packed his bedding. March 4, 1928, will be remembered by all members of the expedition as the end of their captivity in the inhospitable uplands of Tibet. The dawn was softly coloring the sky and the snow peaks of the Shang-shung Range lay behind a deep blue veil. Columns of smoke were rising above the houses of Nag-chu but the streets were still completely deserted. The yak drivers, or lakto in Tibetan, were taking their tea and tsam-pa , squatting near big fires laid on the pavement outside the gates of the expedition headquarters. The loading of yaks began, and it took us almost three hours to start the heavy caravan. We decided to leave Nag-chu after all the transport with the heavy baggage had gone.

At half-past eight everything was finished and the last section of the yak caravan left and disappeared in a narrow gorge across the river. We started, followed by two do-nyers from the dzong , who were to accompany us to Nam-ru. Our Mongol servants who were returning to Tsaidam bade us farewell and presented ceremonial scarfs. They were very sorry to see us depart and feared trouble from the Joint-Governors. They had heard that the Tibetan servants of Dr. Filchner were thrown into prison and apprehended a similar fate. While in Nag-chu, I had spoken to the local authorities on their behalf and they promised to give our men every assistance on their journey to Tsaidam. The expedition issued to them certificates in Tibetan, requesting the Civil and Military Authorities of Tibet to give them adequate help and protection on Tibetan territory, and a high Tibetan lama, who was traveling to Tsaidam, agreed to take them with his caravan. Some four months later, I was gratified to hear that all of them had safely reached their native pasture lands.

Having crossed the river, which was still frozen hard, we ascended a spur, and the rocky silhouette hid Nag-chu with its monastery from our view. An indistinct path wound along the southern slopes of a low ridge stretching to the west.

From now on, we would face the west for some fifty days. The sun warmed the air and the march was pleasant. We had finally escaped from Nag-chu with its refuse heaps and were now on our way to India. We knew that the road would be difficult and that the country which we had to cross lay above 15,500 feet, and had several high mountain passes. The chang-thang or the uplands of northern Tibet is a country of elevated plains and valleys intersected by mighty snow ranges, the highest of them stretching from west to east. Being obliged to change some of our caravan animals on each stage, we had to halt near nomad encampments, and this sometimes necessitated long circuits. Most of the nomad encampments are in mountain valleys, protected from winter winds and gales.

After a two hours' march along the southern slopes of the ridge which lay almost parallel to the snow range of Shang-shung, we ascended a small pass on the northwest. The progress along the northern slopes of the ridge was difficult, the horses frequently sinking deep into the snow. Descending, we reached a narrow valley which lay west of the pass. The frozen pools of water between hillocks made the march difficult. We followed the valley for another two hours and at two o'clock reached a group of nomad encampments pitched on the grassy slopes of nearby hills. Here we received the unexpected news that our heavy transport on yaks, which followed another route from Nag-chu, along the crest of the ridge, was traveling by the northern route some two miles distant from the place where we were. We hastened northward and after fighting our way through a valley buried under deep snow, we reached a small nomad encampment only to learn that no caravan had been seen.

A strong southwest wind blew soon after we reached the place and we decided to wait and send men to look for the caravan. The wind lifted clouds of thin powder-like dust from the heaps of dung left near the encampment by flocks of sheep and it was impossible to stay there. The Colonel and I accordingly went in search of a cleaner place for the camp. Across the valley we succeeded in finding a dry patch of ground sufficient to accommodate all our tents, and waved to the rest of the party to join us. We remained there for about an hour, squatting on the ground among stones and protecting ourselves as far as possible from the bitter southwest wind. Suddenly a rider appeared, struggling in the deep snow of the valley. He proved to be Ochir, one of our Mongols, who reported that our horses and camels had reached the encampments on the southern route and that the heavy transport on yaks was due to arrive by the same route. We at once started back and recrossed the snow-buried valley. Our dog Kadru, tired by the long march, refused to follow and had to be taken in the saddle.

On reaching the southern route, we saw approaching the first group of yaks, carrying our tents and the rest of the equipment required in the camp. Toward six o'clock camp was pitched and we were able to rest a little after the long, tiresome march. The wind abated and we were able to enjoy a calm, warm evening; the thermometer was at -1°C. The last section of our heavy transport arrived only about eight o'clock in the evening, the animals being exhausted by the snows along the crest of the ridge. We experienced some difficulty in finding good water and fuel, for the local nomads had prepared everything for us some ten miles farther on by the banks of a small river called Tsang-po. Considering the late hour and utter exhaustion of the caravan animals, it was impossible to break camp and proceed farther, and after some talk we got everything we needed for the night. There was no water in the valley and we were obliged to thaw snow. The guide who was sent with us from Nag-chu declared that he knew nothing of the route for the next day and we had to arrange for another guide.

A lama from Amdo, who happened to live among the nomads of this locality, tried his best to arrange everything comfortably for us. I asked him how long it would take us to reach the frontiers of India. He waved his hands and declared that Jagar or India was very far off and that it would take us at least six months of strenuous journey!

On the morning of the next day, we had difficulties getting fresh horses. Some of the relay horses supplied from Nag-chu proved to be unsuitable for long, hard marches through snow. The population had no horses and it was with great effort that we succeeded in getting a fresh horse for our cameleer. The morning was fine, and we admired the scenery—the glittering snow peaks of the distant Shang-shung Range.

After a two hours' march along a broad valley with a slight ascent, covered by snow in some places, we reached the banks of the Tsang-po, near which the local headmen had pitched an ornamented white tent for us, and several black tents to accommodate our servants and the two do-nyers from the dzong. This was yesterday's stage and we therefore decided to continue our march and to reach Parta, which was our halting place for the day, according to the official letter or da-yig sent ahead of us from Nag-chu announcing our arrival. We crossed a broad plain covered with snow. One month earlier a passage across the plain would have been most difficult because of deep snow, but now it had settled sufficiently and in some places one could see the quills of the rough grass of the Tibetan uplands sticking out of the snow. On the southwest rose a snow-covered range, apparently a northern branch of the Shang-shung.

Having crossed the plain, we ascended a low ridge and reached a narrow valley with traces of former encampments. A bitter southwest wind which started about noon made the march difficult. The sun disappeared behind heavy, gray clouds which drifted with tremendous speed.

At one o'clock in the afternoon we reached the broad, circular valley called Parta. Several tents were pitched for us in a narrow side valley. A crowd of Chang-pas or northerners with their headman helped us pitch the tents. It was a difficult task, for the wind continued with unabated force and soon turned into a howling hurricane. Everyone hastened to take shelter under the tents. It was getting dark but it was useless to light a candle because of the draught inside; therefore one had to sit quietly and listen to the noise made by the drifting snow and sand striking the outer flies of the tents. The valley of Parta southeast of our camp disappeared behind a veil of rapidly drifting clouds of snow and dust. There was something uncanny in this whistling and howling among the barren mountains and snow-covered valleys of this God-forsaken country. The whole night the wind stormed our camp, as if all the forces of nature were loose and were trying to overcome our tents in a desperate charge, uprooting the tent pegs and breaking the ropes. The Colonel's tent broke down under the terrific pressure. Toward daybreak, the force of the wind abated, and gave us a little rest before breaking camp.

We left at about six o'clock in the morning and having ascended a low ridge to the west, emerged on a vast plain surrounded by low, stony ridges. The violent southwest wind which began soon after we started, made the progress exceedingly trying. The fierce wind of the Tibetan uplands, sometimes blowing with the force of a hurricane, unsteadies the nerves of the traveler and blackens his face beyond recognition. For some time we followed the northern shore of the lake Nam-mar phün-tshok, the first important lake on our way. To the west of the principal lake there were numbers of smaller lakes, surrounded by swamps. We were now entering the region of the Great Lakes, this great drainage-less basin, stretching in a succession of tremendous salt lakes north of the Trans-Himalayas. A passage near the shore of the lake was impassable because of swampy ground. We had to make a long circuit northward, and passed a number of geysers surrounded by layers of salt.

Toward half-past one we reached the halting place for the day. The local headmen had prepared several tents for our party and heaps of argal lay near the entrances of the tents. Out of the hundred yaks, only twenty-nine carrying our tents reached the spot. The rest, unable to proceed farther, stopped for the night near the lake Nam-mar phün-tshok. It was disappointing news as it meant we would have to stay another day, waiting for the heavy transport. The wind abated toward sunset and we hoped for a quiet night. Golubin, who had had fever for the last few days as the result of riding in snow and wind, was feeling worse and our doctor gave him some aspirin and brandy to keep him warm.

On the next day, the heavy transport arrived at half-past eleven in the morning. The drivers related how they had lost ten animals on the way. I tried to get fresh animals from the local nomads to replace our tired ones, but the headmen complained that most of their cattle had died during the hard winter and although the southwest wind, which had been blowing every day for a fortnight, carried away a great deal of snow from the plain, still the cattle had nothing to graze on. After much talking we succeeded in obtaining only fourteen pack animals to carry our baggage to the next stage.

The broad plain on which we were staying was still called Parta, which is the name of the pasture grounds situated along the plain for some forty-five miles. The valley was surrounded by low, stony ridges with eroded surfaces. A snow-covered peak to the southwest of the camp was called by the local population, Jung, and was some 18,900 feet high. It dominated the whole of the Parta grazing grounds. The mountain ranges toward the south of the valley were higher than those to the north and made a gradual transition toward the higher ranges forming part of the northern branches of the Trans-Himalayas. The grazing grounds of Parta are part of the Nag-chu-ka district.

On the next day at seven o'clock we broke camp and followed the plain, which gradually rose toward the west. Having rounded the peak called Jung, we stopped after traveling for two hours and a half, near a number of encampments surrounded by the usual argal walls. It is the custom of local people to make very short marches in order to give sufficient time for the animals to satisfy themselves on the poor grazing grounds of the Tibetan upland. We tried to make the caravan men proceed a little farther, but they declared that they would not move, as it was the stage for the day. All further arguing was useless.

The next day was very cold with a slight southwest wind. We passed the Jung peak and crossed the snow-covered Lamsi Pass. The descent, although not steep, was difficult for horses, because of the hard frozen, slippery snow. For two hours we continued to follow a stony trail in a south southwesterly direction ascending and descending numerous low ridges, all branches of the higher range in which the peak Jung was situated. Having reached a narrow snow-covered valley immediately south of the peak, we halted for the day. The local nomads had pitched three tents for us. Toward the west at a distance of some six miles, we noticed a huge salt lake, stretching north northwest by south southeast. The local people call it Nya-shing tsho. The valley in which we pitched our camp was still called Parta and represented the frontier of Nag-chu-ka. Tomorrow we would enter the district of Chö-khor, a property of the Sera Monastery in Lhasa.

We broke camp the next day about eight o'clock. Having ascended a snow-covered ridge to the southwest of our camp, we emerged on a sandy plain, which stretched south of the Nya-shing Lake. The lake was seen toward the northwest, beautiful with its dark blue surface encircled by white salt-covered shores. Local people affirmed that some kind of "aquatic cow" was found in the lake. These cows were sometimes seen at sunrise and the air resounded with their lowing.

