ON ANCIENT CENTRAL-ASIAN TRACKS

a brief narrative of three expeditions in innermost Asia and northwestern China

Sir M. Aurel Stein
Oxford Clarendon Press, 1933

New edition edited with introduction by Jeannette Mirsky
New York Pantheon Books, 1964

This book is Stein's summary of events from all three of his expeditions. Some of the most important archeological sites he visited on both the first and second expeditions. Others he visited on both the second and third expeditions. And of course he visited Kashgar and Khotan on all three expeditions. In order to avoid repetitions he sometimes includes in one chapter descriptions of a place, such as Niya or Lou-lan from two separate visits which results in some jumping around out of chronological order. He focuses on the most important archeological excavations with less attention to some of the extended geographical- topographical explorations. Thus he skips his remarkable direct crossing of the Takla Makan from north to south and his extended months of survey on the western Tibetan plateau enroute back to India over the Karakorum, both of which took place during his second expedition. The story begins with his travel from Kashmir to Kashgar on the second expedition, leaving out information about the first and third trips. And it ends with his departure from Kashgar to London while completing his third expedition, leaving out the conclusions of the first and second tours. And also it ends with Stein at Samarkand again, leaving out his final exploration into south-eastern Iran.

In addition to all his other capabilities, Stein was an excellent writer of both detailed, professional, official reports and of more popular style narratives of which this is one. He includes enough of the substance of what was found and its significance with a rather dry humor filled account of his own activities and emotional responses to events and settings. The book serves as a useful popular account for the casual reader, but is no substitute for study of both the official reports and personal narratives of the individual expeditions. Unfortunately the official reports are voluminous and available only in rare book libraries. And the personal narratives also are out of print. To some extent, for the dedicated student, these have been scanned into digital form and are available on the Internet. But reading them there is a laborious exercise. I have been able to obtain the original reports thanks to extremely helpful interlibrary loan officials and have prepared summaries and comments in this series. Especially important for a reader to understand Stein's texts is reference to adequate maps, which indeed he did provide, but not for this book. The official reports also contain massive volumes full of illustrations and scholarly analysis in appendices.

Chapter I - A bird's-eye View of Innermost Asia
In this chapter Stein provides an useful summary of the geography of the Tarim Basin and surrounding mountains as revealed during his three expeditions. He notes that these consumed near 7 years in total and about 25,000 miles of travel. He points out that the slow travel (horseback and on foot) over an area of 1500 miles extent west to east and 500 miles south to north, and the detailed topographic surveying accomplished along the routes gave him the opportunity to study in detail this terrain. And indeed the detail described in his lengthy reports is amazing. Each long day was filled with both doing a wide variety of physical activities and with making copious notes in a diary on what was done and what was seen and experienced. The predominate geographical reality, he notes, that governs all else is the lack rainfall and of water except where it is provided by the melting of snow and glaciers in the high mountains around the rim. The Takla Makan desert, he describes, is much more inhospitable even to nomadic herders than the typical deserts known to the West from America and Africa. And even the vast expanse of the high plateau behind (south of ) the front ranges of the K'un-lun is a desert devoid of vegetation. He points out the essential role of the few major rivers, all fed from glaciers, in providing what water for irrigation there is. Thus the population is confined to oases drawing their water either directly from the rivers or from springs that bring glacier water flowing underground back to the surface. These oases exist in a narrow semi-circle between the sharp rise of the mountains and the desolation of the sand- desert. He ascribes the retreat of cultivation and expansion of the desert over the last 2000 years to contraction of the glaciers and resulting reduction in the river flow. Yet, it was through this basin, actually in routes along its edges, that commerce and culture passed between ancient Chinese and the European and Indian peoples. But it was the overall lack of water with resulting vast desert that prevented this vast plain from becoming the home and migratory route of the peoples (Huns, Turks and all) who inhabited the vast steppe north of the T'ien-shan and from the Carpathians to the Altai and Gobi. The chapter provides an excellent background for the reader of Stein's more detailed reports.

Chapter II - Chinese Expansion into Central Asia and the Contact of Civilizations
In each of his official reports and personal narratives Stein devoted extensive sections to discuss the Chinese role in Central Asia, not only for the sake of the ancient history itself, but also by way of interpreting the archeological materials he was uncovering and relating this history to modern circumstances. He explains that Chinese interest in the region began with their effort to defend against the attacks of nomads such as the Huns. In the course of pushing their defenses west through the very narrow Kansu corridor along the steep slope of the Nan-shan Mountains, they discovered the wealth of the societies residing across the Pamirs and Karakorum ranges - Transoxiana, India and Iran and all. The Emperor Wu-ti (140-87 B.C.) began this expansion with the potential of commerce in mind. Chinese superior military power enabled them to push the Huns northward and exercise suzerainty over the many small and weak 'kingdoms' throughout the Tarim. But, he notes, expanded and continual power relied more on diplomacy than warfare. The main campaign and then commercial route, once west of the narrow corridor between Nan-shan Mountains and Su-lo-Ho, led along the northern edge of the desert in the foothills of the T'ien Shan. Stein describes the process in detail, naming successful generals and events. An alternative route soon was opened along the southern edge of the desert through Charchan and Khotan. Chinese power and ability to control the Tarim area waxed and waned with the unity or lack of it in China itself, but also with that of Turkish and Tibetan peoples pressing from north and south. Ultimately it was the religiously motivated expansion of Arab (then Turk) Islam from the west that detached the Tarim from Chinese control until modern times. This new control was only accomplished by Emperor Ch-ien-lung in 1755. The history of this process and era is retained in the Chinese archives and the memoirs of a very few travelers. But, Stein notes, the details of every-day life and the realities of the constant battle against desert was little known prior to the rescue of the vast quantity of documents and artifacts he brought out from a series of long-forgotten and ruined settlements. Since the shift of trade from the desert caravans to the sea between China (and Asia) and Europe, the region has reverted into obscurity for Western peoples.
In a major conclusion he ties the eventual defeat of the settlers and abandonment of their villages to a gradual reduction in the glaciers far away to the south and resulting in diminution of the rivers and loss of irrigation.

