ANCIENT KHOTAN

Detailed Report of Archeological Explorations in Chinese Turkestan

Sir M. Aurel Stein
Oxford Clarendon Press, 1907

This is the official report on Stein's first expedition in 1900-01 from India into Chinese Turkestan. Since this expedition was not nearly as lengthy in time or space as the second and third expeditions, this report is shorter than Serindia or Innermostasia. Nevertheless it is full of detailed scholarly footnotes and the second volume, attached, contains the many photos of artifacts and diagrams Stein drew which full several volumes in the larger official records. Copies of some of the illustrations are here. Stein repeatedly also refers the general reader to his 'personal account' of this trip, Sand-buried Ruins of Khotan in which he comments further on the expedition.
The text is a combination of sections in which Stein records in detail his daily activities in chronological order and sections in which he writes general commentary and scholarly conclusions based on information including material learned much later during the expedition. From the Table of Contents one can see how carefully he has organized the material into these separate sections . The expedition is included also in Stein's later summary work - On Ancient Central Asian Tracks

Readers may find the extensive sections on the history of Khotan of interest apart from the archeological expedition. Stein has brought together as much of the history as he found in Chinese annals as well as medieval and modern visitors and authors. The chapters on the desert ruins at Dandan, Niya and Endere are also interesting from the historical point of view apart from the details of archeological finds. He based his exploration during all three tours on the British consulate at Kashgar. Unfortunately although he did include historical information on Kashgar, especially its various previous names, he did not survey the city or describe its physical layout and fortifications, which have now been destroyed by 'urban renewal'. Stein constantly provides names for locations, both as he visits them and in his historical digressions. One is well advised to keep maps handy for reference. The most detailed of these maps are those he included in Serindia. and Innermostasia

As an archeologist Stein is meticulous in establishing personally the exact relative location and time each artifact was recovered. He carefully distinguishes between items personally found and those purchased from locals. As a geographer and topographical engineer he kept detailed records of temperature, elevations, weather, and hydrology; and created excellent site plans and detailed full scale maps. As a student of Indian (Buddhist in particular) art and culture he continually comments of the connections between the sites he excavated and Gandhara (the Graeco-Buddhist area now in Pakistan). He seeks to find connections also with ancient Rome, Scythia, Sogdiana, Afghanistan and the West in general. As a student of history he traces the history of each place through Chinese or other official documents and traveler's memoirs. As a student of linguistics and anthropology he listens to the speech of locals and takes measurements of their facial features to provide to experts later. As a leader of a large and varied team he spends hours keeping financial and other official records. As the organizer he plans for months and even years ahead to order transportation and logistic support. As a visitor to many different ethnic peoples he is unfailingly able to fit in and treat everyone with great respect. These peoples include the many different groups in India (Sikhs, Gurkhas, Punjabi, and others), various relatively primitive mountain tribes, Afghans, Sarikolis. Turki, Chinese, Uigurs, Khotanese, Lopniks, and numerous others. As a correspondent he devotes hours to writing letters both official and personal to a vast group of government officers and friends in India, England, and throughout Europe. As a military leader one would note that he was his own S-1, S-2, S-3, and S-4, enginner, transportation, intelligence and personnel officers.

Stein is always effusive in his acknowledgment of the permission and support he received from the British Government in India and London, and the contributions of many scholars who have studied the massive collection of artifacts he brought out of the deserts and mountains of Turkestan. Much of their contributions is included in the appendices and also in the footnotes. He dedicated this volume to Sir Henry Yule, among many other things the author of the definitive edition of Marco Polo's memoir. Stein remarks that he carried Yule's volume with him always, along with his favorite Chinese memoirs. He also unfailingly gives credit to the numerous Indian, Afghan, Yarkandi, Chinese and other associates, helpers, and servants who accompanied him and performed heroic work under trying conditions.

This expedition was conducted, as noted, in 1900-1901 but this account was not published until 1907 due to the extensive work for its preparation amid his other duties. Meanwhile he did publish the more popular style personal narrative Sand-burried Ruins of Khotan, in 1903. He wrote this introduction in July 1906 while on his way back to Khotan during his second expedition. Therefor it also contains Stein's comments based on further exploration. I consider it amazing that all during this and the subsequent explorations into desolate mountains and deserts Stein remained in contact with India and England via post carried by 'dak' runners across this terrain. Thus he is able to comment, while exploring glaciers at 15-17,000 feet elevation, when he takes a break for a day to work on the proofs of an edition sent to him from London and then send them back directly by postal courier. .