Toward noon we reached a small rivulet flowing toward the lake, and having crossed it, entered a hilly region cut by narrow valleys. For some time we continued to follow the course of a stream, but on reaching the foot of a pass, were stopped by one of the Nag-chu do-nyers, who came rushing back, reporting the disquieting news that the official letter from Nag-chu had been wilfully kept by the local headman. It was impossible for us to continue on our way as no arrangements had been made ahead of us. It was necessary to find the guilty man and to arrange for another dispatch rider to carry the letter to Nam-ru. I insisted that he should be brought to me at once. The culprit was produced, a fellow with a criminal face and shifty eyes. He was silent to all the questions we asked him and I ordered him to be tied up, pending further inquiry. We decided to hand him over to the higher local authority for proper punishment, according to the custom of the country.

In the evening we had the great misfortune to lose the fine riding pony of Professor Roerich. It was a Karashahr pony and one of the best horses among our riding animals.

During the night a strong southwest wind tore some of our tents. Heavy bags of flour and fodder had to be placed on the tent pegs. Toward sunrise its force abated and we were able to start in mild weather. On crossing the pass west of our camp we ascended a succession of low, sandy ridges covered with good grazing, well preserved under the snow. Good grazing was usually found on patches of land covered by this detritus sand loess. This is generally true of Tibetan uplands. On a vast, sandy plain southwest of the Nya-shing tsho, the local peasants had prepared a tent for us. As we were only two hours on the march, we decided to go for another hour toward the head of the valley, and ordered the people to transfer the tents to another place. We rounded the northern shore of a small salt lake, separated from the Nya-shing tsho by a low-lying plain. Apparently it was formerly connected with the larger lake, but at some time had shrunk and the two lakes were now separated by the narrow strip of land. The smaller lake, during the period of summer rains, filled the dry bed. The local people told me that the length of the big Nya-shing tsho equaled three days' march, that is some fifty miles. The valley which we were crossing was surrounded by low, sandy ridges. Near the lake swampy patches of land were covered with scanty grass, which afforded pasture for the cattle owned by the nomads of Chö-khor. I was told that the local population had lost most of its cattle during the severe winter. We halted near a brook, among sandy hills situated northwest of the valley. The thing that surprised us most was the scarcity of big game. Since leaving Nag-chu, we had seen only two kyangs. The local nomads said that all the game had retreated far to the north in search of pasture. On the way we saw some carcasses of young Tibetan antelopes which had perished during the winter from want of proper grazing.

The local nomads had little information about Nam-ru. Some said we should reach the district of Nam-ru in four days, others said ten. The headman told me that the lake had another name, Dung-tsho, so called because of the lowing of "aquatic cows." I failed to find out what kind of animals were meant by the name. We first thought of otter, which we noticed at Chu-na-khe and at other places of the Tibetan upland but then the water of the lake was salted. I tried to obtain the skin or the skull of the animal, but the local people had none.

We broke camp as usual in the early morning and having crossed the broad valley in a westerly direction, passed a series of small lakes. The largest of them was called Chag-tsho, others again called it Chang-tsho. It was extremely difficult to find out the correct names of places, as most of them had several.

After a four hours' march, we turned into a mountain valley stretching toward the northwest and halted near a nomad encampment. Toward the southwest could be seen the small lake of Pü-tsho from where the local population got its salt. The valley belonged to another district called Sa-skya which was the property of the great monastery of Sa-skya in the Tsang Province. The region was barren, with an almost complete absence of vegetation. The local nomads seemed to be entirely degenerated and wild, as if the barrenness of the country had deprived them of the spark of humanity possessed by their neighbors to the east.

On the next day, we continued our route across a succession of low, sandy ridges. A piercing southwest wind, blowing from the Pü-tsho, froze us despite our Siberian fur coats. The continuous ascents and descents were exceedingly trying for the poor animals. After a two hours' march, we emerged on a rolling plain, stretching from the lake Pü-tsho toward the mountain ranges to the north. We camped in a narrow valley northwest of the lake, on the bank of a rivulet flowing in several beds. The local nomads still belonged to the Sa-skya Monastery. They refused to make the necessary preparations and the do-nyers from Nag-chu mildly whipped two of them. We found that the population of church lands was averse to obeying orders from Lhasa and our do-nyers had difficulty procuring the necessary supplies and relay horses. South of the Pü-tsho Lake rose mighty snow peaks. During the day's march we saw many carcasses of Tibetan antelopes. The local gem-po or headman assured me that we should reach the district of Nam-ru the next day. So far we had been obliged to change guides every day, as the men knew the route only for one day's distance from their own encampment.

We broke camp at half-past seven in the morning and having ascended a rocky spur to the west, rode again on the plain of Pü-tsho. We had to cross the broad plain toward the mountain ridge southwest of it.

The plain was covered with gravel, alternating with patches of loess covered with grass. In the distance we noticed the first herd of kyangs. We did not try to shoot any of them as shooting was strictly forbidden by our passport from Nag-chu. About noon we halted near a nomad camp. One of the do-nyers went ahead in order to reach Chag-lung-khar, or Nam-ru dzong the same day and had asked us to halt here to give him time to reach the dzong and prepare animals for us.

There was another route to Nam-ru, straight along the northern bank of Pü-tsho, which shortened the distance by two days. The route was completely deserted and no encampments were found along it.

The local nomads, called Chang-pas or northerners, looked quite different from the Hor-pas. Here the Mongoloid type prevailed with developed cheek bones and round heads. The physical development was below normal. Scurvy and rachitis were very common. Today was the first warm day since our departure from Nag-chu; in the evening the temperature was about -5°C.

We continued our march across the plain toward a narrow mountain valley in the ridges southwest of the plain. The plain was everywhere covered by good grazing and we noticed that the headmen lived in stone houses at the foot of the hills, a great improvement in comparison with the Chang-ri wa nomads of the Nag-chu district. After a four hours' march we entered the stony valley leading to a small pass. We halted at the foot of the pass near some tents. We had a miserable afternoon, heavy clouds covered the sky and the wet snow and cold wind were most unpleasant.

Next day we started at eight o'clock and crossed the pass northwest of the camp. We descended into a vast plain reaching far to the north. The plain was covered with gravel, alternating with sand. There had been a heavy snowfall during the night and the surrounding hills were covered with fresh snow. The barren hills that surrounded the plain were greatly eroded by the continuous southwest winds and the amplitude of temperature. Along the foot of the hills were moraines.

The southwest wind is one of the greatest factors in the physical geography of the Tibetan uplands. In the winter it brings snow, in the early spring it melts and carries away snow and erodes the southwest slopes of the mountains. In the period of summer rains (July and August) it brings moisture and sometimes causes heavy rains. The country in which we had been traveling for the past few days could be appropriately called a geological cemetery.

For some time we went along the foot of the hills that rose south of the plain, then we turned westward and having crossed a rocky spur crowned with an obo, we headed toward the mountain ridge (18,000-20,000 feet) which rose to the south.

After an hour's march past sandy hills, we entered the mountain valley in which Nam-ru dzong, the administrative center of the district, was situated. The valley was full of hummocks that made progress extremely difficult. To reach the settlement, we had to cross a large patch of swampy ground, frozen and slippery in winter. The village, a collection of dirty, half-ruined stone huts, was situated on the bank of a small river that flowed northward toward the plain. Large heaps of refuse and walls of argal surrounded the houses. Numerous nomad tents were scattered in the vicinity. The house of the dzong-pön in the center of the village was a miserable building with several prayer flags on its roof. People said that formerly the district of Nam-ru was governed by a local chieftain, whose house we had seen near the entrance to the valley, built on a windy spot facing the plain to the north. Since the flight of the Tashi Lama, a dzong was established at Chag-lung-khar to protect the route along the western shore of the Chang Nam-tsho or Tengri-nor. In official correspondence the place was known under the name of Nam-ru dzong, but the local people still called it Chag-lung-khar. Two Lhasan officials were stationed in the dzong, a layman or dun-khor, and a lama official or tse-drung. The dun-khor was absent in Lhasa, but his colleague, the tse-drung, was staying at the dzong.

On our arrival at the village we were met by the do-nyer from Nag-chu and a mixed company of local chieftains, tribesmen, and traders from inner Tibet. We were shown a place for our camp near the dzong, but as it was filthy and thickly covered with dung, we decided to move across the river. Here a number of terraces, with deep-cut, eroded canyons, afforded suitable dry ground for a camp. Having pitched camp, we sent word to the dzong-pön of our intention of coming to see him. The messenger returned and informed us that the dzong-pön would be very pleased to receive us. We recrossed the river by a narrow, wooden bridge and proceeded to the dzong. A number of retainers were waiting for us outside the gate.

The courtyard of the dzong was densely packed with bales of wool and butter, all local taxes being paid in these commodities. A black nomad tent was pitched in the center of the courtyard. We were ushered into a small room which was the apartment of the dzong-pön. The official himself, a young man of about thirty, was seated on a low seat near an altar, on which stood a clay figure of Sakyamuni, the Buddha. Behind the altar hung several thang-kas or painted images representing Avalokitesvara in his four-armed manifestation, the white Tara, and Yamantaka. Heavy silver cups and an offering lamp stood in front of the images. On the wall opposite the window hung a large collection of firearms. Tibetan officials are passionate buyers of arms and pay exorbitant prices. The dzong-pön of Nam-ru had in his collection about thirty Russian service rifles, several Japanese Arisaka rifles and German Mausers, besides several German Mauser pistols and Nagan revolvers. A number of richly adorned Tibetan swords hung with the firearms. Among them I noticed short Bhutanese daggers, long, double-hilted swords carried on the back by Tibetan foot soldiers, and several long, heavy sabers or sho-lang. I asked the dzong-pön what use he made of this quantity of arms. He smiled and replied that in case of necessity he had to protect the dzong.

We were seated on a long bench near the window on the right hand of the Governor and Tibetan tea was served on a low table in front of us. The dzong-pön was the brother of the wife of the Governor of Nag-chu and pretended to be our friend and well-wisher. He informed us that the passport or lam-yig from Lhasa had not yet been received and that he did not really know what to do with us. But as we were recommended to him by the authorities of Nag-chu, he would send us off to the next dzong. He had not received letters about us and was unaware of our coming. We should have to remain for one day at Nam-ru and give him time to collect the necessary animals for our caravan. We urged him to hasten our departure and to speak at once with the chieftains and the local headmen. The headmen were accordingly invited. The chieftain maintained that we should have to pay high prices for the caravan animals as it was not the proper season for travel. The local population had no pack animals in the vicinity of the dzong, except some dzo kept near the tents. Because of the heavy snowfalls during winter, most of the nomads decamped toward the north and were staying at places four or five days' journey from here. It would take some fifteen days to get to Nag-tshang and caravan animals would have to be changed on each stage. This was a great inconvenience, for in case the headmen on the stages would be unable to supply us with a sufficient number of caravan animals, we should be held up indefinitely on our way.

We insisted that the Governor should do all in his power to make the prices a little more reasonable and to arrange everything for the following day, as we were in a great hurry and the state of health of some of the members of the expedition was giving much anxiety. In case of the death of any of the members, there would be a source of friction and endless trouble for Tibet. The dzong-pön inquired whether we were aware of the fact that many Europeans had been murdered lately in China and he desired to know about the actions taken by the European and American Governments. I replied that the Government of China had had to accept full responsibility for crimes committed on its territory and would have to make good the inflicted damages. The dzong-pön said that he was unable to order the peasants to bring the animals at once, for his post was newly created and the peasants, accustomed for generations to be ruled by their tribal chieftains, were very reluctant to obey an official from Lhasa who was a stranger to them. There had even been cases of the murder of officials by local nomads. The dzong-pön expressed his deep surprise that the Government had sent us by such a difficult and circuitous route, instead of allowing us to cross inner Tibet somewhere in the Tsang Province past Lhasa and Shigatse.