Chapter III -Across the Hindukush to the Pamirs and K'un-lun
Stein extols the beauty and solitude of his mountain home in Kashmir. He ascribes his life there as the root cause for his life of exploration, archeological excavations, and study of Sanskrit and other languages of India and Turkestan. He specifically mentions what is obvious from study of his three expeditions, namely that he purposefully chose different routes across the Himalaya and Hindukush from Kashmir to Kashgar. The first route went on the eastern side from Kashmir to China through Gilgit and Hunza. For the second tour, his favorite, he went via a central route from Peshawar through previously inaccessible local chiefdoms of Swat and Dir into the Dard area of Chitral. From there he had secured the rare approval of the King of Afghanistan for passage through the easternmost part of the Wakhan. On the third trip he followed a western route since he was authorized to pass through Darel and Tangir (also not previously accessible to Europeans) to the Taghdumbash Pamir. Considering that all three expeditions took place only 100 years ago and that British rule over that extension of what was still India between Afghanistan and Tibet and south of China was then still very recent and dependent of agreements with local tribal rulers and their personal control over very warlike and unruly isolated mountain peoples the reader can recognize what political difficulties remain even today.
While the third expedition involved crossing more and higher passes in northwest India, Stein thus chose in this chapter to describe the Indian portion of the second expedition. As usual his description is humorously understated in many respects. As always, also, he cites the value of his assistants - in this case Rai Ram Singh (very experienced professional Indian surveyor), Naik Ram Singh (corporal in the K. G. O. First Bengal Sappers and Miners) and Jasvant Singh (high caste Rajput from Kangra and cook for the Indian Sikh surveyors). Stein admires Jasvant Singh's quality and wished he could have benefited from his experience but notes that such high caste prohibited Jasvant Singh from cooking or attending to a European.
(The Hindus and Sikhs also refused to eat with non-caste people, even at Chinese banquets.) Thus Stein notes he had to make due with Mohammedan cooks of quality best left undescribed. Other Indian assistants who participated in the third expedition appear during this narrative by name but seemingly from nowhere. Once in Turkestan Stein hired a high-quality group of expert pony and camel men (mostly the same, but not always, for each expedition).
He proudly notes that for this second expedition he managed to load its extensive baggage at the outset on 14 mules. ( In this book he does not describe in much detail the actual large quantity of equipment he needed, survey instruments, hundreds of glass plate photographic negatives and developing chemicals, tents and winter gear, even cast iron tanks specially built to carry water and ice.)
Departing the famous Malakand fort on 27 April, Stein passed through the Swat Valley, which he naturally connected with Alexander the Great's campaign. ruincathay1.htmHis personal narrative describes his encounters with the descendents of the ancient tribes and harrowing passage through narrow gorges and over snow-covered passes. Here is mentions the Lowarai, which he surmounted in early May, well before the normal season. For the crossing teams of porters were conscripted to replace the mules and he left the ponies behind while proceeding on foot. (But he does not describe the prior logistics planning required to arrange for one group of porters from one ethnic tribe to carry the loads up to the crest while another group from the other side has been sent up to bring the loads down into the next valley.) Surveying was not allowed in these valleys, but Stein made detailed anthropological measurements of the ancient peoples. He also took careful note of the various languages spoken in each valley. He would have loved to remain in Chitral for months to gather evidence of the antiquity of the Dards, but the huge project he had prepared for Turkestan beckoned. Throughout his journey, while still in India as well as in China, he sought to confirm the topographic and other descriptions made by Chinese medieval pilgrims and warriors. Duly noted were each Buddhist shrine or rock carving along the way. He was fascinated by the epic Chinese invasion of Yasin and Gilgit in 747 AD by led by General Kao Hsien-chih. With this objective Stein mounted the Darkot Pass (15,400 feet) on 17 May just to have a look, then actually crossed the Baroghil (12,400 feet) into Afghanistan, at the headwaters of the fabled Oxus. He remarks that Kao Hsien-chih's feat ranks greater than those of Napoleon, Suvorov and Hannibal in the Alps. Alas he was never able to fulfill his life-long dream of exploring the middle Oxus in Bactria.
He gratefully acknowledges the significant hospitality and assistance rendered by the Afghan military and local Wakhis commanded by Colonel Shirindil Kahn. Limited time and local resources forced Stein to move quickly on into Sarikol with a brief look at the Little Pamir Lake (13,000 feet). There his mind was drawn to the passage of his Chinese 'patron', Hsuan-tsang, and Marco Polo, both of whose journals he kept with him at all times. (He finally traveled on their footsteps in the Pamirs during his third expedition.)
From the Wakhan corridor Stein moved up the Ab-i-Panja tributary of the Oxus and crossed the Wakhjur Pass through deep snow. This time it was yaks rather than mules that proved inadequate. Stein remarks that only the fear of the Afghan army detachment watching from the foot of the pass induced the Wakhi and Kirghiz laborers to carry the heavy baggage. Starting at 3AM it was midnight before they reached the Chinese side. This now was the Taghdumbash Pamir, which Stein had crossed already in 1900 but coming from a different direction. (See the map made to illustrate that expedition). Hsuan-tsang passed through the high valley in 642 AD, and Stein made special note to confirm all the landmarks in the pilgrim's itinerary including several fortifications. His reports on this are in the accounts of the expedition. The Taghdumbash Pamir is a long, elevated plateau or valley from south to north on the Chinese side of the international boundary. Stein did not stay long in Tashkurghan, the capital of Sarikol. This time he passed by mighty Muz-tagh-ata ( which had defeated his effort to reach the summit during the first expedition) over the Chickiklik plateau and through narrow gorges to Kashgar.
At Kashgar Stein stayed once again with Sir George Macartney, the Indian Government representative. In addition to the many services Mr. Macartney rendered by obtaining official support from Chinese officials and purchase of the many needed supplies, ponies and camels, Stein was particularly grateful of the appointment of Chiang Ssu-yeh as his secretary and interpreter. Stein's experience with the incompetence or weakness of the interpreters he had during the first and third expeditions is well described in the relevant reports. Chiang became a favorite and devoted friend, much more than an interpreter, as he braved all the hardships of mountain and desert and assisted with supervision of the archeological work as well.
The next leg of the journey began on 23 June as Stein, now with an expanded team and large caravan, departed Kashgar for Khotan. He was eager to return to the desert ruins uncovered during the first expedition, but this was not possible during the heat and sand-storms of summer. This delay gave opportunity for more exploration in the K'un-lun south of Khotan. After a brief return visit to Yarkand, Stein moved into the mountains to Kok-yar where (amazingly enough) he proceeded to edit sections of the report of the first expedition - Ancient Khotan. Amazing because throughout these arduous expeditions Stein was constantly working simultaneously on reports and conducting a massive correspondence via 'dak carrier' with Europe and India. Amazing not only with respect to Stein's persistent efforts but also to the skill and endurance of these 'dak' runners who could find him in mid- winter or summer at no known address in mountain fastness or desert sand-dunes. (His two biographers make excellent use of the archive of Stein's correspondence.) But never to loose an opportunity, Stein also used the stay at Kok-yar to conduct more anthropological measurements and collect data on the Pakhpu peoples of the hidden valleys high in the K'un-lun. He decided that these Pakhpu people were descendents of the "Homo Alpinus" race that extended its presence from Iran eastward through the mountains. He was also busy with plane table to extend his survey from Yarkand to Khotan.
At Khotan he was welcomed by his chief friends from the first expedition among the Turki and Afghan residents. The reader will recognize that the hearty welcome that Stein always received and instant offers of hospitality and assistance were due to his remarkable personality which enabled him to honestly enjoy the company of such a variegated population of Indians, Chinese, Turkis, Khotanese, Afghans, and many other less numerous groups. But also he was always treated as a 'Sahib' with official status conferred by the British Raj in India. From Khotan, it still being summer, Stein promptly moved again south into the highest K'un-lun in his determination to establish the sources of the Yurung-kash river. He had tried and been defeated before, during the first exploration. (For unknown reasons Stein completely skipped discussion of this effort in his official report of the First Expedition but included it in his Personal Memoir) This time he reached the Nissa valley by mid-August (the first time was already in October and November). In this chapter Stein writes a very vivid description of the terrain he encountered and crossed. He managed to reach the foot of the Ortrughul glacier at 16,000 feet and surveyed peaks up to 23,000 feet. But in this chapter he refrains from describing the recalcatrance and near mutiny of the mountain men who had been conscripted to support his campaign. This is what, as much as the impassability of the terrain, forced him to abandon the effort until two years later, when he attacked the Yurung-kash from the south-eastern side of the K'un-lun. But he skips this part of the second expedition entirely. Another vivid description with detailed maps is found in his official and personal reports.

Chapter IV - First Explorations at a Sand-buried Site
Stein in this chapter reverts first to a description of his first expedition at Khotan. He briefly mentions his work around Khotan finding the ancient Buddhist sites described by Hsuan-tsang. He also comments on the way in which 'treasure hunters' have unearthed remnants of the medieval city, Yoktan, while sifting the lower levels in search of gold flecks. He was alerted to his first desert venture by an old desert denizen, Turdi, from Tawakkel. Stein also located and hired two desert hunters, Ahmad Merghen and Kasim Akhun who had led Sven Hedin during previous explorations. So on 7 December he set out for Tawakkel, where he hired (conscripted) reluctant local laborers to dig in the forbidding desert. He added local donkeys to his camels to transport the baggage, food, and water, leaving the ponies to return to Khotan (everyone including Stein had to walk in the deep sand). On 12 December he departed Tawakkel into the real desert of sand-dunes. After some erroneous guidance from the two hunters, Turdi was able to show Stein the ruin at Dandan-uiliq. Stein's description in this chapter is identical to that in his narrative. The chief ruins were of Buddhist shrines set in orchards and irrigated fields. The hardships of working long hours in below freezing weather were amply rewarded with the results carefully taken to the British Museum. The examples of art showed the influence Buddhist interpretation of Hellenistic style transmitted into Central Asia from the Gandhara area on the Indo-Afghan border. One especially intriguing painted panel Stein could not interpret until, at the end of his third expedition in south-east Persia he found a similar representation of the Persian mythical hero, Rustam (illustration in this volume). The painting at Koh-i-khwaja in Iran was dated to Sasanian era of the 7th century AD. Some Chinese documents unearthed at Dandan-uiliq were dated 781 - 790 AD. Other documents in the local Khotanese language also were considered to be 8th century. And copper Chinese coins found there all predated 760 AD. Stein relates all this to the close of the T'ang power in Central Asia about 791AD.
Stein is always at pains to use his knowledge of Chinese annals and memoirs to give historical context to his findings and to use his findings as confirmation of descriptions in the Chinese records. He also uses his interpretation of the archeology, geography and topography to draw conclusions about such general issues as when, how and why a community such as that at Dandan was abandoned. In this case cultivation at Dandan being dependent on irrigation from canal water the site was abandoned when the life-giving water no longer could reach so far into the desert.

Chapter V - Discoveries at the Niya Site
Stein moves on to the most significant of his archeological exploits during the first expedition and one expanded on during the second. He first mentions in passing that there were more sites south of Dandan-uiliq visited later. On the first tour Stein moved from Dandan-uiliq to the Keriya River and up it to the district headquarters at the Keriya oasis where he was informed about another potential ruin further north. From there on January 18th 1901 he moved on to Niya oasis and then down the Niya river to the Imam Ja'far Sadik Mazar (shrine). Stein credits ( he always does) in this case his devoted chief camel-man, Hassan Akhun, with showing him the first very exciting wood tablets on which was script in ancient Kharoshthi, known in India only from inscriptions (carvings) and never from documents.
Imagine the thrill for the explorer who had devoted years to the study of Sanskrit and ancient temples and shrines in India. This was in 1900 before the Indus civilization and so much more were discovered. Stein here details the interesting manner in which these documents came to him, just as he does in his report and narrative of the first exploration. The responsible party was one Ibrahim 'the miller' who had unearthed a whole pile of such wooden articles, a year before, while hunting for 'treasure' and then thrown them away as worthless. Stein was now desperate to secure this priceless (for him) trove (while wondering if Ibrahim was truthful) before they were lost forever to desert erosion or picked up by someone else. The rest of the chapter describes in detail both the excavation process and the wooden documents. Stein notes that no paper documents were found at Niya despite the evidence that the place was not abandoned until several centuries after paper was invented in China. In addition to the very large volume of Kharoshthi text he recovered, Stein was especially delighted to find that the clay seals used to secure each 'envelope' showed Greek figures including Heracles, Athene, Eros and other Hellenistic or Roman symbols. For study of the horde of ancient writing Stein enlisted many of the relevant experts but the volume of documents still prevented as of his writing the full decipherment and interpretation of the contents.
Enough has been done to show that they are all in Kharoshthi scrips of an early Indian Pakrit language with Sanskrit intermingled. They include official orders and financial records. Stein's conclusion is that these records match the information in ancient Chinese and Tibetan archives that the Tarim basin area around Khotan and east to Lop was invaded and occupied by settlers from India (Greek Taxila) in the 2nd century BC. Some documents use the official style of the Kushan Empire - Indo-Scythians who ruled north-west India, Afghanistan and parts of Trans-Oxiana. No such information in life of that period has been found in India itself. He notes confirmation of the dating from the evidence of Chinese copper coins of the Later Han Dynasty - pre-220 AD. There were also Chinese documents dating from the Emperor Wu-ti in 269 AD. Rather than lack of irrigation, Stein concludes that the Niya site was abandoned long before that at Dandan and as a result of political disturbance. Again, lack of time forced Stein to move on February 13, 1901 without completing full excavation of all the buildings.

