On April 13, 1927, everyone in the expedition rose before dawn. We hoped to start early in order to cover as much ground as possible. The first stretch of the route lay along the valley of the Tola River and according to our informants was comparatively easy for laden cars. The three trucks of the expedition were loaded the day before, and there remained the two Dodge touring cars destined for the members of the expedition and their light baggage which was to be tied on the running boards. By eleven o'clock everything was ready and the large gates of the expedition headquarters were opened to let the cars out. Friends, both Europeans and Mongols, assembled to bid farewell and Godspeed to the members of the expedition. The Mongols were chatting with their relatives, and lamas were seen murmuring their prayers and hanging ceremonial scarfs or khatags on the expedition cars. Our Tibetans were surrounded by a crowd of their countrymen who had come to see us off. The Tibetan personnel of the expedition was very proud of the rifles and insisted on riding with fixed bayonets, decorated with khatags attached to the rifle barrel. Our men looked very martial in their khaki fur coats and ammunition belts. The whole presented a colorful scene, in which were blended European and Asiatic elements.

Suddenly a disagreeable surprise was sprung by one of the motor car drivers. The man had forgotten to renew his Mongol passport and obtain a frontier pass for his car. Two weeks before our departure we had warned the drivers to have all their papers ready before April 13. Only two days before I had asked the men if they had done so and they all assured me that they had passports for themselves and frontier passes for their cars. The whole affair was very disagreeable for we had to postpone our departure until late in the afternoon and rush to the Government Transport Office to get the required passes without which the custom authorities would not allow the expedition cars to leave the city. There was no probability that the required motor car passes would be obtained in such a short time. The usual procedure took about three or four days. But thanks to the untiring help of our Mongol friends and the courtesy of the Mongolian Minister of Finance, they were issued and signed.

About five o'clock the expedition convoy of cars started on its long track across the western Gobi. For some distance it was accompanied by our friends and co-workers, Mr. and Mrs. Lichtmann of New York, and Dr. Jamtsarano, secretary of the Mongolian Scientific Committee, and other friends from Urga.

At a point beyond the Tola River bridge, we all alighted from our cars and bade a last farewell to the civilized world and our American friends who were to return by airplane from Urga to Verkhne-Udinsk.

"Greetings to all friends in New York! Await our future news!" were the last words of Professor Roerich, when the expedition column started westward along the valley of the Tola. The heavily loaded trucks with men perched on top of the loads moved on like huge monsters.

For the first five miles the route was good, notwithstanding the melting snow that made it muddy and slippery. On the seventh mile we ran into a thick patch of snow; one of our touring cars sank deep into it and had to be rescued. We all tried to pull it out, but met with little success. The hind wheels were buried deep. Meanwhile the three trucks, which delayed a little at the Tola River bridge chose another circuit route following the bank of the Tola River and were now some distance ahead of us. We fired several shots to attract their attention. One of the trucks stopped and men came running with spades and ropes to our assistance. It took some time before they reached the spot where the car was stuck. With united efforts we pulled it out and were able to continue the journey. It was getting late and we could hardly see the road ahead of us. About ten o'clock in the evening we decided to camp close to the river bank under some willow trees. These were the last trees we saw in outer Mongolia. The day's march was short, only some twenty miles, but we were all glad to be again in open country and feel the desert ahead of us.

After considerable hustle in the darkness, all the tents were pitched and we could look for a well-deserved rest after a tiresome day. Some six miles away on the farther bank of the river we could see the dim lights of the Sangin Veterinary Station. Southeast of our camp rose the western offshoots of the massive Bogdo ula and toward the southwest were the dim outlines of rolling hills, through which lay the valley of the Tola. The night was beautifully clear but bitterly cold, the thermometer registering some -25°C.

We broke camp on a fine sunny morning, with the surrounding mountains standing out sharply on the pale background of a northern sky. We intended to make a long march. The country presented a typical landscape of Mongolian grazing lands. Low ridges of rolling hills in soft undulating relief, covered by grass, were seen everywhere. The Tola River valley is from three to four miles wide. In many places flat terraces descend to the river. The road was generally good and the ground still frozen hard; but in a few places it was intersected by patches of sand, and we crossed one or two belts of sand dunes in which some of the cars stuck. For the greater part of the day's march the ground was free from snow, which lay in narrow side valleys and ravines. We met very few people on the way. Neither cars nor caravans interrupted the silence of the surrounding country.

Late in the afternoon we passed a couple of Mongol tents with some flocks of sheep near by. We camped for the night on an open plateau, situated on the bank of the Tola River. The day's distance amounted to over sixty-two miles. The night was again cold, the thermometer registering -15°C, and a piercing northwestern wind made the cold sharper.

On April 15 we again made an early start with the hope of reaching Nga-Wang Tseren küren, a monastery situated not far from the bend of the Tola River toward the northwest. The country presented the same landscape as the day before. Here and there the monotony of the flat grassy hills was broken by rugged masses of granite and porphyry. The road was easy and so even and straight that our cars could make considerable speed, especially on the last stretch of fifteen miles. So far we were following the Sain-Noin route, taken by the Third Asiatic Expedition of the American Museum of Natural History in 1922-23. The geology of the Tola Basin has been studied and described by Professor Charles P. Berkey and Dr. Frederick K. Morris, the geologists of the expedition, and it is unnecessary to give another description of a region visited many times by American and Russian scientists.