After much talking prices were finally settled—the hire of a riding horse from Nam-ru to Shentsa was eight ngü-sangs and five sho. The hire of a yak was five ngü-sangs. A mounted messenger would be dispatched at once to Shentsa to announce our arrival and to prepare yaks on the stages, so that we should not be delayed on our way. The departure was fixed for two days later.

The whole of the next day was spent in trying to obtain some fodder, frozen milk, and mutton. To make the Governor a little more friendly, we presented him with a pair of Zeiss field glasses. This made a good impression and we were supplied with twenty bags of fodder for sixteen ngü-sangs and five carcasses of sheep. We distributed small presents to the yak drivers who came with us from Nag-chu. Those who reached the camp first, received somewhat more. All of them busily occupied themselves with the sharpening of their swords on stones, as they were returning together to Nag-chu and did not trust each other.

Late in the afternoon I went to speak to the dzong-pön about the do-nyers who were to accompany us to Shentsa. I asked him to send a strong and alert man, with a good knowledge of the route and the country we should have to cross. He promised to send the same man who accompanied the Filchner Expedition to Shentsa dzong. I took leave of the Governor, and returned to our camp to prepare the departure.

March 18, 1928. The early morning was clear and the day promised to be sunny. We wished to hurry our departure in order to cross the mountain ridge situated west of the Chag-lung-khar, before the beginning of the daily afternoon storm. The pack yaks were collecting very slowly. One could see riders driving small groups of yaks before them, coming toward our camp. Because of the heavy snows in the winter, most of the animals were driven far to the north and the local population had difficulties supplying us with a sufficient number of pack animals. Notwithstanding the delay we started at ten o'clock and rode toward the mountain pass Ta-mar-khyer (about 16,700 feet).

The trail followed a narrow gorge with a rather steep ascent toward the summit of the pass. The gorge was blocked by bowlders, and large accumulations of débris were like ramparts, erected for the protection of the pass. From the top of the pass we descended into a circular mountain valley sheltered on all sides by high mountain ranges. From the summit we obtained a view of the small lake of Nam-ka tsho, lying in a vast intermontane plain, north of the pass.

Several tents were pitched for us near a ruined stone hovel, formerly occupied by a wealthy headman. The local headman, an old man whose head trembled and twitched continually, met the caravan of the expedition. We had to change our caravan animals again, but the change was effected smoothly, thanks to our guides from Nam-ru, who proved to be exceptionally efficient fellows. The chief guide was a Tibetan nomad from the region of Gyangtse, a man of good local knowledge, who had only quite recently served as guide to the Filchner Expedition. He was dressed in a dark violet coat trimmed with fur, and wore a dark red turban on his head. A long Kham-pa sword was attached to his belt, and a huge, silver charm box was slung across his shoulders. His assistant, with the official title of mi-ser thü-mi or "representative of the peasants," was a pious man selected from the local nomads. He was a quiet man and had a considerable influence over the local tribesmen. He was mounted on a yak and managed his animal with wonderful dexterity.

We continued our journey westward, following a narrow trail leading to another pass, the Na-tra La (16,900 feet). The route was indicated by several mani -walls built on either side of the trail. This route is usually used by pilgrims from the nomad districts of Tibet on their yearly pilgrimages to the holy Kailasa Mountain. From the summit of the Na-tra La opens a wonderful panorama on the surrounding mountain region and the vast, sandy plain north of it. On the southern horizon rose a high, rocky range with an average elevation of about 17,000 feet. At the foot of the range stretched low spurs of sandrocks and fiery red limestone. In the rays of the morning sun the desert landscape sparkled with varying shades of red, purple, violet, orange, and yellow. The country bore a vivid resemblance to the desert region of Arabia, and the dreary kingdom of sand of Chinese Turkestan. A small river, the Cha-war tsang-po, emptying into the salt lake of Pang-gong tho-cha lying in the plain toward the north, supported scant pastures along its banks. The descent from the pass was effected by a narrow trail leading through an erosive gorge cut through a succession of sandstone ridges. We halted for the day at the foot of a low, sandy spur, not far from several nomad encampments. The day was warm and windless but the night was unexpectedly frosty, the temperature falling to -12°.

March 20. We started comparatively late, at eight o'clock in the morning and having crossed the spur west of our camp, descended into the plain. We camped on the opposite bank of the Cha-war tsang-po, having covered only four miles. We protested, but the guides informed us that we had reached the frontier of another district and had to wait for fresh pack animals. Our present caravan animals were not supposed to go any farther. The place of our new camp bore the name of Sho-de na-mo-che and belonged to the great Tibetan monastery of Tashi-lhun-po. The local headman lived in a poor, stone hovel, surrounded by several courtyards used as inclosures for sheep.

March 21. Before starting, we had an exciting time with our new relay horses. The animals were fresh from the pastures and had spent a whole winter in the open without being saddled or ridden. The result was that all of them were wild and pranced and kicked.

The trail followed a mountainous country of eroded hills of granite, sandstone with layers of red limestone. Numerous mani -walls indicated the route. After a five-mile march over a highly intersected hilly region, we reached a vast plateau covered by scant weathered grass. This plateau was a favorite grazing place for wild game and we observed many Tibetan gazelles ( Gazella picticaudata ).

Having crossed a low mountain col, the Chyang-ku La, we descended through a narrow gorge into a narrow valley situated between two sandstone spurs. The numerous dry beds of temporary rivulets indicated the considerable amount of precipitation received by the locality in the rainy months of the summer. From the top of the pass, we obtained a fine view of the valley of the Buchu tsang-po, a small river emptying into the lake Tsho Ziling. The locality bore the name of Lung-mar, because of the red limestone present in the valley.

March 22. We started early notwithstanding a frosty morning, with a temperature of -12°C. A piercing west wind made the progress of the caravan trying. Leaving the camp, we ascended a low mountain col called Lung-mar La and descended into a valley lying southwest of the pass. For about five miles we followed the valley in a southwesterly direction, then turned southward and skirted a large massif of granite. The local nomads told us that the Tashi Lama passed by this route during his flight in 1923. After a sixteen-mile march we camped in a narrow valley sheltered by undulating, grassy hills. The place was called Do-ring or "The Lone Stone" because of curious megalithic monuments found in its vicinity. These monuments were the first of that kind discovered in Tibet. Until now, only a few of the sanctuaries of the primitive Bön religion have been discovered. These are mostly crude stone altars or lha-tho , and a number of them are known to exist in western Tibet, where they have been studied by the late Dr. A. H. Francke, of the Moravian Mission at Leh.

The megalithic monuments of Do-ring, situated some thirty miles south of the great salt lake of Pang-gong tsho-cha, date back to the pre-Buddhistic period of Tibetan history. They consist of important alinements of eighteen rows of erect stone slabs. Each of these alignments was drawn from east to west, having at its western extremity a cromlech or stone circle consisting of several menhirs arranged more or less in a circle. The menhirs are vertically planted with a crude stone table or altar in front of them. It was evidently a sanctuary of some primitive cult. But what was its age and use?

If one compares the famous megalithic monuments of Carnac in Brittany, to the discovered megaliths of Tibet, he is at once struck by the remarkable similarity of the two sets of monuments. The Carnac alinements are situated from east to west and have at their western extremity a cromlech or circle of stones. The Do-ring monuments have precisely the same arrangement. The sacerdotal use of the Carnac monuments remains unknown to the present day, although numerous explanatory theories are advanced. It seems to me that we possess a clue to the explanation of the megalithic structures of northern Tibet. The megalithic monuments of Do-ring have a large figure in the shape of an arrow laid out with stone slabs, and situated at the eastern extremity of the alinement with its point toward the alinement. The arrow is an important symbol in the ancient nature cult of Tibet, and is connected with the cult of the sun and heavenly fire in the form of lightning, which it symbolizes. The present-day nomads wear ancient brass arrowheads as amulets, which are said to represent petrified lightning after it has struck the ground. Incidentally, it must be said that the arrow is sometimes regarded as a symbol of King Kesar, whose connection with nature worship has been definitely demonstrated by Dr. Francke. The presence of the arrow figure at the eastern extremity of the Do-ring monuments indicates that the whole structure was dedicated to some nature cult and very possibly to that of the sun, of which the arrow is a symbol. This is an important conclusion, since up to now no megalithic monuments could be satisfactorily explained.

It is interesting to notice that most of the discovered megalithic monuments are found along the great pilgrim route south of the Great Lakes that leads toward the Mount Kailasa, the abode of gods, and the sacred places on the Nepalese border. In a future publication on the Bön religion, I hope to show that this pilgrim route was a migratory route in the past, and as such acquired a religious significance which it still preserves.

The megalithic monuments discovered by the expedition are completely unknown to the modern population. At Do-ring, some of the big menhirs inside the cromlech had traces of butter libations, and I was told by a local headman that the stone was the abode of a lha or god protecting the route and travelers. Neither he nor the men of his tribe knew anything about the origin of the stone.

March 23. At the start, we had some trouble with local horses, which had never seen camels before and dashed in all directions. Some time passed before we were able to bring all the animals back to camp. The route followed a mountain gorge, which stretched in a southwesterly direction. Wind-eroded granite massifs rose on either side of the gorge. The path led across a low, sandy ridge, from whose top we looked over a broad grass valley dotted by encampments. The valley was inclosed by high, sandy ridges with eroded surfaces. After crossing the valley, we entered a system of low, sandy ridges, deeply cut by dry river beds.

At this point we found another group of megalithic monuments, three menhirs with stone slabs arranged around them in a square. The offering table in front of the largest of the menhirs, standing in the center, was missing and there were no traces of libations. The sanctuary was probably long ago abandoned. The trail then turned southward and followed the course of a small stream. The way was lined by numerous mani-walls. Tents had been prepared for our use in a small side glen. On all sides of the glen rose precipitous granite rocks. The process of denudation had laid bare the granite structure of the western slopes of the mountains, exposed to the continuous action of the prevailing southwest winds. Near the camp a moraine formed a kind of rampart, blocking the valley from the southeast. The granite slopes of the ridge bore many traces of glaciation.

Close to our camp, Professor Roerich discovered several graves probably belonging to the neolithic epoch. They were inclosed by stones arranged in a square. Each grave was laid out from east to west, and at its eastern extremity was a large bowlder. Evidently the body was buried head eastward. Judging from the outward appearance of the graves, they belonged to the same epoch as the megalithic monuments discovered by the expedition. It was a matter of great regret that Tibetan authorities strongly objected to scientific excavations and we had to content ourselves with taking photographs and making a survey of the site. The local population was wholly unaware of the existence of the graves.

The conical stone huts which serve as storehouses for the local population are highly interesting. They are covered with hides and plaster. Local women wear strange headdresses consisting of an oval frame with black cloth stitched on it and somewhat reminding one of the national headdresses worn by Slavonic peasant women. The surface of the cloth is covered with coral, turquoise, and brass and silver ornaments, among which Tibetan silver coins and Indian rupees are prominent. Over the ordinary nomad coat, the local women wear a kind of cape made of gray homespun cloth, gathered in folds at the back. The costume of the men is the usual fur coat and large fur bonnet, covered with red or blue cloth.

The place was called Ratri and is twenty-two miles distant from Do-ring and four days from Shentsa dzong, the administrative center of the Nag-tshang district.