Chapter VI - The Niya Site Revisited and the Remains of Endere
In this chapter Stein returns to his return to Niya during his second expedition. On this occasion he excavated first at Domoko, as described in his report and narrative, before moving on to Niya on 15 October 1906. This time the locals were eager to participate. (Another indication that Stein had become something of a charismatic figure throughout the region.) This time he had 50 workers and ample supplies for a month plus camels for transport. As described in Serindia his prior experience had taught him to create more means for carrying water and ice. His antiquarian instinct and classical education prompted him to the thought that the remnants of ancient orchards all around "this modest Pompeii" had existed at the time of the Roman empire.
Again, in this chapter Stein describes in detail his feelings of delight and accomplishment as well as the specific procedures followed and incidents involved in the careful archeological digs. This time he had more help also, both the engineer-carpenter, Naik Ram Singh, and the Chinese secretary- interpreter, Chiang, were now fully prepared to supervise as well as perform their main duties. The narrative flows as room after room is excavated, surveyed, photographed and examined for its architectural aspects. And simultaneously the expanding contents fill Stein's hoard of Kharoshthi and Chinese documents. One cache in particular Stein describes as it was uncovered from a special hole beneath the floor of one room. He already had over 100 more documents but this find was special, an archive rather than random abandoned documents. And their preservation was near perfect due to their location. The seals and strings were in perfect order indicating that they were official records, such as deeds, requiring authenticity. The name "Honorable Cojhbo Sojaka" appeared on many. And the seals were fresh with images of Zeus, Heracles, Eros, Pallas Promoachos, and other Greek divinities. Besides dwellings and offices Stein found fences, orchards, gardens, ancient canals, river beds, even footbridges now across dry depressions.
From Niya Stein moved east with a much larger program in mind than that of 1900-01. In a couple of paragraphs he mentions his first visit from Keriya to Endere and the excavation of the Tibetan fort and Buddhist shrine that he dated to around 719 AD from a Chinese inscription. This and his concern that Hsuan-tsang when passing this location in 645 had recorded there was no settlement existing but a legend of a much older abandoned city. So Stein was gratified that during the second expedition, when he again visited Endere briefly enroute to Charchan he found the missing evidence of that much more ancient ruin. More Kharoshthi documents found at a previously overlooked ruin confirmed that indeed there had been a settlement at Endere in the Indo-Scythian (Tukhara) era contemporary with Niya. It had been abandoned at the same early period and then a new outpost with fort had been built in there by the Tibetans much after Hsuan-tsang's visit. Later Stein found that the fort rampart he surveyed in 1901 had actually been built right on top of the earlier remains dating from centuries past.

Chapter VII - The Ruins of Miran
Endere was the furthermost east of the sites Stein visited in 1900-01. In this chapter he continues the second expedition account with the discovery and excavation of fort and shrine south of the village, Miran. In December 1906 he stopped to investigate Charkhlik, the ancient government seat of the Lop area and still a government headquarters, but of a vast area containing a much reduced population. The ancient Chinese records name the area Lou-lan and then Shan-shan. The history is fully described in Stein's official reports. Here, he naturally notes that Charkhlik was dutifully mentioned by Hsuan-tsang and Marco Polo. Stein had practical reasons for visiting Charkhlik also, as the last inhabited place of sufficient size where he could recruit local labor for the work at Miran and Lou-lan. Again, he had some difficulty due to the combination of superstitious fear of desert and the obvious privations to be experienced from hard work in desolate desert during the height of winter. A combination of the Amban's demands and the assurances given by another two desert hunters Stein recruited enabled him to assemble a work force.
On the way to Lou-lan Stein passed by Miran and discovered the ruined fort. He made some preliminary excavations there in early December and returned for more extensive work in January 1907. Miran fort is located on the barren gravel plain that extends down from the K'un-lun mountains to the Lop-nor marshes and Su-lo-Ho delta. Stein's work there is fully described in the personal narrative and official report. Stein here simply recalls that the three-weeks of work was a misery for all concerned. All the assistants save Chiang Ssu-yeh became ill. He notes that in December he recognized that the ruin of a Buddhist stupa far preceded the construction of the Tibetan fort - thus he had another location similar to Endere, where two entirely different groups had built entirely different communities at the same locale, but centuries apart.
On his return to Miran on 23 January 1907 Stein got down to work digging into rather disgusting refuse dumps in some places 9 feet deep.
Stein remarks, "Nothing but absolute indifference to dirt could have induced the occupiers to let room after room of their closely packed quarters be turned into regular dust bins, choked in some instances up to the roof. I have had occasion to acquire a rather extensive experience in clearing ancient rubbish heaps, and know how to diagnose them. But for intensity of sheer dirt and age-resisting smelliness I shall always put the rich 'castings' of Tibetan warriors in the front rank."
These revealed masses of documents in Tibetan, but not one in Chinese, fragments of all sorts of personal items including armor, and implements and furniture. The illustrations of the lacquer armor are particularly interesting. Also, he notes, a small group of documents in runic Turki script all relating to military activities shows that Turkish mercenary warriors had entered service so far east as Miran. Stein reckons that the Tibetan fort was built to guard the route between Khotan - Charkhlik and Tun-huang. He surveyed the fort and shrine and provided excellent diagrams and photos in his official reports.
Stein includs in this volume illustrations of the two winged cherubs that are in other reports. One can find all necessary visual examples in those reports (see Serindia). The fort itself is interesting but not particularly important. But the many frescos and sculptures Stein removed or photographed at the Buddhist stupa are of extreme importance for the history or art.
Stein notes that the thousands of documents provide a unique view into both Tibetan every-day life, at least on a frontier, and of specific military affairs during the century in which the Tibetans held sway over most of the Tarim Basin. He points out that well-known Tibetan literature is almost all about Buddhist religious beliefs and affairs. He identifies from the documents the correspondence of "Castle of Great Nob" with Charkhlik and "Castle of Little Nob" with Miran. Also, he notes that the Buddhist stupa must have been in the far more ancient "Yu-ni" in Chinese Shan-shan.
He turns to description of his work and results at the Buddhist stupas. Right off he gives some dimensions not mentioned (but illustrated) in the reports. One detached, giant head of Buddha measured 17 inches across. The measurement across the knees of seated statues was 7 feet. But nothing above the lower part of these sculptures remained. The style corresponded to that of Graeco-Buddhist Gandhara. He found bits of Brahmi scrips dated no later than 4th century and several silk banners that indicate the abandonment was in late 3rd century.
Stein's vivid description of his feelings and careful work to preserve these unique artifacts is even more moving in this book than in the personal narrative. His view was expressed, "As in eager excitement I cleared head after head with my bare hands, I rapidly convinced myself that the approach to classical design and color treatment was closer in these wall paintings than in any work of ancient pictorial art I had seen so far, whether north of south of the K'un-lun." In the official report and personal narrative Stein describes in detail the techniques he used with professional help by Naik Ram Singh to detach and pack these frescos and deliver them undamaged to the British Museum. He here describes some of the painted scenes, but photographic illustrations are best seen in the report and narrative. Here he points out that the iconography is strict Indian Buddhist, but the artistic style is Hellenistic. The winged 'angels' could well be Indian Gandharvas.
While at Khotan toward the end of the second expedition Stein dispatched Naik Ram Singh back to Miran in hopes of securing more frescos, but the sturdy Gurkha suddenly was blinded by glaucoma and unable to complete the task. Then when Stein returned to Miran in January 1914 during the third expedition he found that an inexperienced 'archeologist' had ruined much of the frescos Stein had carefully buried in a futile attempt to remove them. Stein, did manage at that time to remove an additional portion. But he laments that forever the only record of these remarkable frescos and relievos will be the photographs he managed to make.

Chapter VIII - Explorations at Ancient Lou-lan
Stein jumps back to December 1906 and his preparations at Charkhlik to explore Lou-lan. He gives appreciative credit to the Amban - Lio Ta-lo-ye for his essential help in securing enough reluctant local cultivators for work in the forbidding desert. He managed to assemble 14 more camels in addition to his own 7 to carry the essential bags of ice as well as equipment and food supplies. He enlisted two experienced desert hunters, Mulla and Tokhta Akhun as guides. The logistic planning he organized was worthy of a major military commander. Everything was planned out. All non-essential baggage and collected artifacts were left at Abdul hamlet along with Tila Bai and Chiang Ssu-yeh. The ponies also were left. The caravan consisted of camels and donkeys divided into sections. Donkeys and camels each have characteristic strengths and weaknesses (needs for fodder or water) which determine how they will be employed. Intermediate supply and ice depots were established along the route from which the donkeys on one leg and the camels on another could move back and forth to move ice and supplies onward. Then, during the stay at Lou-lan, the animals were dispatched to oasis where they could graze and find water. While Stein supervised excavations Rai Ram Singh performed plane table surveys and reconnaissance missions into adjacent areas.
Stein describes the difficult trek from Abdal across forbidding terrain with only compass and a copy of Sven Hedin's sketch map as guidance. Along the way they found Neolithic stone tools and coarse pottery lying on the hard clay surface. Stein devotes 7 pages to summary descriptions of the work and results, which can be studied in more detail on the official report. They departed Lou-lan on 29 December 1906 with, again, Stein taking a separate route back to Charkhlik and Miran in order to expand the area being surveyed.