We were forced to camp a little earlier in the day than we had intended, close to a small ravine full of snow. Several Mongol cameleers whom we had met on the way reported that no water was available until we reached Nga-Wang Tseren küren, for the Tola River was still frozen and ran a considerable distance from the route. We had to thaw snow to get drinking water. The night was again bitterly cold, about -17°C., but the wind that had started soon after sunset abated toward midnight and the rest of the night was quiet.

On April 16 we drove for two hours over good gravel road before we saw the dugan or cathedral of the Nga-Wang Tseren küren on the farther bank of the river. Not far from the monastery we left the river valley and continued our journey across a low ridge—the easternmost branch of the Dolon-Khara Mountains. About noon we reached the important monastery of Tukhumun Dugan or Dugan-sümä, situated north of the salt lake, called Ikhe Tukhum-nor. The lake occupied a shallow depression about four miles long and three miles wide; it was still covered by ice and considerable layers of salt crust were seen on its shores.

The monastery is known for its du-khang or assembly hall built in Tibetan style with some good wooden carvings. In the winter months it is almost deserted, but during the summer the number of inmates exceeds 300. The monastery was visited by General P. K. Kozlov on his way to the dead city of Khara-khoto during his memorable expedition of 1907-8. We made only a brief halt in order to inspect the buildings and continued our route to Mishe-Gun.

After a two hours' ride over good road, during which we crossed several shallow rivulets, we reached the Mishe-Gun küren, a large lamasery with a colony of Russian tradesmen living near by. We halted outside the monastery and the settlement to have lunch, and after an hour continued our route southwestward across hilly country. Immediately southwest of the monastery is situated a low but steep pass, and it was only with considerable difficulty that we succeeded in bringing our heavily loaded trucks over it. We drove for some twenty-five miles more over flat ground and camped on the southern shores of two small salt lakes situated in a shallow depression surrounded on all sides by low hills. The ground around the lakes was very saline and in many places swampy.

The days were becoming decidedly warmer, and we observed several marmots that had ended their hibernation.

Next day we again struck an upland desert, imperceptibly rising toward the southwest where lay the mountain country of the Khangai. A cold piercing western wind blistered our faces and penetrated the heavy fur coats. The ground was stony and firm and the convoy of cars kept well together. We crossed a succession of rolling hills among which rose solitary weathered cliffs and sharply cut crags. In one of the valleys lying between two low ridges our first car, in which I rode, ran into a strange oblong object, wrapped in rags and deposited on the trail. I jumped out to remove it and found it was the corpse of a small child wrapped in weathered cloth and partly devoured by wolves. The nomad Mongols throw the bodies of their dead children on a nearby route, so that passing travelers may offer prayers for their souls. On the twentieth mile we passed a huge human figure of stone—one of those figures which abound on the steppe country of northern Mongolia and Jungaria north of the T'ien Shan Mountains. These figures probably represented funeral monuments of departed chieftains, which as a rule are represented holding in their hand a chalice with a burning flame—a sign possibly related to the ancient fire cult of the Mongols.

A local legend concerning the stone image ran that in ancient times there lived a mighty brigand chief, who inspired terror in the surrounding country and devastated the nomad encampments and their herds. When he grew old and infirm, he pledged himself to work for the good of his countrymen. In due course he died and his spirit was found to have been incarnated into the stone image. At present the figure is regarded as a kind of local protecting deity, guarding cattle and men. Libations of butter and offerings of colored cloth are made by local nomads and the rare travelers who happen to come this way. This story was told to me by our Tibetan guide, who had spent many years of his life in this region collecting Mongol offerings due to His Holiness the Dalai Lama after his residence in Urga in 1904.

Some miles away from the spot where we had found the stone image, we came across a small camel caravan which was traveling from Uliassutai to Urga by the southern route. We were very anxious to find out about our caravan with the supply of gasoline for the cars, which had started from Urga a month before us. The camel drivers were unable to give us any information for they had neither seen it nor heard of it. Evidently our caravan leader had chosen another route of his own or followed the northern route through the monastery of Erdeni Dzu.

Later in the afternoon we reached the northern branch of the Ongin-gol. Before reaching the river, the trail crossed a vast gravel plain with firm ground. The crossing was situated close to three Mongol felt tents and their occupants were eager to assist us. Our three trucks stayed behind, delayed by an accident with a tire, and we decided to cross the river in our touring cars and await the rest of the column on its farther bank. It flowed through a broad flat valley, and had several arms. Its bottom was covered with gravel and for the most part was firm. Only my car ran into a sand bank and had to be rescued.