March 24. Because of an unavoidable delay with caravan animals, which had to be brought from beyond the high ridge to the south of the valley, we started at half-past nine. After crossing several sandy ridges we entered a broad valley in which lay the lake Go-mang tsho. It is a salt lake with saline crust and sandy beaches on its shores. Fresh water rivulets, rising in the surrounding hills, empty into the lake. The route passed north of the lake. This particular portion of the route was rather dangerous because of quicksands intersected by belts of sand dunes. On passing the lake, we entered a broad, gravel plain with scant, coarse grass. Herds of kyangs and Tibetan antelopes were seen grazing in the distance. While crossing the plain we met the dispatch rider sent from Nam-ru to Nag-tshang announcing our arrival. He was on his way back to Nam-ru, and reported that everything was ready at Nag-tshang and that we could start from there without any further delay.

Crossing the plain, which was about eight miles in width and seemed endless, we entered a broad mountain valley, sheltered by low hills of sandstone. A strong, western wind made the going extremely difficult and we were glad to find tents pitched for us at a place called Ching-kar, a bleak spot inclosed on every side by mountains. We obtained here a number of bronze arrow heads. It appears that arrow heads are often found in the locality and are believed, as has been noted, to be found on places where lightning has struck the ground. The local people are convinced that these arrows are not fashioned by men, but represent petrified lightning which assumes the shape of an arrow after ten or twelve years of lying under ground. Such arrows are considered to have miraculous powers and they are worn as amulets.

March 25. We had another delay with caravan animals and the expedition started at half-past eight notwithstanding the fact that all rose very early. Mrs. Roerich was fortunate in making an interesting discovery, a brass buckle, representing a two-headed eagle in a circle. The motif of the double-headed eagle is frequently found on antiquities from the northern Caucasus, and can be traced back to the Hittite art of Asia Minor. This buckle demonstrated once more the important fact of the intrusion of central Asiatic motifs in animal style into Tibet.

The route lay across a vast, intermontane plain, covered by strips of sand and gravel. The landscape was extremely monotonous and we were glad to enter again into a narrow mountain valley southwest of the plain. Here the country suddenly changed and we found ourselves traveling through an intersected country of barren and eroded ridges. In narrow mountain glens were seen nomad camps. We passed the house of a local headman, a stone structure surrounded by a stone wall of rough masonry with the usual prayer flags fixed on it. Local nomads seldom erect permanent structures. All the stone structures in the region are either the dwellings of local headmen or serve as storehouses to a community of nomads.

To the west rose a mountain ridge with an average altitude of some eighteen thousand feet. Many of its peaks were snow capped and its long, jagged wall was white with winter snow. After a two hours' ride through this mountainous country, we came to a vast plain inclosed by low, rolling hills. Several black nomad tents were waiting for us at the foot of an eroded massif, which gave to the plain its name of Lug-thang trak-na or "The Rocky Nose of the Sheep-plain." Herds of yaks were seen grazing at the foot of the distant hills. A strong and steady western wind prevented the pitching of tents and we had to wait fully two hours before we were able to fix our camp. After the camp was pitched, I went to reconnoiter the plain, which was free from snow. It was undoubtedly a former water basin, as numerous fossils proved. The distance of Lug-thang from Ching-kar was about eighteen miles. We had to take special precautions to fix the tent pegs with heavy bowlders and bags of grain, because the steady, western wind resumed with renewed violence soon after sunset.

March 26. The road lay in a south southwesterly direction across the southwestern section of the broad Lug-thang plain. The day was glorious and we thoroughly enjoyed the ride across the plain and the mountain valley which we entered. A gradual ascent led to a flat mountain col, across a longitudinal ridge.

The descent was not at all steep and led into the basin of the gYung-drung tsang-po. The colors of the river valley, and the jagged walls of cliffs that sheltered it from the southeast, reminded us of the mighty ranges of westernmost Tibet. We forded the river opposite a basalt massif with a mani inscription on it. The water reached the stirrups and the current was swift. Along its banks the river was still frozen hard, but in the middle it was already flowing free of ice. It was extremely difficult to make the animals plunge into ice-cold water and then again make them climb the slippery surface of ice. The opposite bank of the river was covered by swamps and we had to keep close to the rocks. We halted for the night on the left bank of the gYung-drung tsang-po which had its source in the lake Mu-chyu tsho, lying toward the southwest and emptied into the lake Kyer-chung tsho, which lay on the northwest. To the south of the river valley rose a high, rocky ridge. According to information received from the inhabitants there is another route south of the lake Mu-chyu tsho. The place where we were camping for the night bore the name of gYung-drung kong-ma or "The Upper gYung-drung." The distance from gYung-drung kong-ma to Lug-thang is about seventeen miles. We again found some arrow heads of the same shape as those of Do-ring, Ratri, and Ching-kar.

March 27. We were again delayed with our caravan and were only able to start at seven o'clock. For some time we followed the valley of the gYung-drung chu and then ascended a sandy spur. Toward the south we suddenly beheld a grandiose panorama. Mighty rocky ridges stood shrouded in a deep violet veil; then suddenly the rays of the rising sun struck it and the whole mass of rocks sparkled with red, crimson, and purple colors. The layers of sand at the foot of the mountains lit with a fiery red and golden yellow. The undulating hilly uplands of northern Tibet were left behind and we entered another country, a region of ranges and mountain valleys, forming the northern buttresses of the mighty Trans-Himalayas.

We descended a broad, gravel plain that stretched north of the range. This plain also represented a dry sea bottom, as the presence of fossils demonstrated. On the plain we met the first yak caravan carrying barley, which was bound for the distant encampments situated on the northern upland. The traders told us that until now the passes south of Shentsa dzong had been blocked by snow and all the caravans from the Tsang Province had to wait on the southern side of the mountains. At present the snow was rapidly melting and a passage with loaded animals was possible. It was welcome news for us, for we were a little worried by the condition of the route over the high mountain passes of the Trans-Himalayas, north of Saga dzong.

On crossing the valley in a southwesterly direction, we entered a mountain gorge, walled on either side by cliffs and eroded rocks. A small mountain stream flowed through it and directed its course southward. The surrounding landscape was of exceptional grandeur. Mighty faults, signs of gigantic moves of the earth's crust, formed a fantastic design on the rocky walls that sheltered the narrow gorge. After an eighteen-mile march we pitched camp on a broad plain, surrounded by low hills. A terrific windstorm started to blow soon after we had pitched our tents. Thick clouds of dust enveloped the mountain tops and everyone was forced to remain inside his tent holding the tent pole, which swung in all directions in imminent danger of breaking down with the tent. Tent ropes flew in all directions and tent pegs were carried about the camp with a rattling noise.

I had an unpleasant experience with my Tibetan dog Kadru. The dog started chasing a herd of goats, and succeeded in killing one. The owner came to the camp and complained and I had to pay one ngü-sang for the damage. These Tibetan wolfhounds are known for their bad manners and very often attack goats, sheep, and even domestic yaks.

The storm abated toward evening and the night was quiet and warm, the thermometer registering -2°C.

March 28. We started early this morning in order to reach Shentsa dzong early in the day. After following the steadily rising valley for about two hours, we reached the Dig-pa La. From the summit of the pass (17,600 feet) we obtained a unique view of the high snow massif of Shentsa Jyel-khang, towering on the southwest of the broad plain of Shentsa. One of the numerous snow-capped peaks of the range rose to the height of 22,000 feet. The vast plain on which the dzong was situated was clearly visible as far as the lake Chyaring tsho, on the northwest of the plain. The descent from the pass was very steep and we had to dismount and walk the whole distance. On the southern slope of the ridge we saw Tibetan gazelles and several kyangs . The countryside was free from snow, which lay only in ravines and on the northern slopes of the mountains. Before reaching the village of Shentsa, we crossed the frozen Shentsa chu, flowing in several narrow channels.

Shentsa dzong is the administrative center of a large province stretching north of the Trans-Himalayas and including in its territory most of the great salt lakes of the Tibetan upland. The village is a motley collection of some sixty half-ruined structures of sun-dried bricks with roofs covered with sod, and there were heaps of refuse in front of the houses. The houses were formerly whitewashed but apparently for many years no paint had been applied to the walls with the result that they were of a dirty gray color. In the center of the place rose the only two-storied building—the dzong or official residence of the local Governors, both of whom were away in Lhasa. As is customary, this official structure was painted yellow and had jyal-tsen or black religious banners on its four corners.

The streets of the village were, as usual, extremely filthy but thanks to the dry, cold air, the refuse heaps were still frozen. The population consisted of some fifty inhabitants, many of whom were in a state of utter poverty. There was no cultivation and all the grain and tsam-pa supply was brought at a considerable yearly expense from the Brahmaputra Valley. The Governors, who were absent in Lhasa, were replaced by two ku-tshap (sku-tshab) or delegates, their personal servants, who acted as governors in their absence, but were afraid to take any responsibility upon themselves.

We pitched our camp outside the village and soon received the visit of the two officials. One of the two men wore a pair of huge Chinese spectacles. He was a local nomad with some smattering of Lhasa Tibetan. His colleague was a plain headman with but little dignity about him. They presented us with the usual ceremonial scarf and informed us that the passport in which our future route was outlined had been received from Lhasa. We had had secret hopes of being able to proceed by one of the many mountain routes across the Trans-Himalayas south of Shentsa dzong. Now that the route was clearly indicated in the document, we had to follow the circuit route which would be nevertheless highly instructive. The route indicated in the passport was Shentsa dzong-Saga dzong-Shekar dzong-Kampa dzong-Sikkim. Our outgoing Tibetan frontier outpost on the Sikkim frontier was Kampa dzong, from which place we would probably have to travel via Lachen to Gangtok.

The two officials informed us that all of the caravan animals needed for our transport were ready and that we could start the next day. We had to obtain a supply of grain to last until Saga dzong, for there was but little hope of obtaining sufficient grain along the route, which passed through a thinly populated region. After great exertions we succeeded in collecting eleven bags of barley and dry peas, each bag costing the exorbitant price of fifteen ngü-sangs . We were badly in need of butter and were glad to obtain four pounds of it and one piece of poram or barley sugar.

We were to follow a route skirting the southern shores of lakes Chyaring tsho, Ngantse tsho, and Dangra yum-tsho. From Chok-chu, a place situated west of Dangra yum-tsho, we would have to turn southward and cross the high Trans-Himalayas to Saga dzong. This was a difficult route, considering the poor state of our animals and our rapidly diminishing funds.

March 29. All night loud shouts of herdsmen announced the arrival of parties of pack yaks. Long strings of yaks formed a black wall around the camp. We started at seven o'clock and having skirted a limestone spur northwest of the dzong , headed across the gravel plain toward the southwestern shore of the Chyaring tsho Lake. The plain was watered by the Tag-lung tsang-po, a river that emptied into the Chyaring tsho. We found the river still frozen hard and had no difficulty in crossing it. The trail crossed several old moraines descending toward the lake and, ascending a low spur, entered into a highly intersected mountain region immediately south of the lake. The trail climbed the crest of the range, towering above the lake shore, which lay about three hundred feet below the trail.

We crossed a low pass, the Laptse nagri (the Laptse na-mo La of Sir Henry Hayden's map), from which we obtained a fine view of the lake and the surrounding country. The mountains were close to the lake shore, which occupied a depression some forty miles long with an elevation of 15,448 feet. At this time of the year the lake was still covered with a crust of ice, except its southwest corner, which with its dark blue surface formed a sharp contrast to the white expanse of ice on the lake.