Table of Contents - pgs v - xv

Addenda and Corrigenda - pgs xv - xvi

List of Illustrations, - pgs xx - xxi

List of Abbreviations Titles (bibliography) - pgs xxii - xxiv

Chapter I - From Kashmir to the Pamirs - pgs 1 -21
Section I - The Gilgit Road and Kisanganga Valley - pg 1
Section II - Chinese Historical Relations with Gilgit - pg 4
Section III - The Chinese Occupation of Gilgit and Route to Kashmir - pg 8
Section IV - Ancient Remains in Gilgit and Hunza - pg 17

Chapter II - Sarikol and the Route to Kashgar - pgs 22 - 46
Section I - The Geographical Position and Ethnography of Sarikol - pg 22
Section II - Early Chinese Accounts of Sarikol - pg 27
Section III - Historical Sites of Sarikol - pg 33
Section VI - From Sarikol to Kashgar - pg 40

Chapter III - Historical Notes of Kashgar - pgs 47 - 72
Section I - The Old Names of Kashgar - pg 47
Section II - Kashgar During the Han epoch - 52
Section III - Eastern Turkestan Under the Ta'ngs - 57
Section IV - Notices of Kashgar During the T'ang Period - 65

Chapter IV - The Ancient Remains of Kashgar and the Oases of Yarkand and Karghalik - pgs 73- 93
Section I - The Stupas of Kurgahn-Tim and Kizol-Debe - pg 73
Section II - The Ruins near Khan-ui - pg 79
Section III - The oasis and City of Yarkand - pg 86
Section IV - Khargalik in Chinese Records - p 89

Chapter V - The Route from Karghalik to Khotan: Its Ancient Topography and Remains - pgs 94 - 122
Section I - By Desert Edge to Khotan - pg 94
Section II - The Oasis of Guma - pg 99
Section III - The Tati of Kakshal - pg 103
Section IV - The Ancient Remains at Moji - pg 110
Section V - From Moji to Khotan Oasis - pg 115

Chapter VI - The Khotan Oasis: its Geography and People - pgs 123 - 150
Section I - The Oasis in its Geographical Features - pg 123
Section II - Agriculture and Industries In Khotan - pg 130
Section III - The Population of Khotan: its Distribution and Character - pg 136
Section IV - The Population of Khotan: its Physical Characteristics and Racial Origin - pg 143

Chapter VII - Historical Notices of Khotan - pgs 151 - 184
Section I - The Early Records and Names of Khotan - pg 151
Section II - The Legendary Traditions of Khotan - pg 156
Section III - Khotan in Chinese Records, From Han to Sui Dynasty - pg 166
Section IV - Khotan during the T'ang Period - pg 172
Section V - Later Chinese records of Khotan - pg 177

Chapter VIII - Ancient Sites of the Khotan Oasis - pgs 185 - 235
Section I - The Hill of Gosrnga - pg 185
Section II - The Culture-strata of Yotkan - pg 190
Section III - The Site of the Ancient Capital - pg 199
Section IV - Antiques acquired from Yoktan and Khotan - pg 206
Section V - Buddhist Sites Described by Hsuan-tsang - pg 223

Chapter IX - The Ruins of Dandan-Uiliq - pgs 236 - 303
Section I - The Desert March to the Site - pg 236
Section II - First Excavation of Buddhist Shrines - pg 242
Section III - Art Relics of Site Dii - pg 249
Section IV - First Finds of Ancient Manuscripts - pg 256
Section V - Discovery of Dated Documents - pg 264
Section VI - Records of the Hu-kuo Convent - pg 273
Section VII - Other Ruins of Dandan-Uiliq and General Observations on Site - pg 281
Section VIII - List of Objects excavated or found at Dandan-Uiliq - pg 288

Chapter X - From Dandan-Uiliq to the Niya River - pgs 304 - 315
Section I - The Rawak Site - pg 304
Section II - A Judeo-Persian Document - pg 306
Section III - Keriya, Niya, and Iman Ja'far Sadiq - pg 309

Chapter XI - The Ancient Site Beyond the Niya River - pgs 316 - 416
Section I - The Ruin Ni and the First Finds of Inscribed Tablets - pg 316
Section II - Excavation of Ancient Residences, Nii, Niii, Niv - pg 328
Section III - Discoveries in an Ancient Rubbish Heap Nxv - pg 338
Section IV - Ancient Documents on Wood and Leather - pg 344
Section V - Chinese Documents from Nxv and the Writing on Wood - pg 358
Section VI - Decipherment of Ancient Documents Kharosthi and Chinese - 363
Section VII - Exploration of Ruins Nvi-xii and General Observations on Site - 374
Section VIII - List of Antiques Excavated at Niya Site - pg 385