The trucks did not reach the river until late in the evening. In complete darkness we could see their headlights rapidly approaching across the vast gravel plain. The trucks were unable to cross the river the same evening and had to camp on the other bank. We were left without tents and supplies—an unpleasant situation considering the cold of the night and possible snow- or windstorms. We finally decided to hire some horses from local Mongols and ride across the river to bring the necessary tents, bedding, and supplies. After some negotiations and arguing about the prices, eight saddle horses were brought into camp and we rode toward the river. It was by no means easy to make the animals enter the cold water. They snorted and pranced. After much exertion and shouting we crossed and loaded the horses with tents and supply cases. Some of us had to carry light things on the saddles. The horses shied and galloped away with their riders, one of our car drivers having an exciting time on his mount. My own pony kicked desperately while I was putting a tent bag on the saddle.

Late in the evening the camp was pitched and everyone retired to his tent. The night was clear, the wind having driven away the clouds that hovered over the mountain chains southwest of our camp.

The next morning was spent in bringing over the three trucks. Two of them crossed the river without any mishap but the third stuck in the sand and had to be rescued by riders, who pulled it out. About noon the whole of the expedition convoy assembled on the western bank of the river. We resolved to pay a brief visit to Udzen-Wang and to engage a Mongol guide who knew the route to Yum-beise küren. Our motor car drivers, who knew the road up to Ongin-gol pretty well, knew nothing about the remaining stretch to Yum-beise. An hour's ride brought us to Udzen-Wang, comprising a large lamasery, a local yamen or administrative center, and several Russian trading establishments, including a local branch of the Mongol Central Cooperative. While some of us were looking for a good and reliable guide, a big crowd of red-clad lamas, Mongol officials, and laymen assembled around our cars. They were eager to learn where we were going and what were our intentions.

After considerable waiting, the required guide was produced — a wrinkled old man of sixty, who wore a lama fur hat and some discolored yellow rags that served him for a coat. We were told that he was an experienced guide who knew the country well.

After the brief halt at Udzen-Wang we started again and crossed the gravel-covered plain to the south. The road ran through a mountainous country in a southwesterly direction. After a two hours' ride we climbed a steep sandy hill. The heavily loaded trucks had to be helped by all our men and a good deal of time was spent before all three of the trucks were safely brought over the pass. The descent was gradual and led toward a vast open gravel plain. We camped on it for the night, some thirty miles southwest of Udzen-Wang and not far from the main channel of the Ongin-gol. Luckily the evening and night were windless, otherwise it would have been difficult to keep our tents in position in an open space.

During the day we observed the first gray goose ( Anser anser ) and some turpans ( Casarca casarca ). During the night the camp was visited by a pack of wolves that kept howling in the neighborhood until dawn.

After crossing the Ongin-gol, which occupies a small portion of a broad valley, we again entered a hilly region, highly intersected and difficult for motor cars. It was surprising how our Dodges stood the rough tracks with frequent ascents and sharp turns on the very edge of a precipice. We were now in the southeastern hills of the Khangai Range, an important mountainous area of Mongolia that lies between the lake region of northwestern Mongolia and the basin of the Tola River. The central mass of the Khangai Range lies near to the city of Uliassutai, and is known under the name of Tarbagatai. Its highest peak, the Ochir-Wang, rises to a height of some 12,000 feet.

Some of the most important rivers of Mongolia, like the Orkhon, rise in the Khangai Mountains. These mountains are rich in former volcanoes. The southern slopes of all Mongolian mountains are as a rule rugged and rocky, often weathered and bearing traces of strong erosion effected by the climatic conditions and winds of the deserts that lie south of them. The northern slopes are often covered by grass and sometimes are forested.

After crossing the Tatsa-gol, an insignificant stream in its upper course, we continued our way along the southern foot of the Artsa ula Range. Not far from an old ruined Chinese stone hut, which was used as a storehouse for a Chinese trading company, one of our cars had a rather bad accident, and we had to stay behind to repair it.

While the driver was busy with the car, I made a short trip round the country and discovered a large number of ancient tumuli or barrows. There were eighteen large ones and six smaller ones. All of them had stone cairns on their tops and were surrounded by concentric rounds of stone slabs. These were probably the tumuli described by Professor Pozdneev in his work Mongolia and the Mongols , although the exact geographical position of the barrows was not indicated by the Russian scholar and explorer. Judging from their outward appearance, the tumuli were quite analogous to those of Noin ula Mountains, excavated by the Kozlov Expedition.

After a delay of two hours, we resumed our way and found the rest of the expedition encamped in a narrow mountain valley on the banks of a tiny stream. Not far from our camp passed the great caravan route Sair-usu – Uliassutai.

April 20 was a hard day for our cars and the personnel of the expedition. The only practicable route was a dim camel trail. Numerous sandy ravines made the journey very trying. The cars stuck frequently and had to be rescued by putting under the wheels planks of wood or large sheets of canvas sewn on felts, which had been brought from Urga for this purpose. After rescuing one car, we all had to rush and pull out another; and so it went until everybody was utterly exhausted and we decided to camp. Portniagin and I, accompanied by one of our drivers, went in search of a better road for the next day. We did not take our guide, for the man proved to be of very little use, knowing only the usual camel trail which was often impracticable for motor cars. We climbed several sandstone ridges in search of a route, but in vain. As far as eye could see there lay the same barren and hilly country, intersected by former river beds, full of bowlders and sand banks, and deep and narrow mountain valleys blocked by huge accumulations of débris. We had to return and postpone the search until the following day.