While descending from the pass, we met a group of horsemen escorting four men on foot with their hands tied behind their backs. These were robbers who had recently murdered a party of traders and were now being taken by the local tribal militia to the dzong . One of the militia told us the story of the outrage. The traders were staying for the night in a nomad tent, and while asleep, were brutally murdered with swords by the nomads of the locality. All their goods were taken away. The robbers were unable to keep the affair secret and the brigandage was discovered. The local authorities dispatched a detachment of militia to arrest the murderers. This was soon accomplished and the criminals were sent to the dzong to suffer an adequate punishment. Such cases were reported to take place frequently along the route, for the nomads easily and rapidly turned into robbers. The usual victims are traders and this accounts for the fact that all of the rich caravans have an armed escort.

We camped for the night at the foot of a spur situated to the southwest of a valley open toward the Chyaring tsho. The place, which was a relay station for dispatch riders, was called Ngan-dzom and was fifteen miles from Shentsa dzong. It was famous for strong winds. This reputation was soon justified for about noon a southwest wind started to blow and turned into a real storm with clouds of sand and loess dust blowing across the valley toward the lake. Sir Henry Hayden camped on the same place during his exploration in 1922.

March 30. The trail followed a small valley occupied by several nomad encampments. Herds of domestic yaks and flocks of sheep were seen grazing in mountain glens and on the slopes of the mountains. From the Tsung La, a mountain pass of some seventeen thousand feet, we obtained a fine view of the snow massif Shentsa Jyel-khang, which stood dominating the landscape on the southeast. Descending from the pass, the path followed for about three miles the crest of a range stretching westward, and then crossed two more ridges and entered the basin of the lake Sera tsho, a small lake about half a mile long and a quarter of a mile broad, situated in a flat depression amid the surrounding mountains. Numerous marine fossils were found on the shores of the lake.

The route followed the northern shore of the lake and then entered a broad gorge surrounded by limestone hills. We camped on a plain watered by the Paro tsang-po, a river that empties into the Chyaring tsho. The place of our new camp was called Kar-tsang. The plain had a rather large population and herds of domestic yaks and horses were grazing at the foot of the surrounding hills. As usual on the Tibetan uplands, the best grazing lay along the foot of the mountains.

At Kar-tsang, we received an unexpected reception. Eight big tents were erected and two hundred and sixty yaks and twenty riding horses were held ready for our expedition. A large crowd was in attendance. On inquiry as to the cause for such a reception, we were told that the messenger from Shentsa who brought the official letter announcing our arrival, had stated that we were in need of eight big tents on each stage and three hundred pack animals!

Toward evening our camp was visited by traveling singers, a man and a woman with a child. The man played on a stringed instrument, and the woman sang, while the child danced.

March 31. A fine morning, warm and windless. We crossed the Kar-tsang Plain toward the Paro tsang-po, which flowed along the foot of the hills, forming the western rim of the plain. After crossing the river, the trail entered a narrow mountain valley and ascended toward a low col, called the Lam-lung La. The summit of the pass was crowned with several stone cairns with the usual multicolored prayer flags. The descent was gradual and led into a broad plain surrounded by low, limestone ridges. On some of the loess-covered patches of ground we found grass well preserved under the snow.

The descent was made exceptionally trying by the bitter cold, western wind that blew across the plain and with terrific force struck the western slope of the pass. The horizon was covered by thick clouds of loess dust, and a yellowish mist hid the outlines of the mountains that rose south and west of the plain. The camp was prepared on a very bad spot, open to all winds, but there was no possibility of transferring it to another place. The spot bore the name of Chu tra-ri and was known for its muddy and brackish water.

April 1. We decided to make a double march this day, that is, to combine two short stages and to change yaks and horses at the end of the first stage. We started early, at six o'clock and crossed the Chu tra-ri Plain, heading southwest. The track was even and good. After some twelve miles over a gravel and loess-covered plain, we halted for an hour at a nomad encampment to change the pack yaks and riding horses. So far we continued to follow the route of Sir Henry Hayden in 1922. From Gon-khyok, the name of the encampment, our route left his and struck the southern shore of the lake Phung-pa tsho, whereas the route of the British explorer skirted the northern shore of the same lake.

The change of pack animals was effected with remarkable quickness and in about an hour the caravan column was again in marching order. The trail continued across a series of sandy ridges. On the northern and western horizon rose high, snow-capped latitudinal ranges. After another eleven miles, we camped on the southeastern shore of the lake Phung-pa tsho, the Daru tsho of Nain Singh and the Murchu tsho of Sven Hedin. Nain Singh passed north of the lake and Dr. Sven Hedin passed west of it. Sir Henry Hayden passed the lake along its northeastern shore on the way to Wang-po on the northern shores of the Dangra yum-tsho. The Phung-pa tsho is connected by a very narrow channel with another small lake lying west of Phung-pa tsho, and called Majyar tsho. Both lakes were frozen and it was impossible to ascertain whether there was any flow from one lake into the other. The place of the camp was called Mur-chen and was inhabited by nomads all the year round. The local nomads are a little better off and possess fine herds of yaks and horses. The yaks particularly struck us as being much larger and stronger than the animals we saw at Nag-chu-ka or at other places of the Nag-tshang district.

April 2. We started early again, in order to reach the shores of the Ngantse tsho on the same day. The track crossed the plain of Mur-chen and then led across a succession of low, sandy ridges that fell steeply toward the Majyar tsho. The weather, which was sunny and fine in the early morning, turned into a bad snowstorm, and progress was made difficult by biting wind and wet snow. Our local guide, who was unable to see the track ahead of him, lost his way and we wandered for some time amid the howling of the storm until the wind and snow suddenly abated and the dense white fog that surrounded the caravan column lifted. We found ourselves standing on the crest of a mountain spur stretching toward the Majyar tsho. From this place we easily found our way into a narrow, sheltered valley, where our next stage was situated, a group of encampments called Pong-chen. We found all the relay animals ready and were able to continue our route at once.

A narrow valley led toward the Pong-chen La, a pass with an elevation of 17,621 feet. The ascent was very steep and stony. Our poor camels had an exceptionally hard time on sharp stones. Unfortunately, a cold, western wind and wet snow started again soon after we had reached the summit of the pass and the storm continued until we had descended into a broad plain in which lay the lake Ngantse tsho, which was clearly seen from the pass.

On the pass we crossed the route followed by Sven Hedin on his way to the Brahmaputra Valley. We camped near the house of a local headman, a miserable stone structure with filthy courtyard. The place bore the name of Chi-buk de-nga. Game was rather plentiful in the plain. We observed herds of kyangs , gazelles, and flocks of snow cocks. Toward the south of the plain rose several snow-capped peaks of an unknown range. The basin of the Ngantse tsho represents a vast plain covered with grass land, swampy in many places. Like most of the lakes of the Tibetan highlands, the Ngantse tsho has considerably contracted in size and Sir Henry Hayden thinks that the three lakes, Phung-pa, Majyar, and Ngantse once formed one much larger lake, filling the whole of the vast valley. The lake was thoroughly studied by Sven Hedin during his memorable journey in January, 1907.

April 3. The track crossed the plain south of the lake. After several hours' march across the gravel and loess-covered plain, we camped for the night at the foot of a low, sandy ridge not far from the southwestern shores of the lake. It was called Nga-tam-chen, and has a small population of nomads living in tents, pitched at the foot of the hills. The local population has almost no knowledge whatsoever of the neighboring regions. They know about Nag-tshang to the east and Chok-chu to the west. An old man remembered two Europeans who had visited the northern shores of the lake, probably Dr. Sven Hedin or Sir Henry Hayden. Our camels were the first that ever reached the place and created an unusual stir among the nomads. Crowds assembled to see the strange animals enter the camp.

April 4. We started at six o'clock notwithstanding some difficulties with relay horses which were placed at our disposal. Half of the animals were quite wild. The others hardly stood on their feet, being exhausted after a snowy winter. The morning was exceptionally warm, the thermometer registering +12°C. The horizon was hazy, and thick mist hung over the lake and the mountain chains to the east of it.

The trail which we were following led southwest across a complicated system of ridges. Toward the north lay a broad latitudinal valley. The caravan of pack yaks with the heavy baggage followed the valley. All of the horsemen with the detachment of yaks carrying the camping outfit, followed the much shorter route along the crest of the ridge. After seven miles' march, we reached a low pass with a summit marked by stone cairns and mani-walls. The view was wonderful and we could see the whole of the vast expanse of the holy Dangra yum-tsho, one of the most beautiful lakes of the Tibetan highlands. Straight toward the south rose the imposing snow-capped meridional range of Tarkö La-jyap or Tarkö kang-ri. Dr. Sven Hedin has proved that the range should be considered as an isolated massif. The range continued north as far as the gap in the ridge, through which passed the route to Chok-chu. From the summit of the pass, the track turned southward, and suddenly, as is often the case with Tibetan roads, became a broad highway lined on either side by mani-walls. The ground was sandy and soft and our animals rapidly trotted toward a group of nomad encampments, which were said to be our destination.

At the foot of the pass a narrow strip of swamps had to be crossed. The camp was pitched for us on a small square patch of level ground a mile distant from the southwestern shore of the lake. The place was called Lo-wa na-mo and was the seat of a local headman, an imposing man with very civil manners.

The Dangra yum-tsho, which was first discovered by Nain Singh during his splendid journey from Ladak to southwestern Tibet and Assam in 1873-75, is forty-five miles in length by twenty-five in breadth at its widest part. This lake has also considerably contracted in size. The existing erosive terraces clearly delineate its former large basin. Nain Singh was of the opinion that the smaller lake of Thang-yung to the north formed one lake with the Dangra yum-tsho. The lake receives a large volume of fresh water from the Tarkö tsang-po, a river that flows in a terraced channel, with terraces twenty-five to thirty meters high. A probable explanation of these terraces is that the contraction of the lake resulted in a contraction of rivers, which deepened their course.

We were the first Europeans who actually camped on the southern shores of the lake. During his expedition of 1907, Dr. Sven Hedin reached a point three days' journey south of the lake, where he was stopped by local nomads acting under orders of higher authorities. Nain Singh, who discovered the lake, visited only its northern shore at a place called Wang-po (dBang-po).

The valley of Wang-po has a sedentary population. The place being somewhat lower than the surrounding uplands, barley is cultivated. According to a local tradition, recorded by the Pandit, many decades ago the valley of Wang-po was a powerful kingdom. The local chief had his residence at Khyung dzong, the ruins of which are still visible. He was overcome by the king of Lhasa. In 1921 Wang-po was visited by Sir Henry Hayden, engaged in geological survey work for the Tibetan Government. According to Nain Singh, the population of Wang-po belonged to the Bön-po sect. The population to the south of the lake was Buddhist.

Lo-wa na-mo, where we were camping, was a group of nomad encampments. Some of the families lived there permanently, and the tents were encircled by walls made of stones and sod to protect them from the spring and winter gales. There were no traces of cultivation. Some of the men present in the camp knew well the routes to the Tsang Province, by which Dr. Sven Hedin tried to approach the lake. I even found a man who had been to Nepal, and had a smattering of Hindustani.