Chapter XII - The Endere Ruins - pgs 417 - 442
Section I - The March to the Endere River - pg 417
Section II - Excavation of the Endere Temple - pg 421
Section III - The Ruined Fort and Stupa of the Endere Site - pg 430
Section IV - The List of Antiques from the Endere Ruins - pg 438

Chapter XIII - Khara-dong and the Search for Hsuan-tsang 's P'i-mo - pgs 443 - 469
Section I - Expedition to the Kara-dong Ruins - pg 443
Section II - Hsuan-tsang's Notice of P'i-mo and Marco Polo's Pein - pg 452
Section III - The Sites Uzun-Tati and Ulugh-Ziarat - pg 457
Section IV - The Tuga-dong Mounds and the Keriya- Khotan Route - pg 465

Chapter XIV - The Ruins of Ak-sipil and Rawak - pgs 470 - 506
Section I - Hanguya Tati and the Site of Tam-Oghil - pg 470
Section II - The Remains of Ak-sipil and Kighilik - pg 474
Section III - The Rawak Stupa - pg 482
Section IV - The Sculptures of the Rawak Vihara - pg 488
Section V - The Date of the Rawak Remains; The Jumbe-kum Site - pg 500

Chapter XV - Departure from Khotan - pgs 507 - 520
Section I - Islam Akhun and his Forgeries - pg 507
Section II - Last Days at the Khotan Oasis - pg 514
Section III - From Khotan to London - pg 517

Appendices
Appendix A - Chinese Documents From the Sites of Dandan-Uiliq, Niya and Endere - pgs 521 - 547
Appendix B - Tibetan Manuscripts Sgraffiti Discovered at Endere - pgs 548 - 569
Appendix C - The Judeo-Persian Document From Dandan-Uiliq - pgs 570 - 574
Appendix D - Inventory List of Coins found or Purchased - pgs 575 - 580
Appendix E - Extracts from Tibetan Accounts of Khotan - pgs 581 - 585
Appendix F - Notes on Specimens of Ancient Stucco from Khotan Sites - pgs 586 - 587
Appendix G - Notes on Sand and Loess Specimens from the Region of Khotan - pgs 588 - 590
Index of Objects found, purchased, etc - pgs 591 - 596
Plates Described or Referred to in text - pgs 597 - 598
General Index - pgs 599 - 621

Chapter I - From Kashmir to the Pamirs

Section I - The Gilgit Road and Kisanganga Valley

Stein begins with an explanation of his long-term desire to explore Turkestan and his choice of the Route through Gilgit as the shortest way to Kashgar. He always had in mind accomplishing many tasks simultaneously including - archeology, topographic surveying, geographical study, ethnology, linguistics and more. He discusses the geography of Kashmir and the route into Central Asia in terms of the historical trade routes. He mentions the Tragbal, Razdiangan, and Dudhhuk Passes and the Dard occupation of the region. During his several expeditions he was much interested in the Dards, an ethnic group that occupied the remote north-western valleys of India toward Tibet and Afghanistn. He refers to Sanskrit documents. He notes that the Chinese pilgrim, Wu-k'ung, visited Kashmir from Gandhara from 759 - 763 AD. and prepared a topographic account that Stein has fully proved accurate.
There were three roads out of Kashmir, east, southwest and north.

Section II - Chinese Historical Relations with Gilgit

Stein turns to Chinese records - The Official Annals of the T'ang Dynasty. After the destruction of the empire of the Western Turks - 658-659 AD The Chinese gained control not only over the whole Tarim Basin but further into the Oxus River valley and southwestward across the mountains into Kashmir. But the Tibetans increased their power and occupied Kashgar and the Tarim Basin from 670 to 692 AD. Then the Arabs under command of Quyatba ibn Muslim in 705-715 were victorious and expanded eastward along the Oxus River and into Sogdiana. The Chinese drove the Tibetans out of the Tarim Basin but the Tibetans then sought alliance with the Arabs across the Pamir Mountains. Chinese Emperor Hsuan-tsung (713-755) developed diplomatic and military campaigns in which Gilgit played a strategic role. Gilgit was central to communications between the Upper Indus via Yasin and the Baroghil Pass to the Central Pamirs and Oxus River valley. The Chinese strategic policy was to keep the Tibetans from reaching the Arabs. For this the Chinese wanted to keep Udyana (Swat valley) away from the Tibetans.
At the time of Hsuan-tsang's visit around 631AD the frontier of Udyana extended northeast as far as Darel, near Chilas.. Thus the Chinese placed great strategic importance on Gilgit. In 722 the Chinese sent a small army of 4000 troops from Kashgar to aid Gilgit against the Tibetans. Another campaign took place in 737. But later the Tibetans succeeded in gaining control over the small 'kingdoms' in and around Gilgit.