The next march was difficult and trying. The trail lay over a barren country of intersected sandy ridges and canyons formed by streams that once flowed from the mountains to the north. Soon after we left our camp, we met the first caravan coming straight from Tibet. It belonged to a wealthy Lhasan merchant, Kusho Kudrupa, one of those "state merchants" who traded in the name of His Holiness the Dalai Lama. He was accompanied by a number of Tibetan retainers and was going to Urga and Peking to buy silks, brocades, and other articles. The caravan men reported strong bands of brigands in the Gobi Desert north of An-hsi chou. Among the caravan drivers, we found two Kurluk Mongols from northern Tsaidam—a graybeard and a young man of twenty. For some reason or other they had decided to leave their native country and go to Urga. We observed their strange costumes and the picturesque shirts with turned down collars edged with otter fur. The men wore pointed hats and large leather trousers. They examined our cars—the first they had ever seen—but were doubtful whether we should be able to reach Yum-beise, for the route we were using was never before covered by cars or any other wheeled conveyance.

On our way we passed a large encampment of Mongols in process of pitching their camp. Several felt tents were already pitched and others were coming on camels. We saw long strings of camels marching through the valley. Some of them were covered with canopies under which sat women and children. A large crowd watched us while the cars were crossing a small rivulet with a muddy bottom.

There seems to be a perceptible difference between the type of Mongols of the grass country of northern Mongolia, and that of the Mongol nomads dwelling in the Gobi, south of the Khangai Mountains and north of the Mongolian or Desert Altai. The northern Khalkha Mongol is stronger in build and is generally better dressed, being fond of bright colored silks and cloth. The nomads of the barren country round the Gobi Altai are often of inferior physique and smaller in stature, as if the unfavorable conditions of their life in the desert had checked their physical development. The crowd that surrounded our cars was poorly dressed; some wore old rags, others dirty sheepskins, worn over their bare skins. Their long unkempt tresses of black hair gave the company a wild look. On the whole, the crowd resembled more the Hor-pas of northern Tibet, than the fine looking northern Khalkhas.

On April 21 we broke camp early in order to find our way through the intricate mountain system, and cross the muddy Toin-gol. Most of us had to walk in order to help out the cars. For considerable stretches of sandy plains the cars had to be pushed, with planks and canvas sheets put under the wheels to give them a grip. It is a marvel that they did cover this difficult ground and stand the severe test. About eleven o'clock in the morning we reached the Toin-gol—a river difficult to ford because of its muddy bottom. We had to stop and find a suitable place. After a long search one was discovered, but the river bank was too steep to allow the cars to enter the water. We had to make a route and level the steep bank. After three hours' hard work, all five cars were safely brought across the river.

Beyond the river we encountered again a very heavy road. Having driven for about ten miles, we found ourselves amid impassable sandy hills and had to turn south in search of a passage. Our Mongol guide was arguing with our Tibetan, who was of the opinion that we should have followed the camel track which lay south of the direction we had taken. It was difficult, almost impossible to retrace our steps and we therefore decided to follow a dry river bed, which if passable would bring us across the barren mountain ridge. The going was very trying. Every ten minutes one of the cars would sink into the sand and considerable time would have to be spent in getting it out again.

About five o'clock in the afternoon we decided to camp for the night on a level stretch of ground close to the dry river bed. A reconnoitering party was mustered to examine the possibility of a route farther up the dry river bed, which we were following. After an hour's absence the men returned with the report that the road was heavy with accumulations of débris, but that a passage could be effected along a side valley higher up the stream. We decided to try the route suggested by our scouts and set out. The country around us was absolutely barren. Scant shrubs of saksaul (Haloxylon ammodendri) and kharmik (Nitraria scholeri), typical Gobi plants, grew on the banks of the small stream that once had emptied into the larger stream, now dried up.

Next day we again made an early start and drove, following the indications of our Mongol guide, westward up the dry river bed. After an hour of heavy road, we turned into a side valley and ascended a low mountain spur that seemed to be a suitable pass across the ridge that barred our passage. We hoped, having crossed it, to find our way into the vast gravel plain that lay south of it. Our expectations were vain, for behind the pass we found the same intersected mountain country of rugged sand hills with scant outcroppings of granite. The only outlet open was a narrow mountain valley leading west. We followed it for some five miles till we came to a small circular valley sheltered on all sides by mountains. It was evident that there was no practicable road farther ahead. The Mongol guide continued to insist that we should cross the mountains southwest of the valley. This was easier said than done for most of the mountain side was steep and rugged. We could not possibly climb the steep ascent. Myself and a driver ascended one of the nearby hills, from which we could easily orient ourselves. We found a heavy but passable track on the very edge of a narrow canyon and the driver was of the opinion that we could risk a passage. We accordingly descended and guided our cars up the steep hillside and the difficult track that lay along the edge of the ridge. We had to advance very cautiously, for one faulty movement on the part of the driver would send the car down into the canyon. After the light cars were brought over, the heavy trucks ascended with the help of the men. Sand and stones rolled down when the heavily loaded trucks climbed the mountain, men were pulling from in front and behind, and it was a great relief when all trucks had safely arrived at the summit. We found ourselves in sight of the sandy plain with faint outlines of rocky mountains far to the southwest. Before reaching the plain we had to cross several successive ravines full of sand, in which our cars stuck more than once. The mountain ridge from which we descended continued far into the desert plain south of it, stretching its sandy spurs for a considerable distance.