It was difficult to gain reliable information about routes to Saga dzong. In our Lhasan passport, it was stated that the authorities of Shentsa dzong were supposed to give us transport until Saga dzong, but we soon discovered that the two representatives of the dzong knew nothing about the route to Saga and were planning to leave us at Chok-chu. The local headman told us that he was not at all sure that caravan animals were ready for us at Chok-chu and we therefore resolved to dispatch one of the Shentsa representatives to Chok-chu the same evening to make all the necessary arrangements. He rode off on a fresh horse, accompanied by a local guide. The men had received instructions to ride in all haste and to arrange for another representative from Chok-chu to accompany us to Saga dzong.

I knew that a southern route branched off from Lo-wa na-mo to Saga dzong and crossed the Tarkö La, lying to the southwest of the lake. This route was shorter by ten days and had the great advantage of passing through a populated region. It was officially known under the name of lho-lam or the southern highway, and the nomads along the route always kept ready a certain number of pack yaks and riding horses for the use of the governmental ula or relay system. I tried hard to persuade the local headman to send us by this southern route, which offered great advantages to us and to him, as he would have to supply us with animals only for one stage. He agreed that this would be the right solution for both parties but as he had direct orders from the authorities of Nag-tshang to send us to Chok-chu, he was quite unable to change our route. He was inexorable in his position and although I pressed him hard, he would not give up his decision to send us by another route.

The day continued to be calm and warm, the thermometer registering +13° at two o'clock. We observed flocks of wild geese and teals on the numerous water pools in the swamps east of the camp. It was the first spring-like day and the surrounding mountains were already assuming that deep blue and purple coloring of the mountainous regions of inner Asia in springtime.

April 5. Again a fine and wonderfully clear day. We started soon after the rising sun had illumined the snow peaks of the Tarkö kang-ri. It was an unforgettable spectacle, a foretaste of the grand mountain scenery of the Trans-Himalayas. The track followed the southern shore of the lake. The ground was sandy and ravines intersected the plain south of the lake. The Tarkö tsang-po, which empties into the lake, was just breaking up and was covered by masses of floating ice, which struck and cut the animals' legs. The camels had a hard time at the river. They have a natural aversion to water and we had great difficulty in bringing them safely across the swift stream. On the opposite bank of the river lay a vast plateau, stretching at the foot of the Tarkö kang-ri massif and covered with juniper, the first shrub we had seen in six months of travel over Tibetan highlands.

The track crossed the plateau in a northwesterly direction. We camped for the night at the foot of the range on the southwestern shore of the lake. A clear stream of water descending from the mountain side, fed by glaciers and thawing snows, supplied the drinking water for the caravan. The camels enjoyed the salty water of the lake. The site of the camp was known under the name of Thu-gu lap-tse, and its chief defect as a halting place for caravans was the total absence of grass, except juniper shrubs which were good for camels, but not for horses. The day was exceptionally warm and the thermometer registered +28°C. We sat until late into the evening, admiring the unique panorama of the lake and the surrounding mountains, lit by the rays of the setting sun.

April 6. We broke camp very early, at six o'clock, for the stage had to be a long one. The track followed the southwestern shore of the lake. At some places it was difficult going, especially for camels, as the trail climbed rocky cornices hanging above the lake and was so narrow that the loads had to be taken down and carried by the men, while the caravan animals were led, one by one. This all delayed us. To the west rose the northern continuation of the Tarkö kang-ri and the trail led over a vast, sandy slope, gradually descending toward the lake. After some four miles, we passed a small lake about a quarter of a mile long. Dark clouds enveloped the snow peaks of the Tarkö Mountains, and about noon a violent western wind with wet snow began to blow, making our progress very trying. We halted for the night in a high mountain valley with scanty grass, called Chumgo mar-po. Eleven tents were pitched for our use and the local people assured us that they had spent seven days on this spot, waiting for us. The yak caravan arrived only late at night and we sat a long time on our saddles, which were placed on the ground, while the storm howled around us.

April 7. A dull day with heavy gray clouds on the horizon. The trail crossed the low, sandy ridge west of the valley of Chumgo mar-po and then led across a succession of sandy ridges, heading northwest. After a three hours' march, we reached a vast plain, and halted on the shores of a small lake. Here the headmen of Chok-chu had prepared a camp and brought pack animals. We seldom saw such degenerate types, robbers by nature, and liars by conviction. The chief headman, an indescribable creature, assured us that they had prepared everything for us along the route to Nga-ri dzong on the Ladak route, but that they had no information about the route to Saga dzong, by which we were now to travel.

The route to Ladak was a jya-lam or highway, but that to Saga was only a local trail. According to the headmen, it was very wrong of the De-wa-shung to send us to Chok-chu for there was no ula from there to Saga. We should have to wait for the arrival of the Tarkö garpön who would decide about the route to Saga. I had a sharp conversation with the headmen, and told them that we would start tomorrow and that they would have to accompany us to get animals on the next stage, and that unless this was properly done, we would have to complain about them to Lhasa. I added that their professed ignorance of the route was just an excuse for they were bound to know the road to the dzong, because Chok-chu paid the annual tax to Saga dzong. After much talking and arguing, the headmen finally agreed to escort us to the Tarkö garpön. I tried to get some information about the local country, but the nomads had a very limited knowledge of it and I succeeded in finding only a single old woman who remembered having been on pilgrimage some thirty years ago to the Lake Manasarowar and the holy places on the Nepal border.

The information she gave me was far from being reassuring. According to her the country to the south of Chok-chu was highly mountainous and Saga dzong was a big settlement, with a thriving trade and shops where Nepalese, Indian, and Lhasan goods were sold. I doubted the correctness of her statement, for Captain Rawling and Sven Hedin, who had visited the place, gave quite a different description of it. The old woman had probably visited the place during the annual autumn fair, and had seen Nepalese and Lhasan traders with their temporary shops, attending the fair.

April 8. It was about seven o'clock when everything was in marching order and we were able to start the ride southward across the vast Chok-chu plain toward the mountain some six miles distant from our present camp. Our conversation of yesterday probably had some effect on the headmen, for they all appeared in the morning and escorted us personally. According to them, there were two routes to Saga dzong, one through Tarkö, the other across the high Sangmo Bertik Pass.

The vast latitudinal plain of Chok-chu is an intermontane plain for the most part covered by sand and loess, which is covered by good grazing during the summer months. On our way we passed herds of kyangs and Tibetan gazelles. The chief headman rode ahead of us to inform the garpön of our coming. After a two hours' ride, we were suddenly stopped by the headmen at a group of tents not far from the foot of the mountains. The headmen explained their action by the fact that the garpön had sent a messenger asking us to stay there for one day, as he was coming personally to see us.

The garpön of Tarkö arrived in the afternoon. He was a young man of about thirty, tall and well built. He was attired in the usual puru coat of the wealthier class, and wore a Tibetan fur cap, bound with a red turban, following the Kham-pa fashion. A long sho-lang, or saber, ornamented with turquoise and coral, was fixed to his saddle. A white tent with ornaments sewn on it was pitched outside our camp and was used as temporary headquarters for the district chief.

A preliminary conference was held by the garpön with the local headmen and our Tibetan guide. After an hour, the garpön requested to see our Lhasan passport and announced his desire to visit us. He was received in my tent and expressed his deep regret that we had been sent by the present route and thus delayed by three days. According to him, the proper route for us was the "southern route" which started from Lo-wa na-mo on the southern shore of the Dangra yum-tsho. This route crossed the Tarkö La and led to the headquarters of the garpön, a place called Tarkö La-jyap. From there it went across the Pendong La, a pass situated in the same range as the Dong-chen La, to the Nak-po kondro La, situated in the main range of the Trans-Himalayas east of the Sangmo Bertik Pass. The range in which the pass is situated bears the name of Nak-po kondro kang-ri, the Nak-po gongrong gang-ri of Sven Hedin's map.

As we were already in Chok-chu, the garpön thought it advisable to let us go by the Sangmo route, which crossed on its way to Saga the following passes: the Dong-chen La, the Sangmo La, the Gyegong La and the Tsuk-chung La. This route had been traversed and surveyed by Dr. Sven Hedin in 1907. We at once perceived that the real object of the garpön was to send us through the territory of his neighbor chief, the headman of Bum-pa chang-ra (the Bong-ba of Sven Hedin's map), and thus escape the necessity of furnishing us with caravan animals and other supplies along a difficult and deserted mountain route. I succeeded in making him promise to accompany us personally to the encampment of the next headman and arrange with the latter about caravan animals. The garpön hinted that our passport was granted from Lhasa and we therefore had the right to travel through the territory subject to the Government, whereas his territory was church land.

The location of our camp was known under the name of Kang-ru. The pack animals arrived in small groups driven by wild looking fellows, who could hardly understand each other, speaking different nomad dialects. Our Lhasa Tibetans had great difficulty in making themselves understood by the local population.

The Chang-pas of the Trans-Himalayas strike the traveler as being of primitive and inferior physique. The hard life at high altitudes, the grazing valleys of the northern slopes of the Trans-Himalayas having an average elevation of sixteen thousand to seventeen thousand feet, affected the physique and intelligence of the nomads. They are a wild, dirty looking crowd with long, matted hair that blows out in the wind. The local nomads are round-headed with prominent cheek bones. We very often observed primitive types with retreating foreheads and projecting glabella. The lips are of medium thickness. Most of the men are of slender build with rather weak limbs and sunken chests. The hair growth is scant and I don't remember having seen bearded males. The hair when cropped is woolly. The laymen usually wear their hair plaited in many small tresses that hang over the forehead and down the back. The families are small and children are seldom seen in encampments. There is a clear impression that the population steadily decreases. Our doctor observed frequent cases of rachitis, especially common among the children, and scurvy among the adults. Scorbutic conditions of gums are almost general.

The costume consists of the usual sheepskin coat, worn on the bare body, and the usual nomad boots, made of homespun cloth and held under the knee by a garter. Head coverings are seldom used. In winter the fur-lined bonnet is commonly worn. The men carry swords and matchlocks. Modern magazine rifles are seldom seen in the possession of the nomads. The population is said to be addicted to brigandage and such was the impression of Dr. Sven Hedin. During our passage through the district we did not encounter any robber bands, although the existence of a tribal organization of robbers was quite possible in those out-of-the-way places.

The Trans-Himalayan nomads dwell in the usual black tents and live on the produce of their herds, which consist mostly of sheep and yaks. Horses are kept only by well-to-do families. The summer months are usually spent in some of the valleys of the Trans-Himalayas, but in winter the cattle are driven to the north, where there is less snow and grazing is better.

On the northern uplands, the mountain valleys and low hills are covered by sand and loess and afford good grazing grounds, whereas the valleys of the Trans-Himalayas are mostly covered with detritus and the grass is scant, even after the period of summer rains. During our march through the Trans-Himalayas we experienced great difficulties in finding sufficient pastures for our animals. Most of the country was completely barren and the scant level patches of ground were covered by dead and weathered grass.

It was by no means an easy country to cross, not only because of the natural obstacles, high mountain passes, and barren valleys, but chiefly because of the strange distribution of districts, which were separated from each other by strips of land belonging to other districts in a most perplexing way. Chok-chu belongs to Nag-tshang but pays its tax to Saga dzong. Kang-ru and the mountain country south of it are under the jurisdiction of the Tarkö garpön, who is appointed by the Tashi-lhun-po labrang. Nam-chen and Nga-mo dang-khang on the southern shore of the Ting-ri lam-tsho are again under Nag-tshang dzong, whereas the next stage, Le-kar and Lap-sa-ru, belong to the Bum-pa chang-ra Province and are under the jurisdiction of Nga-ri dzong. The valleys of Tsuk-chen and Tsuk-chung belong to Saga dzong.