Chapter IX - Tracking the Ancient Route across the Dried-up Lop Sea
In this chapter Stein interrupts the story of the second expedition (from Miran to Tun-huang) in order to jump to his return to Lou-lan during the third expedition and his most trying trek right across the Lop salt flat from Lou-lan to the limes north of Tun-huang. While surveying the Han dynasty limes at Tun-huang on the second expedition Stein understood that the Chinese annals specified that the central and major military and commercial caravan from Tun-huang to Lou-lan and then to Kucha led directly across the Lop salt sea, not by the southern route through Charchan and Khotan nor by a northern route skirting the Lop sea from the north. So he was determined to find this abandoned route. This he accomplished in 1914.
The chapter begins with his arrival again at Charkhlik on 8 January 1914. He briefly recounts the story told in Innermostasia about the episode in which, just prior to his arrival, a band of Chinese 'revolutionaries' had attacked Charkhlik and murdered the Chinese Amban. But these no doubt stupid adventurers had then been surprised in their sleep by a detachment of tough Tungan troops dispatched from Kara-shahr (far to the northwest) and all disposed of. All the while the local Muhammadan population stood by indifferently to either group. Now the remarkable coincidence was that just prior to his demise the Amban had received an edict from headquarters at Urumchi to prevent further survey work by Stein. But the document lay unopened in the Yamen under seal and according to Chinese civil-military regulations would not be touched by the military commander, pending arrival of a new civil Amban. That worthy did not show up until Stein was safely deep in the desert at Lou-lan.
For this exploration Stein had 30 camels and 35 workers. He got moving again from Miran on 1 February 1914. Along the march Tokhta Akhun showed him several more ruined forts at which he found documents in Kharoshthi, Brahmi, Chinese and Early Sogdian. He mentions here that on the third exploration he had Afrazgul Khan (a Pathan draughtsman, Sepoy of the Khyber Rifles) instead of Rai Singh (unfortunately ill), along with R. B. Lai Singh, (another very experienced professional surveyor from the second expedition) both as surveyors. Stein in this account only mentions the excavations on this occasion at Lou-lan, but notes that they found more forts, a cemetery, and many artifacts. The main subject of the chapter is to recount his daring direct march across the Lop salt sea in dead of winter - some 120 miles if direct, with no sources of water or grazing. Stein prepared with due caution. First they went north to the spring at Altmish-bulak to refresh themselves and the camels and secure more ice. In Innermostasia the story reads like a true adventure. They were not far along the trek guided mostly by Stein's pure guesswork, when they came across a line of some 200 copper Chinese Han dynasty coins lying for 30 yards on the open, hard clay and salt flat. And these were followed by many arrow heads. Stein recognized that these must have fallen out of some sack on a caravan nearly 2,000 years before. So they were on the right track. (What a remarkable scene. - Remarkable also is that the coins could remain as they had fallen on the surface despite the incessant wind erosion. They must also have escaped observation by subsequent Chinese caravans due to being slightly off the standard path.) Further on Stein and assistants found more evidence of a caravan route and also the unusual terrain features the Chinese had called ' White Dragon Mounds'. On the 9th day out from Altmish-bulak they reached the first scanty reeds and bushes and arrived at Kum-kudak, an oasis on the caravan route to Tun-huang. Stein down plays his own exploit but stresses how remarkable it is that the Chinese could conduct major caravan traffic across this 120 miles for several centuries.

Chapter X - Discovery of an Ancient Border Line
Stein reverts to the second expedition on 21 February 1907 when he departed Miran across the 380 miles of desert to Tun-huang. Naturally both Hsuan-tsang and Marco Polo had trod the same track, the latter requiring 28 days. But now, Stein notes, that the route has been rediscovered it is only used in winter when a caravan can carry a sufficient supply of ice through the waterless desert. Stein made the crossing in 17 long days. Stein describes his joy when he came across the first watch tower and recognized its antiquity. He had been hoping for such discovery from having studied the ancient Chinese annals in which the wall was mentioned. The next morning he sighted another and then another with the trace of a wall between them. Many more towers were apparent before Stein had to turn to the south-east to hurry on to Tun-huang. But he immediately realized that he would need again to hire a sizable crew of diggers in order to excavate this new find. Despite the friendly assistance of both the civil and the military Ambans, Stein was only able to dredge up a dozen opium smoking lazybones. Stein in Serindia notes the striking difference in social and political attitude of the Chinese laboring class toward their magistrates in China proper from that of the Moslem Turki working class toward their Chinese rulers in Turkestan.
In this chapter Stein describes both the process and results of his efforts along the limes during the second expedition. He immediately returned north from Tun-huang and surveyed a section of the wall with towers toward the east. Then he returned to Tun-huang and recruited other workers. He then went south-west to Nan-hu and then north-west again to survey the wall and towers toward the west, to the end of the wall. After spending time at the Caves of the Thousand Buddhas, Stein surveyed part of the wall further east to An-hsi enroute to Su-chou. And he surveyed still more of the wall and towers toward the east during his third expedition.
But in this chapter he includes only a brief description of his initial efforts.

Chapter XI - Finds along the Ancient Chinese Limes
Stein again prefaces his description of his work with an excursion into Chinese history. He recounts how Emperor Wu-ti after 121 BC pushed the Huns north and established the security of a caravan route just north of the Nan-shan by establishing military colonies and constructing a defensive wall. This effort, Stein notes, was similar to the Roman system of frontier limes that included both a protective wall and a road system. Thus Stein's continual terminology "limes' for this Chinese counterpart. The Annals tell us that the Chinese system was complete by 108 BC from Su-chou to the "Jade Gate". After a second campaign deep into the Tarim Basin in 102-101 BC, the emperor pushed the wall and towers further west past Tun-huang to the 'Salt Marsh'. Stein here rather proudly notes that he actually found at one of the further west towers a tablet with a date of 94 BC. And, from the last watch tower on the wall itself he excavated a series of detached towers stretching to the south-west along the edge of a marsh to serve as alert and signaling stations.
Stein continues his description of the wall, towers, and contents found in rooms and refuse dumps. He mentions one document dated 68 BC and a pile of over 300 pieces dated 65-56 BC found in an apparent archive. The documents detail military and individual life along the frontier - orders, memos, organizational details, rations accounts, weapons inventory and repair, calendars, recruiting records, leave records and more. He also found a document written in the Early Sogdian language of Samarkand. In Serindia he devotes a full chapter to his analysis and conclusions drawn from study of the entire mass of material. He then turns to the wall toward the east along the Su-lo Ho and beside the marsh and Khara-nor lake. Among the many towers there he found the remains of one built of brick rather than stamped clay. In it he found a medicine chest and large supply of bronze arrow heads.
Throughout his narrative Stein returns to admiration of the skill of the Chinese engineers in the way they took every advantage of specific terrain features when siting each tower and constructing the walls. The construction effort in an area devoid of water and lacking building material beyond clay and reeds must have been enormous and very expensive. Thus the location of every marsh and lake, plateau, ravine and mesa was integrated into the complex design. (The reader must also remark that the correspondence of the location of wall and towers to details of terrain also indicates that after 2000 years the terrain itself has not changed despite the effect of wind and water erosion.) Stein here gives expression to his excitement, elation and fascination as he explored mile after mile of ancient walls and multiple towers never described by Western scholars.
His emotional expression shows in this comment, "Never did I realize more deeply how little two thousand years mean where human activity is suspended, and even that of Nature benumbed, than when on my long reconnoitering rides the evenings found me alone at some commanding watch station.... How easy it was then to imagine that towers and wall were still guarded and that watchful eyes were scanning the deceptive depression northward for that fleet and artful enemy, the Huns!"
He continues this thought by noting that the frequent finds of bronze arrow heads and Chinese reports of raids and ambushes indicate that this was indeed an active military frontier. He thought of the raids the Huns would conduct along this frontier and of the offensives and campaigns they centuries later mounted into the Roman Empire.
Stein noticed two interesting examples of the continuity of human evidence in addition to the towers and wall itself. When the sun hit just right the shadows revealed a long but shallow depression parallel to the wall and about 10 yards from it. This he concluded was the path worn into the surface by Chinese patrols marching along the same path for centuries. The other curious find was the piles of carefully stacked fascines laid out in a pattern adjacent to a tower. He recognized that these must be prepared combustible material ready for use when a fire signal was urgently required.