We decided to camp on the flat top of a low canyon. The night was exceptionally warm—the first warm night since we had left Urga.

In the early morning we started west again, climbing up and down the numerous canyons and spurs that intersected the southern side of the mountains. Some of the canyons were difficult to cross. The Mongol guide from Udzen-Wang insisted that we should continue our journey westward.

About three o'clock in the afternoon we succeeded in crossing a low sandy ridge to the southwest and found ourselves in sight of a vast gravel plain with faint outlines of rocky mountains to the far south. Our drivers knew nothing about the route and from the existing maps we could only find the direction of the Yum-beise küren. We all were of the opinion that our Mongol guide was a failure and that the old man was guiding us in the wrong direction. I therefore questioned our Tibetan who knew the camel trail well and he was positive that our route lay south across the gravel plain. After a long argument, the Mongol guide confessed that he was guiding us toward the relay station of Yum-beise küren. We therefore decided to follow the indications of our Tibetan and cross the plain southward.

We endeavored to obtain some information from Mongol women, who lived in a felt tent completely hidden in one of the deeper canyons. Their men folk were away and the women knew nothing about the route. We had to rely on the knowledge of our Tibetan.

We started again about five o'clock in the afternoon and drove over good firm ground across the plain. It was a pleasure to be able to drive fast again, after all the trials of the previous days. We observed for the first time large herds of wild asses, some two hundred in all (Equus hemionus). The animals stood motionless watching our movements and then suddenly dashed in wild speed across the trail. We followed one of the animals in our car and it was surprising to see the speed it developed.

After an hour's drive we crossed the great highway Koko-khoto-Ku-ch'eng, distinguished as all the Chinese highways by deep wheel tracks left behind by convoys of heavily loaded Chinese carts.

It was getting dark and the after-sunset glow only dimly lighted our path. We decided to camp for the night on the shore of a small lake, the Boro-nor, situated in a shallow depression. It was almost dried up, but according to local Mongols it assumed large proportions every summer after the rains. On its shores we found numerous sea gulls (Chroicocephalus ridibundus), gray goose (Anser anser), and turpans (Casarca casarca). Late in the evening a large herd of khulans or wild asses visited the lake.

We started next day with the object of reaching Yum-beise küren. After we left the lake, the trail again turned toward the mountains and we found ourselves on a heavy road along a dry river bed, leading south. Sharp stones, large accumulations of gravel, and deep sand banks made the progress very slow. Some of the cars stuck and had to be pulled out by the men. It was strenuous exercise and many of us lost considerably in weight. After these twelve days of car extricating we could consider ourselves experts in the job, and indeed it took us considerably less time to pull out a car than at the beginning of our journey from Urga. I had an unpleasant experience with my car, which suddenly caught fire. The situation was a serious one, for I had most of our oil and gasoline supply and several cases of ammunition. We worked feverishly and succeeded in getting the fire under control before it neared our gasoline supply. I had to sacrifice the water from my flask in order to stop the fire that caught the floor of the car. The damages were only slight and we were able to resume our journey.

The last portion of the road to Yum-beise was exceedingly hard for cars. The country was a highly intersected mountainous region and the dry river beds presented great obstacles for wheeled vehicles. After reaching the top of a steep ridge, we suddenly found ourselves in sight of the long-expected Yum-beise küren, situated in a deep valley surrounded on all sides by hills that sheltered the monastery from the severe winds which blow during the winter and spring months. It is a collection of white houses in the center of which rise two du-khangs or assembly halls. A huge crowd of red-clad lamas, on hearing the sound of motor horns, rushed out of the monastery and surrounded our cars.

We found no suitable place for camping close to the monastery and were advised to establish our camp outside the monastery in a place locally called Tsagan Tologoi—"White Head," so named after a mountain west of the monastery. The place chosen for the camp was situated on the banks of a tiny stream, the Tsagan Tologoi-usu. From the northwest and west our camp was well sheltered by the mountains, Tsagan Obo. To the east and south the place overlooked the vast rolling landscape of the Gobi.

We were obliged to spend several days at Yum-beise in order to arrange our farther journey across the southwestern Gobi to An-hsi chou. Our camel caravan with the gasoline supply reached Yum-beise on the same day we did. It had followed a northern route past Erdeni Dzu and Lama-yin Gegen Monastery and spent several days grazing the camels near Lama-yin Gegen. This explained the delay. Local inhabitants were of the opinion that it would be difficult, almost impossible, for us to continue our journey on motor cars south of Yum-beise. According to them the road was almost impassable, and all former attempts made by Mongol cars to cross the Gobi Desert south of the monastery had met with little success. Considering the poor condition of our cars, we unanimously decided to send them back and to continue our journey on camels.