April 9. This morning a thick, white mist hung over the valley of Chok-chu and hid the summits of the surrounding mountains. Snow had fallen during the night and the mountain slopes were covered with it. We crossed a sandy ridge and entered an intersected mountain country, a labyrinth of mountain valleys, a country which stretched as far as the Brahmaputra and represents one of the highest watersheds on earth. A mountain valley in which several encampments were situated ran southwest. A small but deep stream ran down the valley. The stream had already broken up. Only a narrow strip of ice remained along its banks. The bottom was muddy and we had difficulty in finding a suitable landing place. We tried several times and finally succeeded in finding a place which looked fordable. We followed a high mountain valley called Ugpa occupied by encampments. The garpön rode ahead of us, and we could see him galloping from tent to tent collecting animals for us.

On our way we were joined by a number of menservants of the garpön. All of the men had matchlocks slung across their shoulders and wore red turbans. Very few of them had been to Saga dzong and only a single man remembered having once visited the place. The valley we were following rose gradually until it suddenly broadened and formed a small circular valley, well sheltered by mountains. The place was called Nam-chen and a tent was prepared for us by the local headman. The poor man was in despair because he would have to supply us with caravan animals to Bum-pa chang-ra. According to him, we had to go by the southern route across the Nak-po kondro La. We left the garpön and the headman to decide the question of caravan animals between themselves. It was useless to sit and witness the arguing of the two men. After an hour the garpön and the headman emerged from their tent and came to tell us that the headman of Nam-chen had undertaken to transport the expedition to Bum-pa chang-ra and that the garpön would help him in this, supplying part of the pack animals.

April 10. We delayed our start because of the pack animals, which reached the place only at about eight o'clock. Some of the new relay horses were so wild that they broke loose and stampeded and had to be brought back. After leaving camp, we crossed a low pass called Mari-trang La (17,300 feet) and descended into a broad, sandy valley, the basin of the salt lake of Ting-ri lam-tsho (the Teri-namtsho of the maps). We skirted the southeastern shores of the lake. To the south of it rose basalt rocks. In former times the lake occupied a much larger basin, but in the course of time had contracted. The plain southeast and south of the lake is the dry bottom of this former large lake. The lake is salt with layers of salt crust on its shores. We observed flocks of aquatic birds, turpans, teals, Brahmini ducks, and seagulls. The air was filled with their cries and flocks of wild geese were continually passing northward.

We found a camp pitched for us at a place called Nga-mo dang-khang, south of the lake in a valley inclosed by high rocky ranges from the south and open toward the lake. Toward the southwest rose the range of the Drong-chen La. On our way to the camp, we discovered that the official letter which was supposed to be traveling ahead of us, had reached the place only two hours before, for the messenger had spent the night in an encampment and thus delayed the delivery. It was almost impossible to obtain a new dispatch rider as the population had no horses. Toward evening the headman of Nam-chen secured another dispatch rider and sent him off with the letter toward Bum-pa.

The local population was exceptionally wild looking and had never seen a white man before. They wondered at our camels and insisted that we should give them some camel hair to be put into their charm boxes, for according to their belief, a hair from such a strange animal is sure to be a very potent talisman against illness and other danger.

Toward the northwest of the lake there is a monastery, Mendong gompa, belonging to the Karma-pa sect, which was visited by Dr. Sven Hedin during his passage in 1908. The afternoon was very warm, the thermometer registering +32°C.

April 11. The promised caravan animals did not reach the camp and we had to ride on our own horses and leave Portniagin in charge of the baggage to await the arrival of all the pack animals. We crossed the valley in a southwesterly direction and entered a broad mountain gorge that led toward the Dong-chen Pass. After a ten-mile march, we camped at the foot of the pass. The population was extremely poor and we experienced great difficulty obtaining grain for our horses and camels. For the last three days we had obtained only one small bag of grain containing some fifteen pounds of barley. An order was given to feed the camels and horses with tsam-pa and to give to the camels an extra meal prepared of Chinese flour. It was absolutely necessary to reach Saga dzong and to replenish our stores. We had been cutting down the rations for men and animals and our supplies could only last for another seven days.

The amplitude in temperature was very wide. In the morning before sunrise the thermometer registered -13°C. and at two o'clock in the afternoon we had +40° in the sun.

April 12. We started earlier than usual and after an hour's climb reached the top of the Dong-chen Pass. From the summit we beheld a unique panorama of the broad mountain valley southwest of the pass, sheltered by high eroded massifs of limestone. We followed the valley and soon reached the broad intermontane valley of the Sü-tsang-po, the Soma tsang-po of the maps. The camp was pitched on the left bank of the river, which flowed in several narrow channels. The Sü-tsang-po had its source somewhere in the Tarkö kangri and emptied into the lake Ting-ri lam-tsho. The river bed was flat and shallow but the river is said to carry a large volume of water in the period of floods. The local headman was unable to supply the required number of pack animals for all the pack yaks of the district had been driven to far distant pastures. He agreed to hire the yaks which came with us from Dang-khang. The place was known by the name of Le-kar and was still under the jurisdiction of Nag-tshang. The local headman said that some three days southeast of Le-kar there was a wooded valley with plenty of game. The day was cloudy and warm, the thermometer registering +14°C. in the afternoon.

April 13. A fine but cold morning, with a temperature of -5°C. We crossed the plain in a south southwesterly direction and ascended a low, sandy spur. From here we traveled over an intersected country of valleys and gravel plains stretching at the foot of the range, south of the route. We had to halt at a place called Sangmo nga-dum, for the local headman had made no preparations to meet our caravan. There were no tents pitched, or fuel stored for us. We had to camp near a miserable tent of some shepherds and sent a man to the headman, requesting him to appear in person to discuss the situation and comply with the orders from Lhasa. Meanwhile we had to arrange our camp as best we could, collect fuel, and break the ice in the rivulet for drinking water.

The headman appeared with tents and fuel supplies only late in the afternoon, and professed utter ignorance of our coming. He did hear of an official letter announcing our arrival but thought it was not addressed to him but another headman, beyond the mountains! He could supply us with only thirty yaks, the remaining twenty we had to obtain from the Nam-chen headman, who fortunately agreed to go on for another two stages. The whole of the afternoon was spent in negotiations, which take a long time in Tibet and leave but little opportunity to attend to other needs of the caravan.

April 14. We made an early start in spite of the initial difficulty with pack animals. Fresh animals did not reach the camp in time, and Portniagin had to remain again with the baggage to wait for the yaks. At the last moment the Sangmo headman succeeded in supplying us with the required number of yaks.

Leaving the camp we followed the gorge that gradually ascended toward the high pass of Sangmo. Basalt massifs rose on either side. Kyangs grazed in the gorge on the scant grass that grew along the banks of the tiny mountain stream. The track went over hard, stony ground and our camels had a hard time walking over the surface.

After a six hours' march we pitched our camp in a sheltered valley close to a small brook of fresh water at the height of 17,600 feet. The yak caravan arrived only late at night. Gray clouds hid the summits of the mountains, and the Tibetans feared a snowfall during the night, that would render the pass difficult. The yak drivers assured us that some two weeks before the pass was completely blocked by snow, and that even now much snow lay on the summit. It was impossible to trust local informants, and we would have to find out the condition of the pass for ourselves. Everybody felt somewhat excited for the next day we would cross the highest pass on our way to Saga dzong.

Although the existence of a mighty mountain range north of the Tsang-po had been surmised by a number of explorers, and d'Anville (1733), Brian Hodgson and others had marked on their maps a problematic range that cut central Tibet from west to east, Dr. Sven Hedin was the first to survey the range and cross it by at least eight tremendous mountain passes with an average height of eighteen thousand feet, during his memorable journeys of 1906-8. The range forms the watershed between the Indian Ocean and the inclosed self-contained drainage region of inner Asia. To the west the Trans-Himalayas merge with the great Karakorum and it is left for a future explorer to investigate the meeting point of the two lofty ranges. The explorations of Dr. Sven Hedin have shown that the Trans-Himalayan system belongs to a comparatively recent geological period. The structure of its ranges shows that the formation of the system was attended by volcanic activity and the presence of numerous hot springs in the region proves the recent disturbances.

The Trans-Himalayas consist of a series of ranges, of which the principal one seems to be that in which the Sangmo La (19,094 feet), the Se-la La (18,064 feet), Khalamba La (17,200 feet), Go-ring La (19,587 feet), and the Shang-shung La are situated. The southern slopes of the ranges are cut by deep eroded valleys, traces of the activity of the Indian monsoon. The northern slopes merge with the undulating highlands of northern Tibet and receive much less precipitation than the southern slopes. On the northern side, the snow line lies at about eighteen hundred feet and glaciers never reach the bottom of the valleys.

April 15. The night was quiet and snowless. We rose very early and by six o'clock the caravan column began its ascent to the pass. The northern slope was covered by huge bowlders, carried down the mountain slopes. In some places snow was still lying on the ground, and our horses frequently sank deep into it and had to be rescued.

After three hours of continuous climbing we reached the summit of the pass (19,094 feet). Several stone cairns stood by the roadside. From the summit a fine mountain view, of grandeur and severity, unfolded before us. A broad mountain gorge led to the south. To the west and east the view was closed by meridional ranges, descending toward the Lap-sa-ru Plain, south of which rose the snowy massif of the Kang-chung kang-ri. The day was dull and heavy clouds and mist hovered over the ranges. A steep but short descent, during which we all had to walk, brought us to the gorge. On the narrow trail our horses had a hard time among bowlders and pools of frozen water. Men and animals frequently slipped for there was hardly a place for one's foot. While descending from the pass we observed a herd of wild yaks on one of the slopes. They were climbing a precipitous, almost vertical mountain slope, with each step sending stones and clouds of sand down into the gorge. For a moment, their dark silhouettes stood out on the transparent background of the Tibetan sky, then they disappeared behind the rocky wall. It was a picture of wild Tibet, which always haunts the imagination of anyone who has wandered across its unforgettable mountain expanses.

After seven hours of hard marching over a heavy, stony road, we descended into a broad latitudinal valley belonging to the Bum-pa chang-ra district. To our great surprise, nothing was prepared, notwithstanding the official letter which was said to have gone ahead. We sent out a reconnoitering party, which rounded up a couple of individuals who said that they were herdsmen, but refused any assistance. Our Tibetan argued angrily with them and it was decided to keep the two men in camp, pending an inquiry into the situation. The two local men finally agreed to give some fuel, and to send out a messenger to the local headman requesting him to come at once. Everyone assured us that they had heard nothing about our coming. Happily the messenger who had carried the letters was among our yak drivers and he certified that the letter was safely and duly handed over to the local district chief or yu-pön . We witnessed a hot scene between the messenger and the local nomads. As a precautionary measure, we decided to keep all the pack yaks and horses from Sangmo nga-dum until the local headman had produced the required number of caravan animals. The drivers protested but had to submit, for they understood our predicament.

The day was warm, notwithstanding the fact that the valley was inclosed on all sides by snow-capped mountains, with glaciers descending comparatively low into the valley. Some two and a half miles south of our camp lay the small salt lake of Lap-chung tsho. In the evening the setting sun lit the snows of the mountains, and the whole scenery plunged into a deep purple glow. The yu-pön did not arrive this evening and it looked as if we would have to stay a couple of days in the valley and search for the headman, who was trying to avoid us. Our supplies were coming to an end and it was imperative to reach Saga dzong.