Chapter XII - The Cave Shrines of the Thousand Buddhas
Stein dates his interest in and desire to visit the Caves - Ch'ien-fo-tung - to reports he received of the visit there by his fellow Hungarian, Professor de Loczy, who visited the place in 1879. Therefore he had such a visit included as a priority for his second expedition. In this chapter he describes the physical layout of the cave complex and what can be seen in the many shrines, chapels and living quarters built at various heights into the rock wall. The elaborate decorations include both full statues and fresco paintings on walls and ceiling. The original entrances and wood galleries providing access have largely fallen to the extent that caves high on the cliff are no longer accessible. But there is more than enough to see in the lower levels. These include massive statues of the Buddha that rise inside through several levels. There are many statues of Buddha in various reposes and of numerous Bodhisattvas and other saints and religious figures. He dates the majority of the art work to the T'ang Dynasty period. The styles indicate a mixture of Indian iconography of the Graeco-Buddhist type with Chinese taste and technique. He notes also that the area fell under Tibetan and Turki domination with the decline of the T'ang and suffered damage during that time, but regained importance after the Mongol conquest. Thus Marco Polo recorded that the location was a center of idolatry.
Stein was able to photograph (unfortunately only in the glass plate technology of his day). But his great contribution to our view of the art work was due to his discovery and adroit diplomacy. It turned out that some time prior to his arrival the local priest, Wang, who actually was a Taoist rather than Buddhist, had found a massive assortment of documents and paintings hidden in a sealed room which he carefully guarded. It took all of Stein's and Chiang's psychological manipulation, discrete diplomacy, and a sizable quantity of silver to convince this pious monk even to let Stein look at the documents. After which Stein gradually talked the monk into secretly letting him take away (eventually) hundreds of priceless samples.

Chapter XIII - Discoveries in a Hidden Chapel
In this chapter Stein continues the story of his coup. By way of justification for abducting such a trove (24 cases of manuscripts and 5 of paintings and textiles) to safety in London, he points to the less satisfactory results of a subsequent visit to the caves by Professor P. Pelliot who talked Wang Tao-shih into letting him take another group of manuscripts. But the (foolish?) Frenchman proceeded to take them out of China via Peking (not clandestinely as Stein knew he must do) thus alerting the Chinese authorities to what was happening. The government naturally then ordered that all the remaining material be carted off to Peking. Naturally the result of that was that Wang did not receive the promised silver and that many of the documents were pilfered along the way. Still, when Stein returned to Tun-huang in 1914 Wang Tao-shih sold him more manuscripts that he had belatedly recognized had great value and had hidden them from the government officialdom. Now Stein does not mention it in the book, but the whole affair generated such uproar and denunciations of his 'theft' on the part of Chinese officials that he was forbidden to do any such thing again when he applied for a fourth expedition, which was aborted on that account.
Stein refers the readers to Serindia and The Thousand Buddhas for further information on the documents. In this chapter he briefly summarizes the results from scholarly study in London. The total results comprise some 3000 rolls and nearly 6000 detached documents. The documents date from the early 5th century to the early 11th century. Apparently the entire collection was hidden away in the secret room when the region was under attack. The contents are in Chinese, Indian, Tibetan, Tokhari, Turki, Khotanese, and other languages. The texts relate not only to Buddhist but remarkably also to Manichean religion. Other texts are on secular topics that provide information on previously unknown daily life and political administration. There are texts of previously unknown works and others that only were known by fragments or citations. One document dated to 868 AD is the earliest example of block printed book yet known.

Chapter XIV - Buddhist Paintings from the Thousand Buddhas
Stein turns in this chapter to description of the art work - paintings, banners, silk hangings of various kinds. These also, like the manuscripts, require years of study and analysis. As always Stein states his appreciation to the specific individual scholars who have done the expert analysis. The art work, like the documents, originated from many places, some distant such as Tibet and Nepal and from local artisans using a mixture of various styles. A few show they were created by artistic masters, but the majority are clearly from less-skilled artists commissioned by donors for use as votive offerings. But even these provide valuable understanding of the way in which artistic concepts from India, China and the West merged in Turkestan. Moreover, they date from the T'ang era- mostly 7th to 10th centuries, a period of great Chinese art but of which hardly any authentic specimens survived in China. The earliest with a specific date is 864 AD but Stein believes others are older yet. Stein considers the location where the art was created an additional cause for their significance. There, he notes, at the crossroads of India, West and China the student can see the influence of Mahayana Buddhism from India mixed in with the influences flowing out of China.
This is an unusual and important chapter in this book. In comparison with other chapters Stein provides rather more specific details of the artifacts. One has to read it to appreciate the results of the author's knowledge of Buddhist iconography. Here I will only mention the names of Buddhist figures Stein identifies in the pictures. For each he analyzes style, color, clothing and other details. The Buddha naturally ranks first, but also Gautama Bodhisattva, and his life even earlier with Dipankara Buddha. There are scenes from his birth and youth along with his mother, Queen Maya. The Nagas are represented as are Gautama's horse, Kanthaka, and groom, Chandaka. All these paintings mix Indian and Chinese styles and iconography. Among Bodhisattvas, Avalokitsevara, the Bodhisattva of Mercy, is depicted frequently. Many others are anonymous. But Manjusri and Kshitigarba have frequent representations. Manjusri appears on his lion, while Samantabhadra rides his elephant. The Four Lokapalas, warrior kings, (Guardians of the Four Regions) are present. These are Vaisravana (Kubera) ruler of the north with his halberd and model of a shrine; Virupaksha, ruler of the south with sword; Dhritarashtra ruler of the east with bow and arrows; and Virudhaka, ruler of the west with mace. These appear in various paintings with more or less typical Chinese or Indian styles. There are also depictions of several Buddhist Paradises. Several depict Buddha Amitabha seated on a throne between Avalokitsevara and Mahasthama with lesser Bodhisattvas below and the six original disciples to the back and the figure of the donatrix below. Another Paradise, Bhaishajyaguru, the Buddha of Medicine, appears in another painting.

Chapter XV - Explorations in the Nan-shan Ranges
During the second expedition Stein moved from Tun-huang north-eastward to the small oasis, An-hsi, the crossroad of caravan routes from Kansu to north-west and south-west. This had a convenient, secured Ya-men where Stein could store all the artifacts collected to date. From there he moved south into the high Nan-shan for summer surveying. He found a ruined walled town near Chiao-tzu. His digging there dated the ruin to the 12th or 13th century. He notes the impact of continual wind eroding the east walls but leaving the north and south walls practically untouched. Further into the mountains he found another set of caves - Wan-fu-hsia - "Valley of the Ten Thousand Buddhas", an active pilgrimage site even with a title rather more boastful since it was smaller than the "Ten Thousand Buddhas" caves. There he found interesting frescos but no cache of documents. He continued up across the dry plateau's of the Nan-shan, turning eastward, and then back down into the corridor at Chia-yu-kuan, the famous gate in the Ming Great Wall. He describes all this in Serindia. (The fortress has been renovated for tourists and appears in most books on the Great Wall.) Stein returned to this area during his third expedition in which he found more of the eastern end of the Han Dynasty wall that anti-dated that of the Ming by centuries. Here he repeats his analysis written in Serindia, that the Han wall was part of an offensive program to expand Chinese control into the Tarim, while the Ming wall was defensive, to exclude visitors from the west.
Stein then visited Su-chou, the first large Chinese town. He comments on the difficulty he had in enlisting (even with official edits) Chinese laborers who were fearful of going into the high mountains. Once there against their will they attempted to mutiny and desert, prevented by the Chinese military escort wisely provided from the Ya-men. The description of all this is more extensive and vivid in Serindia and Ruins of Desert Cathay. The resulting map that shows Stein's routes across extremely high passes is in Ruins of Desert Cathay. He found isolated camps of gold diggers and nomad herders in the valleys. He spent August riding and walking over 400 miles among the glaciers and snow covered peaks rising 18,000 - 19,000 feet and triangulating also on the main ridge rising to over 20,000 feet. He located the sources of the Su-lu-Ho and Su-chou River (which flow north and that of the Kan-chou that flows there south east before turning abruptly north at Kan-chou city. The survey maps covered over 24,000 square miles. He descended across the Richthofen Range to Kan-chou.
Stein returned to Kan-chou during his third expedition to survey yet more of the Nan-shan mountains. His narrative in this chapter seamlessly switches to the third expedition, in July 1914, described in Innermostasia. During this return into the mountains Stein found more Buddhist cave temples at Ma-ti-ssu. He notes that here he met the watershed between Central Asia and Pacific drainage, which corresponds also with a dry area in which cultivation is by irrigation and a humid area in which agriculture is possible from rainfall. He again crossed passes over 11,000 feet elevation. At that point he suffered a serious accident when his horse fell backwards on him severely injuring his thigh. R. B. Lal Singh managed to complete the survey which doubled the included area. Stein was carried back to Kan-chou in a litter. From there, with his leg gradually improving, he set out in late August 1914 north-west to revisit the area of the T'ien-shan mountains north of the Tarim. For this he crossed the Pei-shan range on the edge of the Gobi Desert by routes never before explored by Westerners. The first objective was the oasis of Mao-mei where the Kan-chou and Su-chou Rivers join to flow as the Etsin-gol toward Mongolia.