Among the chief obstacles to motor car traffic across the southwestern Gobi are the broad latitudinal valleys, stretching between the parallel ridges of the Mongol Altai, which are thickly carpeted with saksaul , called dzak by the Mongols. The drifting sand accumulates around the plants and forms small hummocks that often completely block the way, leaving barely enough space for a camel trail. The passages through the successive ridges of the Gobi Altai lead over heavy, stony ground strewn with bowlders and layers of gravel, and are in most cases impassable for cars, to say nothing of heavily loaded trucks.

Along the foot of these desert mountains stretch large belts of sand dunes and alluvial fans, often difficult to cross by car. Some time before a Mongol car had reconnoitered the route south of Yum-beise and, following a roundabout way, succeeded in reaching a point some four miles north of Shara-Khulusun. Farther south the car was unable to go; it was a light touring vehicle carrying two men and the necessary supply of gasoline.

Late in the afternoon, our camp was visited by the nyer-wa or manager of the monastery. He made an offer to supply us with a sufficient number of first-class camels and to accompany the expedition as guide. It was imperative to get first-class animals in good condition, for the camel season was at its end. The monastery was willing to supply us with camels and drivers up to Shih-pao-ch'eng, an oasis in the Nan Shan Mountains south of An-hsi chou. The route which we were to follow on camels was according to the nyer-wa utterly impassable for motor cars. The Yum-beise route is chiefly used by Tibetan caravans and Mongol pilgrims going to or returning from Lhasa. It lies across one of the most desolate regions on earth—rugged mountains and gravel-covered plains with neither men nor wild animals to break the monotony of the landscape. The region has always been a favorite hiding place for brigand bands, and the deeds and cruel massacres of Ja Lama, the warrior-monk, were still on the lips of local nomads.

The region is impassable during summer, because of the unbearable heat of the desert, the stone surface of which reflects the burning rays of the sun. All the caravan traffic is carried on between October and April. When we were crossing this forbidding country, a passage from Sir Aurel Stein's article on innermost Asia often recurred to my mind. Speaking of the Taklarnakan Desert, the great explorer remarks:

It is this extreme deficiency of water which invests by far the greatest portion of the area we are considering, with the character of what I may call "true desert." Let me stress the word "true" in this expression in order to make it quite clear that the ground over which I shall have to ask you to follow me tonight in tracing geographical movements, differs greatly indeed from those deserts with which biblical stories, descriptions of Arabian or South African scenery and the like have made many of us familiar in a certain sense. These "tame deserts," as I should venture to call them for the sake of distinction, may indeed impress the town dweller, especially if he comes from our centers of congested humanity, with their sense of solitude, emptiness and let me add, peace. But deserts in which whole tribes can wander about for long periods sure to find water and grazing for their flocks at least at certain regular seasons—deserts in which populations driven out from their seats or harassed by foes can safely seek refuge for a time, are not such as face us in most parts of the high basin between the Celestial Mountains and the K'un-lun.

The very same words can be applied to the western Gobi, the vast expanse of desert country south of Yum-beise and north of Altan-usu in the Artsa-yin nuru Mountains. Here the absolute aridity of the region made human and animal existence practically impossible. Traffic is carried on only on camels, which can stand the waterless marches, and is practically limited to the winter months. Large caravans on horses and mules are doomed to peril.

The nyer-wa of the monastery proposed to follow a caravan trail lying east of the usual caravan route Yum-beise-An-hsi chou. This route had the great advantage of being shorter than the usual one. Until recently it was considered dangerous for caravans on account of the brigand activity in its neighborhood, but considering our armament, the nyer-wa was confident that we could attempt the desert crossing by it. This route was often followed by contrabandists and was well known to some of the monks in the monastery. The nyer-wa promised to produce a reliable guide the next day. The only sure way of crossing a desert is to engage a good guide and intrust him with the selection of camping places and finding of water wells. European maps cannot always be relied upon, and moreover there are no route maps of the region of sufficiently large scale. The only means of finding the correct direction in the Gobi are a good guide and the compass.

The Mongol guides are famous throughout central Asia for their wonderful knowledge of the country. It is an almost supernatural capacity that always helps them to find a passable trail or water well. After our sad experience with the guide from Udzen-Wang to Yum-beise, we all were rather skeptical about these qualities of Mongolian guides, but on reaching Shih-pao-ch'eng in the Nan Shan safely, we had to admit that for a whole month of trying desert crossing our Mongol guides—lamas of the Yum-beise küren—never misled the caravan, always finding water supply in a country that seemed utterly barren.

Early next morning I went, accompanied by our Tibetan guide, to the monastery to settle with the nyer-wa the cost of hiring the camels from Yum-beise to Shih-pao-ch'eng and the date of our departure.