Section III - The Chinese Occupation of Gilgit and the Route to Kashmir

The first three Chinese expeditions to regain Gilgit were failures. Then came a fourth expedition in 747 led by General Kao Hsien-chih. General Kao Hsien-chih was specially appointed by Emperor Hsuan-tsang with a force of 10,000 cavalry and infantry. They started from An-hsi (Kucha) and reached Kashgar in 35 days. They reached Tash kurghan in Sarikol in another 20 days. In another 40 days they crossed the Pamirs. Then they turned back east going up the Oxus River valley through the Wakan corridor to the Tibetan fortress called Lien-yun, which is identified now with the village Sarhad on the direct route over the Baroghil Pass and toward Gilgit and Chitral.. The Tibetans had a 1,000 man garrison in Kien-yun and the river was in flood. The Chinese crossed the river and bypassed the fort to engage a Tibetan army of 8-9000 men in the hills. The Chinese gained a great victory. The general left 3,000 men to guard the fort and led his main body over the Baroghil (12,460 feet elevation) and down into the narrow valley to the Mastji River and then over the Darkhot Pass ( 15,200 feet) to the Yasin River and into Gilgit. Stein matches all the Chinese names with the modern names of the various places. The Chinese subdued the Yasin valley and broke the key bridge over which the Tibetan relief army must cross. The general then returned back the way he came. But he had to return in 750 to restore order. However in 751 Kao Hsien-chih was soundly defeated by the Arabs who were over running all of Sogdiana. This forced the Chinese to abandon Gilgit and all western - southwestern areas.
Stein discusses the logistic problems of the Chinese in maintaining a garrison in Gilgit and notes that up to the building of a modern highway there from Kashmir the same problems have existed to the present.

Section IV - Ancient Remains in Gilgit and Hunza

Stein writes that the absence of written records make reconstruction of the history of Gilgit and Hunza very difficult. He passed through the area too rapidly to conduct local archeological study. He found a few Buddhist ruins including a large rock-carved relief near Gilgit Fort. Stein compares this representation with statues of Buddha in Dandan-uiliq. Stein then describes his view of the Hunza valley through which he traveled to the Taghdumbash Pamir. The valley is so isolated by high mountain ranges that it has escaped history. The Burisheski language spoken there has no relation to any of the other languages in NW India or adjacent regions. He did find one Buddhist stupa.

Chapter II - Sarikol and the Route to Kashgar

Section I - The Geographical Position and Ethnography of Sarikol

Stein records his passage over the Kilkut Pass (15,800 feet) in theTaghdumbash Pamir on 29 June. This is close to the location where the Indus, Oxus and Tarim rivers all have their sources. He viewed the area from Kushbel Peak and established a survey station. This is also the meeting place of India, Afghnistan, Russia and China. Although the other Pamir ranges are in the Oxus river watershed, the Taghdumbash Pamir's streams flow eastward into the Tarim basin by Tash-kurghan in the Sarikol mountain district. Tash-kurghan is located where all the routes east to west between the Tarim and Oxus meet in a broad valley. Stein discusses the geographical position and ethnography of Sarikol. Sarikol could support a larger agricultural population but it has been subjected over many years to attacks by warlike tribes hunting for slaves. The local population are Tajiks plus visiting nomad Kirghiz. They speak an Iranian dialect.

Section II - Early Chinese Accounts of Sarikol

Stein again reverts to Chinese archival records to discuss medieval Chinese views of Sarikol. His favorite pilgrim, Hsuan-tsang, passed through Sarikol in summer of 642 AD on his return trip from India and duly recorded his observations. So did several well known pilgrims. The area also appears in the T'ang Dynasty annals. Stein describes all this in great detail. He traces Hsuan-tsang's route by several possible passes leading to Sarikol in the process of which he gives us considerable information about all the passes (their heights) and the glaciers and river gorges in the Great and Little Pamirs. He notes that the Wakhjir Pass (16,200) feet has always been a main part of the merchant caravan's route between Khotan and Badakhshan.