April 16. A sunny, warm day. We were forced to stay the whole day and wait for the headman. The drivers from Sangmo tried several times to get away with their animals, but our guards prevented them from going back. We told the people that we would complain to the officials in the dzong about such a delay and open disregard of orders from Lhasa. The local people finally agreed to supply us with all the yaks in the valley and thus enable us to start the next day. We were to receive thirty yaks from them and the rest of the animals would have to be supplied by the Sangmo drivers. After three hours of negotiations a settlement was reached and we were assured of starting the following day. We also arranged with the owners of our relay horses to accompany us to Saga dzong and gave them extra pay for their services. The men were completely satisfied.

The place of our camp was called Lap-sa-ru and was inhabited by several families of nomads who bred yaks and sheep. In one of the tents we met two blacksmiths from the Tsang Province, who were working for the local nomads. The yu-pön did not arrive and we were greatly irritated by his conduct and deliberate ignoring of orders from Lhasa.

April 17. We started in the forenoon, for there was the usual delay with fresh animals and new drivers. We crossed the Lap-sa-ru plain in a southwesterly direction and skirted the Lap-chung Lake along its southeastern extremity. The lake is situated at an elevation of 17,037 feet. Tibetan antelopes were observed on its shores.

Not far from camp we came across another group of megalithic monuments. It was partly buried by drifting sand so that only the points of the menhirs forming the alinement were visible above the ground. I tried to excavate one of the stones, but it went deep into the ground and my excavations attracted the attention of the local people, who, as all Tibetans, are very concerned with not allowing the gods of the soil to be disturbed by excavations. I had to stop and leave the investigation for the future, when the Government of Tibet would sanction scientific excavations on its territory. What ritual was performed before these stone altars and cromlechs, we shall be able to tell only after a thorough search of the voluminous Bön-po literature. The great Bön collections of sacred texts, some three hundred volumes in all, still remain a closed book for us. Undoubtedly from some of the texts on rituals incorporated in them, we would learn the precise significance of menhirs, cromlechs, and alignments.

The yu-pön or local chief met us on the way. He was a huge fellow with matted hair, attired in a large, dirty sheepskin coat. He stuck out his tongue in the approved fashion, and reported that he had never received the official letter about our coming. He had no yaks at his disposal and was unable to help us. After a lengthy conversation, in which all the members of the expedition took part, we succeeded in persuading him to hire yaks from a caravan carrying salt. He agreed to do that and yaks were hired from traders with an understanding that the animals would have to go to Saga. Late in the evening the yu-pön changed his mind and decided to furnish us with his own animals.

We camped at the foot of a ridge, forming the southeastern rim of the Lap-sa-ru Plain. From here there were two routes leading to Saga dzong, a short one across the Lug La, and a longer one skirting the mountain range. The short route was reported to be still blocked by snow and we therefore decided to proceed by the longer route. The official letter sent from Lhasa was lost and no traces of it were to be found. I had to write another letter myself and seal it with our personal seal. A dispatch rider was sent off at once and to our amazement the new arrangement worked just as well as the old one. The place of our camp bore the name of Rong-se.

April 18. We started at eight o'clock this morning and crossed the low, sandy spur lying immediately southeast of our camp. From here the track followed an ascending valley, past a small lake still frozen hard. From the lake we turned into the valley leading to the Gyegong La, and stretching southwest. The valley is known for its numerous hot sulphur springs and is called Memo chu-tsen. Dr. Sven Hedin mentions these springs. He found their temperature to be 93.6°C.

The ascent toward the Gyegong La (18,012 feet) was rather trying. The valley at the foot of the pass was covered by deep snow, and we experienced some difficulty in finding our way through bowlders and snow. After four hours of tiresome climbing, we reached the summit of the pass. A grandiose mountain panorama unfolded before us—a sea of mountain peaks that stood sparkling in the sun. "Oh, God! There is no end of mountains!" exclaimed Golubin, dismounting from his horse and preparing for the descent. We all stood for a moment looking at this unforgettable scenery. The descent was steep and part of the way the track led over the ice of a frozen mountain torrent. In some places great blocks of ice obstructed the way. It was difficult for our unshod animals. For another four hours we went down the gorge. Frequently we had to cross the frozen stream over ice bridges. The country became more and more intersected.

We were approaching the peripheral regions of Tibet with their deep-cut mountain valleys and powerful streams. Nowhere was to be seen the dreary undulating landscape of the north Tibetan highlands; now there were instead rocky massifs, narrow mountain valleys, and towering ranges. On reaching a place where three important valleys met, we turned toward the southeast and camped for the night in a valley called Tsuk-chung. No traces of encampments or tents prepared for our use were to be seen. We sent out men to reconnoiter the locality and if possible to bring back some men, who could give us information about the local condition and furnish us with the necessary supplies.

After two hours of waiting, we noticed with our field glasses a human figure walking rapidly on the opposite side of the valley. The figure was moving toward us and proved to be an old man with an arrogant, unpleasant face, who at once started to tell us that it was none of his business to assist us and that the Government had no right to send us into the wilderness. We informed him that we had to get caravan animals at any price and that he had to furnish us with fuel and grain for our animals or bring the local headman to our camp. The old man was obstinate and said that he had no yaks, no fuel, and no tents to spare. Our Tibetans became very excited and threatened to arrest him and hand him over to the local authorities at Saga dzong. The man met their words with perfect composure and retorted that since everything had been taken away from him by the Government, he had nothing against losing his head.

We were relieved from the necessity of arresting the obstinate old man by the arrival of a tall, fine looking man, who was a former lama of the Chamdo Monastery in western Kham, and was now acting as local headman in the locality. He at once understood our difficult situation and ordered the recalcitrant old man to bring a tent and some fuel and moreover promised to talk with the yu-pön of Bum-pa about the caravan animals for the next day. Toward evening two tents were pitched and big camp fires lit the scene of the camp. The yu-pön arrived only at dusk but suddenly refused to camp with us and instead camped behind a low, sandy spur. He promised to come in the evening after sunset and talk the matter over. The headman of Tsuk-chung and our Tibetan guide considered his actions to be strange and were afraid that he would try to escape with his yaks during the night. We waited for him until ten o'clock in the evening, but as he did not come, we decided to send a man after him. The man soon brought back the answer that the headman of Bum-pa informed us that he had no business in our camp and that he was unable to keep his word and transport our baggage to Saga dzong.

No time could be lost, if we intended to proceed to Saga dzong without delay. To be kept in the desert without fuel and food supplies for men and animals, meant death to the whole caravan. We had to take intensive action to protect our interests and to insure the prompt execution of the governmental orders. I reported the situation to Professor Roerich and received instructions to follow strict measures for our protection. We decided to send the local headman of Tsuk-chung and our Tibetan guide to the camp of the Bum-pa headman and request him to appear in person in our camp. As the Tsuk-chung headman was afraid to proceed alone, half of our guard was sent with him in order to arrest the headman in case of necessity.

When our small detachment approached the Bum-pa headman's camp, we found him sitting at a camp fire surrounded by a circle of thirty armed followers. All rose to their feet, when the Tsuk-chung headman approached the fire. Our men were lined up a few feet behind him, ready for any emergency. The Tsuk-chung headman reproached the yu-pön for not coming to our camp to settle the matter in a friendly way and asked him to come and discuss the situation. Dead silence followed his words. A murmur rose from the crowd of armed yak drivers and suddenly one of them snatched out his sword. The nomads were apparently ready to resist with armed force. Portniagin, who stood near the man with the sword, suddenly pulled out his revolver and planted its muzzle in his face. This had a magical effect. The crowd started backward. Our men stood ready to open fire if necessary and the extreme tension continued for a few moments more. Then suddenly some one in the crowd ran back and the whole force stampeded toward the hills, leaving the yu-pön sitting at the camp fire. He was arrested and taken under guard to our camp where a general meeting was at once summoned inside the big tent of the Tsuk-chung headman.

Armed sentries were posted about the camp, for we had received information that the yak drivers were discussing the possibility of attacking the camp to recapture their headman. Two sentries were placed at the entrance of the tent which served as temporary courtroom. The yu-pön was brought inside and placed under a strong, armed guard. When everyone had taken his place, the Tsuk-chung headman once more reproached the Bum-pa headman for his arrogant conduct and asked him to close the matter amicably. The obstinate yu-pön once more refused to comply with our wishes and was silent to all our proposals. According to him he had nothing to do with the Lhasan Government, which was far distant from his district. I told him that unless he agreed to our demands, he would be arrested and handed over to the Lhasan authorities for proper punishment for his rebellious words and actions. I gave him ten minutes to think the matter over. The man lowered his head and sat for a while thinking, then rose from his seat and said, "No, no," and sat down again. He was accordingly declared under arrest and his hands were tied behind his back with a leather rope. Two guards were posted to keep watch over him during the night.

A strong guard remained in camp all during the night and the Bum-pa tribesmen who attempted several times to penetrate the camp were warned to keep away. Late in the evening the yu-pön agreed to supply us with caravan animals and to keep his word. His only condition was that we should set him free and forget about the incident. The man probably got frightened and understood that we were in earnest. He was therefore set free, but had to remain inside our camp for the night.

April 19. We started at seven o'clock and retraced our steps, until we had reached a narrow mountain glen leading toward the pass, Tsuk-chung La (18,000 feet). Not far from the entrance into the glen we discovered another cromlech with an alignment of stone slabs. As in Lap-sa-ru, the stones were half covered with drifting sand. The local headman had no knowledge of the stones and according to him they were so arranged by some lha or local deity.

The ascent to the pass was very steep and our poor animals had a very hard time. They often had to stop to regain their wind. From the summit one could see three valleys, the Tsuk-chung on the northeast, the Tsuk-chen to the west, and Saga dzong on the south and southwest. On the western and eastern horizon rose the mighty snow peaks of the Trans-Himalayas. Most prominent among them was the beautiful peak, Saga Jo-chung, which was clearly visible among the peaks forming the eastern continuation of the mountain range in which the Tsuk-chung Pass was situated. The descent was steep and all had to walk down the sandy slopes. A broad mountain gorge led into the basin of the Chorta tsang-po, a tributary stream of the Brahmaputra. Far to the south we could see in the hazy atmosphere the snowy outlines of the northern offshoots of the Himalayas, on the Nye-lam side.

The plain over which we were riding was covered by gravel. Grass was scant and cattle were nowhere to be seen. Saga dzong was hidden by a low, sandy spur projecting into the plain. We obtained the first view of the place from the top of the spur. Saga was a motley collection of dirty, stone hovels, and was built on a spot open to all the winds and gales of the Tibetan upland. Outside the village we were met by the local officials, who conducted us to a camp prepared for our use. Two tents were pitched southeast of the dzong and a crowd of local people and soldiers of the small detachment garrisoning the fort were in attendance. Notwithstanding the fact that the officials had received special instructions direct from Lhasa, nothing was ready and we were told to stay for three days at Saga to give the authorities time to procure the caravan animals. The Governor of the fort was in Lhasa and the place was ruled by his personal representative, the nyer-wa or treasurer of the dzong, who also was away on an official tour in the district. A messenger had been sent to him, requesting him to return at once to the fort. Meanwhile the petty officials resident in the fort refused to discuss the situation.

Our caravan reached the camp late at night and great anxiety was felt for our camels which were delayed on the pass. At about eleven o'clock that night three camels and their driver reached the camp and he reported that the fourth camel had died on the summit of the pass.