The Yum-beise Monastery was situated about a mile north of our camp. It consisted of several narrow streets, on either side of which rose brick walls with small narrow gates leading into courtyards where stood the lamas' houses or tents. The monastery has a population of some 500 lamas, who throng the streets and the large courtyards in front of the temples. In the center of the monastery lay a spacious square, on the western side of which stood the chief monastery buildings, two du-khangs or assembly halls. They were constructed in Tibetan style. The structure itself was built of bricks, and the front portion had terraces supported by eight wooden columns with the usual floral design painted in bright colors. The four corners of the temples were crowned by the usual jyal-tsen or "victory banners," and on the roof over the principal entrance stood the customary emblem—the Wheel of Law and the two deer. Behind the temple rose the gilded roofs of the chapel containing the sacred images. The roof was surmounted by a gilded gañjira in form of a vase or bum-pa .

I had no time to visit the temples, and, deciding to wait until the following day, hastened to see the nyer-wa . The monastery treasurer was living in a spacious Mongol felt tent situated in an inclosed courtyard. The interior arrangements of all the tents occupied by lamas is almost the same. At the northern wall stands an altar with two or three gilded clay or brass images, often photographs of the Pontiffs of Tibet, the Bogdo Gegen of Urga, and other outstanding personalities of the Lamaist Church of Tibet and Mongolia. A low thick mattress spread on the floor serves as a bed for the host and a seat for his guests. In the center of the tent stands a hearth and at the opposite walls tea urns, brass kettles, and other domestic implements. Some of the tents have Ning-hsia and Alashan carpets spread on the floor. The rest of the courtyard in which the tent is situated is filled with argal heaps.

Many of the courtyards had small storerooms, where the wealthier lamas kept their belongings. In some we found covered carriages to be drawn by a camel—one of those high-wheeled vehicles used in desert traveling. Such carriages consist of two large wooden wheels and a wooden covered carriage with small windows on both sides, which are usually barred with a wooden frame. The wood is covered over with blue cloth, and sometimes colored paper is pasted on the windows. Such carriages are used only by wealthy lamas or ladies who are anxious to protect themselves from the winds and gales of the desert uplands. They are drawn by one camel, and are made to accommodate usually only one person, or rarely two.

The nyer-wa met us in his tent and after the usual reciprocal compliments and polite inquiries about health and road, we started to talk business. The treasurer said that it was extremely difficult to find camels, for the season was late and very soon the desert would be impassable for camel caravans. He was quite willing to let us have his own camels, about twenty in all, but the rest of the required number of camels we would have to hire from some wealthy lamas in the monastery.

I asked him to conduct me to some of the camel owners in order that I might have a talk with them. He agreed and we all three went to interview a well-known guide and caravan leader, Lama Sambu, who had spent most of his life guiding caravans from Yum-beise to An-hsi. He knew the route well and moreover had a sufficient number of camels at his disposal. A big man, with a stout face and prominent cheek bones and a thick red neck, he was a great humorist and afterward proved to be the best of the whole lot of cameleers. He willingly agreed to accompany the expedition as guide and to rent his camels. After much talking and mysterious finger signs indicating numbers, made in the long sleeves of the Mongolian coat, the price of twenty-two Mexican dollars was fixed for one camel from Yum-beise to Shih-pao-ch'eng. The monastery treasurer and Lama Sambu undertook to reach Shih-pao-ch'eng in twenty-two days from the day of departure from Yum-beise, provided everything went well on the way. They also agreed to sell us several camel saddles made in Alashan to be used on riding camels. These consist of a flat thick carpet cushion, to which are fixed on both sides large carpet flaps. Stirrup leathers are attached under the top cushion, which serves as a seat. A long broad girth serves to keep the saddle in position between the two camel humps. Some of the camel saddles are finely ornamented and cost considerable money. Portniagin and I had to ride camels, in order to be quick on the move and able to do reconnoitering duty. The rest of the expedition would have to ride on seats arranged on the top of loads. This way of traveling on a camel is quite comfortable, for it permits the traveler to lie down and even sleep. Mongols always use this method of traveling.

After fixing the price and the date of departure, we returned to the camp and the rest of the day was spent in the tedious but necessary occupation of arranging camel loads.

Toward sunset a violent northeast wind set in and soon turned into a wild storm. Thick clouds of sand passed over the camp, penetrating the tents. We all had to seek cover and close the tent entrances. A miserable night it was! The sand accumulated in huge quantities round the tents, forming high ramparts. The noise made by the outer tent fly striking the inner flap was so great that it made sleep almost impossible. Toward dawn the strength of the wind doubled and I felt my tent cracking and then suddenly dropping on me. The two flaps broke down and I was buried under them and under sand that rapidly accumulated over the débris . After some exertion I managed to creep from under my tent and find my way out. The whole country was hidden behind an impenetrable yellow veil and the wind blew with unabating force. Our camp presented a strange appearance. The servants' tent was blown down, but the inmates slept quietly under the canvas. Portniagin's tent was about to go and was balancing on its pole and one peg that managed to remain in the ground. The doctor's tent had most of its pegs out, and it was a miracle that it still resisted the storm. Professor and Mrs. Roerich's tent stood firmly against the blast. Portniagin and I went round the camp fastening ropes and fixing tent pegs. Suddenly a wild crash was heard, as if hundreds of unseen riders had dashed across the camp, destroying everything in their stormy passage. The kitchen tent flew upward and all the water basins and buckets with a loud rattle were carried away over the stony surface. This was the climax. Everyone in the camp woke up. One could hear voices shouting, "What happened?" "Fix my tent!" "Can't get out!" After sunrise the force of the wind abated, and we were able to restore order in the camp. It was evident that we could not remain any longer on this windy spot. We found another suitable place at the foot of the Tsagan Obo Mountains, and Professor Roerich gave the order to move camp.