Chapter XVI - From the Etsin-gol to the T'ien-shan
Stein begins by commenting that he had already visited Etsin-gol in May 1914 enroute to Su-chou, so he back-tracks a bit to recount his search for the eastern end of the Han wall. This river valley was an ancient access route from Mongolia into the Kansu corridor. Stein found many ancient Chinese forts built to block these raids. His narrative here reverts to May and June, that is prior to the second period of survey in the Nan-shan from Kan-chou described in the previous chapter. He describes his search of the great fortress city Khara-khoto (Black town), which Marco Polo named City of Erzina. (Some of the ruin is still a very romantic site for intrepid tourists.) It was an important station on the caravan route between Kansu and Mongolia, 150 miles north of Mao-mei, already when Chingis Khan attacked in 1226 and remained so into the 15th century. Its heyday was during Tangut or Hsi-hsia Buddhist control in the 11th -12th centuries as shown by the numerous Buddhist stupas. Stein recovered many Buddhist manuscripts and block prints in Tibetan and old Tangut language plus stucco relievos and frescos. As usual Stein 'mined' the refuse dumps and retrieved more paper records in Chinese, Tangut, Uigur or Turkish including a printed bank note of Emperor Kublai Khan. He estimated that the city was abandoned when a change in the river made irrigation too difficult. While Stein was excavating Khara-khoto, Lal Singh was expanding the area of the survey.
When they departed Khara-khoto in June the camels were sent for their summer vacation. They rejoined Stein at the end of August. The narration again picks up in September 1914 with Stein, in a pony litter and still suffering from his damaged leg, back from the Nan-shan and Kan-chou and headed across the Pei-shan for nearly 500 miles toward the T'ien-shan from Mao-mai. They traveled again in two separate parties to increase the area being surveyed. He notes here that at Mao-mai he hired two guides who claimed knowledge of the Pei-shan. But in Innermostasia he discusses in more detail the results of their failures. But Stein's dead reckoning and compass work saw them through the 28 days of riding across desolate rocky terrain. After passing the one well at Ming-shui they recognized the peaks of the Karlik-tagh in the far distance. Further adventures awaited in the final crossing of the Pei-shan until eventually they emerged on the southern edge of the Dzungarian plain. They arrived safely at Bai.
Stein spent October surveying the northern side of the eastern end of the T'ien-shan to Barkul and Guchen. As he noted, the Dzungarian steppe, that is the reasonably well watered area between the T'ien-shan and Siberian forests was always the home of the major nomadic tribes from the Great Yueh-chin (Indo-Scythians), to the Huns, Hephthalites, Turks and Mongols. And it was from their base in Dzungaria that these warriors raided south through the Tarim basin as well as migrated west into Ukraine. It was winter by the time Stein reached shelter in and old temple in Barkul. From there he surveyed Guchen (ancient names Chin-man and Pei-ting). Then he moved on through a pass (12,000 feet) in the T'ien-shan south to Turfan. At once he saw the striking difference in climate - the northern slope has conifer trees above the line of excellent grazing, but the southern slope is dry and bare.

Chapter XVII - Among the Ruins of Turfan
Stein reunited his survey teams at Kara-khoja in November 1914. His objective for the winter was detailed survey and excavations in the deep Turfan depression which lies between the T'ien-shan and the Kuruk-tagh ranges. In its deepest section it is 1000 feet below sea level. Its small rivers end in marshes and a dried up salt lake like the Lop sea. But there is sufficient run-off from the streams to provide irrigation for two annual crops, achieved by ingenious underground cisterns and tunnels. The surplus agricultural produce is traded for the produce of animal husbandry (wool, livestock, and all) supplied by the nomads across the low passes of the T'ien-shan. This was the economic basis all during Han times and again during that of the T'ang until the southern side (at Pei-t-ing) was taken in 790 AD by the Turks and Tibetans. A century later the Turkish Uigurs drove the Tibetans back and gained control of the entire Tarim with their capital in the Turfan area. They, in turn, were supplanted by the Mongols in the 13th century. (But it is Uigurs who remain the local population today under Chinese domination.) Early on the Uigurs became Buddhist, but during the Mongol period largely converted to Islam. Stein found many relics of Buddhist culture throughout. In contrast with the southern edge of the Tarim, the Turfan area never lost its irrigation nor succumbed to abandonment in the desert. Thus, Stein found such remains of pre-Islamic society as still existed within the same continuously cultivated zone. The easier access of the region to neighboring Russia also encouraged more extensive attention by Russian, German and Japanese archeologists. Nevertheless, Stein notes, his short visit in 1907 had revealed to him sufficient further opportunities to entice him back in 1915.
Stein divided the work. He sent Lal Singh off for extended surveys of the Kuruk-tagh while he and Afrazgul Khan surveyed the Turfan and dug into Buddhist ruins for 3.5 months. He began with Idikut-shahri near Kara-khoja. This was the capital (Kao-chang) during T'ang era as well as that of the Uigurs. There he found massive walls of stamped clay enclosing many ruined structures. Despite the extensive digging by German explorers followed by that of the locals realizing renewed interest in sales, Stein found by digging deeper enough artifacts to keep him happy. These included a large quantity of metal objects and Sung dynasty copper coins. Stein moved on to Toyuk village with its rock-cut grottoes. There the German explorers had been followed by locals eager to reap rewards, resulting in typical destruction of the sites from an archeological point of view. Even so the ever resourceful Stein found more hidden sites at which to employ another army of local diggers. From Toyuk he brought out frescoes and stucco relievos plus Chinese and Uigur manuscripts.
In December Stein moved on to Bezeklik. There he found more Buddhist temples decorated with frescos from the early Uigur era. The Germans had already made off with many frescos and the locals had followed them in destructive efforts. But Stein's assistants, Naik Shamsuddin and Afrazgul Khan, using the techniques Stein had developed at Miran still managed over 2 months effort to salvage valuable Buddhist art. All this filled another 100 large cases. He proudly states that even with the 3000 mile transport by camel, yak and pony over passes of 18,000 feet they reached India safely. During 1921-1928 they were housed in a specially built museum in New Delhi.
Over Christmas Stein went back across the mountains to Urumchi to see once again his early patron, P'an Ta-jen, who had moved successively from Amban at Khotan to Tao-tai at Aksu to Financial Commissioner of the "New Dominion' (Hsin-chiang). (Stein made a special trip across the desert from Khotan to Aksu during the second expedition for a similar reason.)
January 1915 found Stein at work at Murtuk and Astana. At the latter place he found a huge cemetery whose conical mounds revealed the grave sites. The tombs were in deep caves cut into the sandstone below entrance tunnels that required clearing. Most of the caves had been robbed for possible valuables including the very wood of the coffins. Nevertheless Stein with the sizable work force available around Turfan was able to reopen many tombs which contained a wealth of artifacts. He notes that the extremely dry climate had preserved the human remains and much more. He dated the remains to the early 7th century to middle 8th century, that is the era of the last local dynasty's rule and that of the resurgence of Chinese - T'ang - control in 640 AD. For the T'ang's Kao-chang, near Kara-khoja Astana, was a major government center. Stein describes the extensive quantity of varied artifacts recovered from these graves in far more detail and extent than I can include here. But one cannot skip noting the gold coins copied from Byzantine samples placed in the mouths of the dead and Persian silver coins from Sasanian kings placed in their eyes. Together with the previous materials the Astana articles filled 50 camel loads Stein dispatched early with Ibrahim Beg to Kashgar. He does not mention here that one reason for this rapid departure was his concern about growing Chinese official interest in what he was doing.
Stein concluded his work around Turfan in February at Yar-khoto, the capital of the region in Han era but found little remaining of interest.

Chapter XVIII - From the Kuruk-tagh to Kashgar
Stein continues to note his incapacitation due to the leg injury. Rai Bahadur Lal Singh returned in January from an almost 3 month survey expedition into the Kuruk-tagh. He had surveyed the mountains and the valley around Singer and on to the southeast toward Lou-lan to complete that section of the triangulations. He had also surveyed the desolate area northeast of Altmish-bulak with assistance from Abdurrahim and back again into the deep part of the Turfan. He again set off in early February to survey the western Kuruk-tagh. Stein sent Afrazgul Khan back toward Lou-lan to fill in missing survey areas. Stein started on 16 February with Abdurrahim's younger brother to explore the western Kuruk-tagh. From there with a fresh supply of ice Stein moved south to survey the 'Dry River" that had been the source of water for Lou-lan. Despite the beginning of the spring sand storms Stein continued and found several more cemeteries of local nomads greatly different from the Chinese who were passing by. At that point Stein was relieved of concern when Afrazgul Khan and Hassan Akhun returned from extensive surveying in the Lop desert and northeast of Lou-lan and then again to the southwest along the edge of the Lop Sea and then back again across the rocky plateau to the Kuruk-tagh. So both surveyors had spent the winter far into the desert with a few camels and helpers. The teams then moved west together to Ying-p'an where they found another ruined fort and temple. Stein found evidence of its Chinese garrison during Han times. From there Stein explored the route to Korla finding more fortified watch stations along the 100 mile long route. These massive signaling towers dated back to 100 BC. but were still in use in T'ang era.
Korla is the easternmost of the oases along the foothills of the T'ien-shan and on the main route for Hun and other raiders from the north to attack the Chinese at Lou-lan and along the caravan routes. It is but a half-day's ride south from Kara-shahr. The Kara-shahr valley is wide and fertile. Stein visited there Lake Baghrash-kol which supplies irrigation water to Korla and adjacent regions as well as flow to the Konche-darya that flows west. There also Stein found the extensive walls of Baghdad-shahri near the northern shore of the lake. Stein here jumps back again to the second expedition in December 1907 when he excavated the Buddhist shrines called Ming-oi. which he describes as long rows of detached cells. He found coins dating from the 9th century AD. It appeared to have been ruined by a huge fire, possibly resulting from the Mohammedan invasion. As usual underneath the debris Stein found stucco sculptures and relievos friezes. He was able to note clear comparisons with Graeco-Buddhist art from India but also evidence of influence of Gothic artistic style.
He again describes his activities from the second expedition in January 1907 when he was led into the desert south of Korla by hunters who swore than had seen 'ghost' cities amid the sand-dunes. He accompanied them into the area between the Inchike and Charchak rivers without seeing a single such city but being content to learn more about the geography of the river system. For their part the honest guides were disappointed that Stein's 'magic' had been unable to overcome the evil spirits that concealed these undoubted cities which by accident they had 'seen' during a sand-storm. He does not mention that on the second expedition this wild goose chase was immediately followed by his dramatic exploit of crossing the Takla Makan desert from north to south with the expectation (by no means assured, but realized none the less) of finding the end of the Keriya River where it disappears into the sand. He defers a brief mention to his discussion, further on, of Kucha.
Back on his third expedition again, Stein writes about April 1915. The teams reassembled there and started west for Kashgar. R. B. Lal Singh followed a northern route through the T'ien-Shan foothills. Muhammad Hakub, the second surveyor (who replaced the ill Rai Singh) went south across the Konche and Inchike Rivers to the Tarim to Yarkand. Stein kept himself to the main route in order to look for ancient sites among the oases. This was and remains the main merchant route, some 600 miles from Korla to Kashgar. The continued occupation of inhabited sites and climate conditions prevent the survival of much ancient ruins or artifacts. There was nothing at Bugur but in the desert he did find more ancient watch towers. At Kucha, the second largest of the oases of the T'ien-shan foothills and the crossroad where the east-west route intersects a north-south route from Dzungaria to Khotan. There he found Buddhist temples already explored by German, Russian and French expeditions and Stein had also already explored the place in 1907. No matter, Stein spent 3 weeks with Afrazgul Khan unearthing more material and surveying in detail the area around Kucha. He believed that the rivers that supplied irrigation to the extensive oasis area had greatly diminished in volume since T'ang times.
Here he returns again to January 1908 and inserts a comment about that 'risky journey' from the Tarim across the Takla Makan to the Keriya river and refers to his account in Desert Cathay. So in May 1915 Stein moved west from Kucha while Afrazgul Kahn went south to survey the route through Aksu. They both reached Aksu and continued west for 6 more days to Maral-bashi. He digresses again to mention that during the second expedition he had visited also Kelpin oasis in May 1908 and found another line of watch towers in the desert. With arrival at Faizabad Stein entered the expanse of the Kashgar oasis and reached Chini-bagh on 30 May.

Chapter XIX - From Kashgar to the Alichur Pamir
In his concluding chapters Stein describes his (to him) most interesting exit from Chinese Turkestan. At the conclusion of the first expedition Stein crossed into Russia and went by the shortest route directly to Samarkand and from there by rail and ship via Baku and Ukraine to London (Ancient Khotan) . At the end of his second expedition Stein detoured to explore the western Tibetan plateau behind the K'un-lun to find the sources of the Khotan rivers, an objective that had eluded him twice before. His account is in Serindia and Ruins of Desert Cathay. That effort required a month of traversing high passes and crossing desolate wastes. It concluded with him becoming frostbitten and being carried over the Karakorum on a litter. Better left unsaid.
So these chapters are about the first part of his excursions upon departing Kashgar for the third time. The concise account here is actually clearer and more easily understood than the more lengthy description in the official report. Unfortunately there was not there nor here a detailed map such as are essential to follow his trail throughout Turkestan. Of course he was not allowed to do surveying as he went. There were two part to this phase, first through the high Pamirs to Samarkand and second across Iran to explore Sistan in south-east Iran on the Afghan-Indian border. Only the first is recounted these three concluding chapters.
Back at the Macartney residence (Chini-bagh) in Kashgar Stein in June 1915 completed the administrative chores from his expedition. This included packing 182 heavy cases of acquired antiques for their long journey over the Karakorum to India and thence to London. And during this time the indestructible R. B. Lai Singh accomplished yet more survey work in the Pamirs and T'ien shan north of Muztagh-ata. The great improvement of British-Russian relations on the eve of World War I enabled him to secure excellent support from the Russian government for his proposed extensive tour through the Pamirs. His objective as always was to follow the ancient caravan routes (especially those of Marco Polo and Husein) through the Alai region. Special permission was granted during WWI in 1915 for Stein to travel through extensive regions of the Pamirs as far as the Afghan border on the Oxus. He was fortunate in the timing as no such permission was possible not long before nor after the Russian Revolution.
Stein departed Kashgar into the mountains on 6 July to rest at the Kirghiz camp at Bostan-arche. On 19 July Stein started west to the Ulugh-art pass (16,600 feet) with only Afrazgul Khan as assistant. They were again in a land of dangerous narrow canyons and steep ridges among glaciers. Naturally he thought of Ptolemy's description which he mentions several times. He traveled south-west, still in Chinese territory toward the Taghdumbash Pamir, which he had visited twice before. Riding through the gorge of the sources of the Kashgar River and then over the Kosh-bel pass at 13,800 feet he reached the Trans-Alai range with peaks over 20,000 feet. At that point he crossed the Russian frontier over the saddle at Kizil-art (14,000 feet) and reached the Russian border road between the Oxus and Farghana. He remarks on the pleasure of finding a real cart road after 2 years exploring.
At the border post at Por-dobe he met the Russian commander, Colonel Ivan Yagello. Stein profusely thanks the good colonel for the excellent arrangements the officer had already begun to insure that Stein's full support throughout Wakhan, Shughnan and Roshan. (that is throughout the Pamirs clear south to the Oxus river border). Stein always insisted that personal inspection on the spot was essential to gain an understanding of the geographical and topographical reality that medieval travelers confronted as described in their memoirs. He believed that the Alai valley was the principal route between Samarkand or Bactria and Khotan or Kashgar. This route left the Oxus via the Surkh-ab valley, the Kara-tegin valley and then up the Alai valley. Stein was determined to confirm all this by eyewitness observation. For centuries the Alai valley has been the summer grazing ground of the Kirghiz. Stein recalls his memory of seeing them between Irkesh-tam and Andijan in 1901. He notes that the extensive vegetation on both sides, except on the highest section of about 70 miles, would have enabled merchants traveling through in medieval times to find adequate food, water and shelter. But no longer, because now the western terminus at Balkh and Afghanistan no longer send or receive such trade. Now trade goes by the Terek pass (that Stein used in 1901) between Kashgar and Farghana (where there is a railroad).
Stein stopped next at the Kirghiz encampment, and Russian customs post at Daraut-kurghan. Nearby he found Chat, with its ruined fortification walls. From there he moved south across the snow covered ridges and watershed. This detour was strictly so he could enjoy the scenery of the high Pamirs. Even with Kirghiz ponies ordered up by Colonel Yagello the route was extremely difficult. He describes his enjoyment.
"Rarely have my eyes in the Himalaya, Hindukush, or K'un-lun beheld a sight more imposing than the huge glacier-furrowed wall of the Muztagh as it rose before me with magnificent abruptness above the wide torrent bed of the Muk-su, after I had crossed the Trans-Alai by the Tarsagar, our first pass from Daraut-kurgan. Its boldly serrated crest line seemed to rise well above 21,000 feet, and individual ice-clad peaks to reach a great height above it."
A huge flood blocked the main route Stein wanted to use to reach the Takhta-koram pass. Instead he was forced to cross the Kayindi pass at 15,100 feet over difficult moraines. Once over that he found open valleys toward Sel-dara. Then he mounted the Takhta-koram at 15,000 feet. On 8 August he crossed the Kizil-bel saddle at 14,000 feet to meet the local Kirghiz Ming-bashi, Kokan Beg, from whom he could secure the next relay of ponies. Although he could not do topographic survey of the terrain, Stein did not refrain from his usual anthropological measurement of the various tribesmen he met. He was always on the lookout for representatives of what he termed "Homo Alpinus" a race of hardy mountaineers he believed occupied all the high valleys from the Pamirs through the K'un-lun. His path was blocked again by a huge lake in the Sarwz Pamir created by an earthquake of February 1911. His account of the difficulties overcome in moving through narrow gorges and over rock ridges in Serindia is more vivid than in this book. At any rate he moved along the Tanimuz River crossing it back and forth and finally reached Pasor on 12 August where the population was Tajik rather than Kirghiz. They next reached Saunab (Tash-kurghan) in Roshan a settlement of the Iranian speaking Ghalchas. He secured a new team of mountain porters. Again, more work on anthropological measurement. Further movement south required passage through more narrow gorges and defiles along the Murghab River and over rock strewn ridges. Stein describes the scene he witnessed that was created by the massive earthquake which sent a whole mountain down into valleys. He crossed another pass at 13,200 feet. His Roshani hillmen acting as porters had to build by hand 'rafaks' that is ledges made of wood and stone as cornices attached to sheer rock walls to enable the party to get through. At one point it required 5 hours to advance one mile. Reaching the Langar Pass at 15,400 feet on 20 August the Roshani were replaced by Kirghiz set up by Russian order.