The new spot was well sheltered from the northwest and northeast by high hills. Tsagan Obo is a rugged granite ridge cut by narrow canyon-like valleys filled with débris . Not far from our new camp rose the highest point in the ridge, the so-called Tsagan Tologoi. In the afternoon lamas from the monastery came to inquire how we had fared during the terrible sand storm of the night. The wind had done serious damage in the monastery, carrying away many of the tent roofs and flag poles on the temple roofs. Yum-beise is known for its winds. The place is open from the southeast and northeast; and the early spring is always heralded by wind and dust storms of terrific violence.

The next night and day were quiet and we all went to visit the monastery. In the large assembly hall we found a crowd of lamas and novices busy painting wooden covers ( leg-shing ) for the 333 volumes of the Kãnjur and Tãnjur . The assembly hall of the monastery had the usual arrangement inside. The northern wall was occupied by the state throne of the Incarnated Lama of the monastery and several glass cases containing brass and clay images of crude workmanship. The other walls were occupied by the monastery library. Between the wooden columns supporting the roof were spread the low mattresses that serve as seats for the lamas during religious services. We noticed only a few painted banners. A big one hanging on one of the columns displayed a black and white drawing of the mandala or mystic sphere of influence of Shambhala, said to have been presented to the monastery by order of the late Bogdo Gegen. The rest of the banners were painted in bright colors, but of very inferior design. We searched in vain for the finely executed banners from eastern Tibet and Derge. Most of the brass images in the glass cases either came from Urga or from Dolon-nor. We were surprised to find such a paucity of really good things. The küren situated on the caravan route to Tibet should possess better examples of Tibetan religious art.

After the visit we returned to our camp, but the doctor went to see some stone figures said to be in the neighborhood of the monastery.

In the afternoon I took our expedition guard for some target practice, for we were soon to enter a brigand-infested area and the men needed to know how to use their rifles. The marksmanship was generally good and we thoroughly enjoyed the exercise, when suddenly some lamas interrupted our occupation and begged us not to shoot in the neighborhood of the monastery since the spirit of the Tsagan Tologoi would surely get angry and make his wrath felt by sending a fresh wind and sand storm. We had to return to our camp in order to avoid a misunderstanding with the lama. Apparently the prediction of the lamas was coming true, for a violent wind started shortly after our return to camp and we had to take all precautions to keep our tents in position. It grew stronger during the night and two of our tents broke down in spite of all precautions. Everything was enveloped by thick yellow clouds and we had to spend most of the day inside our tents.

On April 30 bright and windless weather brought some relief, after two successive days of violent wind and sand storms. We were able to make a final distribution of camel loads and the camel drivers brought in ropes and saddles, preparing for the next day's journey.

The morning of April 30 greeted us with a strong and steady northeastern wind. Mist veiled the peaks of the Tsagan Obo Mountains and wet snow lashed our faces. The camels arrived at ten o'clock — a fine body of stalwart animals in splendid condition. The guide, Lama Sambu, was careful to bring with him only young animals fit for the desert crossing. Most of the camels had had a whole year of grazing and looked remarkably strong. At eleven thirty the expedition caravan started in three long columns on its long track across the Gobi. The camels trod gravely and the large caravan bells on the last camel of each detachment droned dolefully. The lashing wet snow and wind abated toward noon and we had quite a pleasant march over the vast gravel plain, bordered by distant rolling hills that lay south of Tsagan Obo Mountains. The country was monotonous, quite different from the country north of Yum-beise with its branches of the Khangai Mountains stretching far into the desert region.

After five hours' march, the caravan halted at a desert well around which grew some shrubs of saksaul, a favorite food of the camels. We pitched our camp on a vast gravel-covered plateau. Far on the horizon rose rugged mountain ridges. To the north lay the Bain Ündür Mountains, a continuation of the Tsagan Obo Ridge. Far to the east, one could discern the faint outlines of the Elgi-yin ula Mountains — one of the many parallel ridges of the Gobi Altai.

The local camel drivers have a curious tradition of never mentioning the name of the camp, while staying in it. If pronounced on the spot, misfortune may befall the caravan and camels may perish. The name is pronounced aloud only after they have left the locality. For this reason the name of our camp, Tsagan Khuduk—"The White Well" was disclosed to me only after we had left.