Chapter XX - By the Uppermost Oxus
The next two days were spent crossing the Alichur Pamir valley which stretches over 60 miles east to west. Stein was again on a medieval trade route between Shughnan and Khotan. Stein again recalls the exploits of General Kao-Hsien-chih who brought his Chinese army through the valley and on to the Darkot pass in 747 AD. Shughnan was also the route of Chinese pilgrim monks to and back from India. The Manchu conquerors of China also sent their soldiers to the Alichur Pamir. Stein found the base of a memorial stele built at a shrine by the lake to commemorate the Manchu victory there in 1759. But the upper part with inscription had been removed by Cossacks to the museum in Tashkent after they too scored a victory on the spot over Afghans. Stein rode for two more days through the Alichur Pamir to Bash-gumbaz-aghzi, the summer camp of the Kirghiz. Another nomadic tribe, another anthropological measurement filled a day. Then it was time to cross another high pass from the Alichur to the Great Pamir on 26 August, at 16,300 feet. Stein again savors the panoramic view of the Great Pamir and Lake Victoria at the border of Russia and Afghanistan. Yes, both Hsuan-tsang and Marco Polo had shared this view. On 27 August Stein camped by Lake Victoria at 14,000 feet with the temperature down to 12 degrees below freezing. He waxes poetic:
"In the peace around, undisturbed by any sign of human activity past or present, it was easy to lose all count of time and to feel as if spiritual emanations of those cherished old patrons of my travels were still clinging to the scene." He continues with explicit references and quotations of the very accurate descriptions of the lake written by both Hsuan and Marco.
Stein notes that Afrazgul Khan shot a mountain sheep and give him a head as a souvenir. From Lake Victoria Stein rode for three days along the right bank of the Great Pamir branch of the Oxus ( the border of Afghanistan he could not cross) to the Wakhan corridor at Langar-kisht where he was welcomed by the local Russian guard post commander. He could see the peaks of the Hindukush across the narrow strip of Afghan territory along the left bank of the Ab-i-Panja River. There also he met the Wakhi Ming-bashi, Sarbuland Khan, on the Russian side and discovered that it was this gentleman's own son who had helped Stein cross the Chillinji Pass into Hunza in 1906. Stein recalls that expedition in which by special favor he had been allowed into the upper reaches of the Wakhan at Sarhand and assisted in crossing the Wakhir. But on this occasion Afghan- British relations had again deteriorated preventing Stein from setting foot in Afghanistan. All he could do was look longingly from across the Oxus and the fall harvest opposite and the mighty Hindukush peaks over 22,000 feet high behind it as he rode down the Russian bank and crossed many foothills. In this he notes his resemblance to Sung Yun, another Chinese pilgrim and to the Jesuit Benedict Goes who rode up the Wakhan in 1602. Not to worry too much, Stein found ample subjects for his anthropological study in the various tribes separated in their individual valleys.
For this reader his mention of the numerous fortresses and outposts he studied, each on its own crag or precipice, is most intriguing and frustrating. He comments. "But what claimed my attention most were the ruins of ancient strongholds, some of them of considerable extent and in part remarkably well preserved, to be found on hill spurs over looking the valley. There was much of antiquarian interest to be observed in their plans, the construction and decoration of their bastioned walls. The natural protection afforded by unscaleable rock faces of spurs and ravines had always been utilized with skill in these defenses. "
Indeed, but he continues.
"This is not the place to describe them in detail, nor need I set forth the reasons which even in the absence of direct archeological evidence such as only excavations could produce lead me to believe, that several of these fastnesses go back to a period roughly corresponding to Sasanian domination or possibly even somewhat earlier. "
Well, he does not describe them in sufficient detail in his official reports and narratives either. One wishes to be able to see the excellent diagrams such as he produced for the forts and towers in Turkestan, but clearly such work was not possible while he was in Russia under special permission. Have they ever been studied in detail with results published?
Ok, he does describe one fortress, Samr-i-atish-parast (The Zamr of the Fire-worshiper), to some extent. It was constructed of successive walls with numerous bastions and turrets built of rough stones or large sun-dried bricks up the slope of a precipitous spur over 1000 feet high and with a circumference of over 3 miles. What a missing picture. Stein considered that the fortress was built as a place of refuge against the frequent invasions mostly from the west. He always draws broad conclusions, in this case, that the construction is evidence that the population of the valley in ancient times greatly exceeded that of today. He mentions that he was informed that the current population on the Russian side of the river was about 3000.
Further down the Wakhan corridor Stein passed into Ishkashm, separated from the main Wakhan by rocky defiles but again specifically mentioned by Hsuan-tsang and Marco Polo. There he found a ruined fortress known as the "Castle of Qa-qa" near Namadgut. It had massive walls thirty feet thick composed of sun-dried bricks on the crest of two adjacent ridges. This one was nearly a mile long with a citadel at the western end. After another day's ride he reached Nut, opposite a larger village on the Afghan side, and where he was welcomed by Russian Captain Tumanovich. Over a two day stay, Stein collected samples of the Ishkashmi, a Eastern Iranian language retained by isolated hill communities. Linguistic surveys was Stein's other subsidiary hobby in addition to anthropological survey, both to be passed on to specialists in the subjects. From Nut Stein continued down the Oxus through Gharan. This isolated valley was the source of Badakhshan rubies noted by Marco Polo. Stein made note of the pits. On September 12 he reached the valley where the Shughnan rivers meet the Oxus. Here was the headquarters of the Russian military "Pamir Division" in Khoruk. Stein met his Russian host, Colonel Yagello. Stein enjoyed another 2-day halt. He confirmed that current accounts ascribe to the Shughni the same warlike and generally violent disposition as recorded by Hsuan-tsang. But raids and invasions are now of the past since the advent of strong Afghan and Russian government control. From Khoruk Stein rode through Shakh-dara to its junction with the Alichur Pamir and then into the Ghund valley. Again, along the route he passed numerous forts and outposts one would like to learn more about. All along he noted the improvement of local conditions as a result of the Russian suppression of raids and exactions.

Chapter XXI - From Roshan to Samarkand
Once again Stein notes that he could have reached Roshan by a direct and easy route, but rather wanted to see the high mountains. So he left ponies at 12,600 feet and enlisted mountain men to carry his luggage up treacherous rock filled ravines and glaciers over a 16,100 foot pass affording 'magnificent views' that were 'a fitting reward'. (For him but I suspect not for the coolies.) After descending along a glacier to a level spot Stein exchanged the hardy Shughnan carriers for a fresh team of strong Roshanis who were waiting. Both teams were courtesy of Colonel Yagello's orders. A day passed for the movement down the Raumedh valley and over more moraines and into the Bartang river valley.
Stein, comments: The march "sufficed to impress me with the exceptional difficulties offered to traffic by the tortuous gorges in which the Bartang River has cut its way down to the Oxus. I now understood why Roshan has always remained the least accessible of all the valleys descending from the Pamirs and why in the stock of its people and in its traditional ways it has retained most of its early inheritance."
From there the route led through more "narrow, deep-cut gorges" and between towering mountains. He crossed the Khaizhez river on a goat-skin raft and then climbed up and down rock faces along narrow ledges or by using footholds a few inches wide. He frequently returned to the rafts pushed by swimmers through the swirling water as he watched the Roshanis mountaineers high above carrying his baggage across sheer rock walls. He notes "they looked like spiders." He passed or stayed at tiny hamlets of rubble-built hovels placed at the mouth of ravines. He used every opportunity to examine the men, finding them examples of the same ancient Homo Alpinus encountered throughout the mountains from west to east.
"Among the Iranian-speaking hillmen of the valleys I traversed in the Oxus region, the people of Roshan seemed to me to have preserved the Homo Alpinus type in it greatest purity."
Stein passed through more gorges "where the track clings to almost vertical rock faces by frail wooden 'Rafaks' or ladders". After two more days they reached Kala-i-Wamar, the main village of Roshan, where he stopped for more anthropological measurements near a ruined castle held by the Mirs of Shughnan. He noted that the wooden decorations of the homes the style was that of the Graeco-Buddhist relievos of Gandhara, Niya and Lou-lan. He departed Kala-i-Wamar on 27 September toward the Kara-tegin across still more ridges that divide Roshan from Yazgulam, via the Adude Pass, a glaciated saddle at 14,500 feet.
At nightfall he reached Matraun where he was greeted by waiting officials from Bukhara alerted by Colonel Yagello. Stein notes that the inhabitants controlled by Bukhara are Sunni, but those of Shughnan and Roshan are Ismaili ( the sect of the Aga Khan). Passing down the Yazgulam Stein again reached the Oxus at the Wanj valley. On 1 October he moved up that valley through orchards. There the population was Persian speaking Tajiks. Rain and snow forced him to stop at Sitargh. By providing the village headmen with a written statement that they were not to be held responsible for damages Stein arranged for transport in the snow across the Sitargh pass at 14,600 feet to Khingab. Once across he walked down between crevasses past the glacier to find a small camping site. On 4 October he reached Pashmghar, the highest village in the Khingab and on to reach Kara-tegin. Stein commented on the maladministration from Bukhara that exacted the meager wealth from the villages he passed in the guise of 'gifts' passed from one chief to the next. But worse was coming all too soon with the Soviet conquest. Rain and snow again forced a halt at Lajirkh on 6 October. Two days were spent in crossing the Girdan-i-kaftar pass and Tupchak plateau. After this meandering travel for 2 months through gorges and over passes Stein was back again on the ancient route to the Alai. Here the harvest was still under way at 8000 feet elevation. At Kara-tegin he was again with Kirghiz settled here in villages. But they were being pushed out by the Tajiks. For centuries the area had been a meeting place for the Turki Kirghiz and the Iranian Tajiks. On 11 October Stein stopped at Gharm, the regional headquarters. After two more days he reached Ab-i-garm where the Ssurkh-ab flowing south enters impassable gorges. He turned west into the open plain of the Hissar. In 9 more days of travel he covered 270 miles through Uzbek lands. From Hissar, Stein notes, he could have taken the easy route southwest to reach the main highway between Termez and Samarkand, but true to form he instead went northwest through more mountains past Kash-kurghan and over the snow covered Karkhush Pass to Shahr-i-sabz which he reached on 20 October. The following day he reached Samarkand and ruined Afrasiab (Markanda).