Section III - Historical Sites of Sarikol

Stein turns to description of the historical sites in Sarikol (called Chien-p'an-t'o in Chinese records), once again relying on Hsuan-tsang's reports. Tash-kurghan was the ancient capital and now has a modern Chinese fortress (occupied in 1892) on a long plateau next to the Taghdumbash River. This river is a tributary of the Yarkand River to the east. The fort is an irregular quadrangle with circumference of about a mile. The walls are of sun-dried brick but the ancient earlier walls include stone. It suffered major damage in an earthquake in 1895. The town also had walls now damaged.

Section IV - From Sarikol to Kashgar

Stein resumes his travel - from Sarikol to Kashgar. Naturally Stein would not take the standard, easier route via the Chichiklik Pass, but wanted to climb the imposing mountain, Muztagh-Ata. This required him to traverse narrow mountain valleys and climb over high passes on a longer route. This passed Muztagh-Ata ( 24,321 feet) which he tried to climb (he managed up to 20,000 feet before being blocked) and then passed through the Gez defile, to Tashmalik and then to Kashgar. He believes that Marco Polo followed mostly on this route. But he believes Hsuan-tsang used the route over the Chichiklik Pass and along the foothills north to Kashgar.

Chapter III - Historical Notices of Kashgar

Section I - The Old Names of Kashgar

Stein reached Kashgar on 29 July. This section is his digression about the city, starting with a discussion of its various names over the centuries. Stein notes that considering the occupation and importance of this location in ancient times, there is scant archeological evidence from which he can gain an understanding of the city. (Actually at that time the city still had its medieval fortifications but Stein didn't want to appear too nosy in a Chinese city.) In this section he focuses on the various names ascribed to the place and comments on the variety of ideas expressed by scholars about connections with various possible sources.

Section II - Kashgar during the Han Epoch

In this section Stein turns to the written records of the Han Dynasty. He notes that the earliest of these relates to the Chinese expansion into the Tarim under the Emperor Wu Ti (140 - 87 BC). This took place after the envoy, Chang Ch'ien, (139-127 BC) returned from his remarkable journey clear to the Oxus and reported on the commercial opportunities that could be developed by trade across the Tarim basin. He described the northern route, along the foothills of the T'ien Shan but first through the narrow corridor west of Sha'chou. Stein notes that this route, still the most frequently used ,is from Sha-chou to Hami, Turfan, Kara- shahr, Kucha, Aksu, to Kashgar. This route then described as passing Su-le and then over the Ts'ung-ling Mountains to Ta-wan. The area named now Sogdiana.
The Chinese government suffered decline during the reign of the usurper Wang Mang (9-23 AD) and lost control of Central Asia. Under Emperor Mig Ti, the second Later Han emperor (58-75 AD) Chinese power expanded again. By then Sule and other small domains in the western Tarim had been brought under control by the King of Khotan. In 70 AD general Pan Ch'ao conquered Khotan and by 76 AD the whole of the Tarim Basin except modern area of Kara-shar and Kucha. We do not know if the route taken by Pan Ch'ao in 95AD went through Kashgar. Kao Hsien-chih did pass through Khotan in 747AD.
Pan Ch'ao's campaign was a great Chinese victory and expansion of power into the west. This was important for the development of the famous trade route with the Romans. This led to the knowledge of western China shown by Marinus of Tyre and then preserved by Ptolemy. Marinus described the land of the Seres from whence came silk. Stein provides the detail of Ptolemy's description of the route. He credits Sir Heny Yule and Sir H. Rawlinson with identifying the places names. He identifies Ptolemy's Stone Tower with the then current Russian customs station at Irkeshtam. This route is the most direct between Kashgar and ancient Baktria. From the Ancient Han annalys we learn that in 107-113 AD An-kuio of Su'le sent hostage to the ruler of the Great Yueh-chih, who ruled over Baktria and eastern Afghanistan and then to Gandahara. It was during this period that Buddhism spread to Kashgar.

Section III - Eastern Turkestan under the T'angs

Stein shifts his focus to the reign of the T'angs. He continues, noting that the loss of Chinese control under Emperor An Ti (107-125 AD) resulted in a lack also of Chinese records of the Kashgar region. During the following 500 years the Uigurs rose to power around Turfan and Hami while Kashgar was dominated by the Yueh-chih. Likewise, during the periods of the last Han Emperor Hsien Ti (190-220) and the Three Kingdoms (221-265) contact with Kashgar and western Tarim basin was lost. Emperor Wu Ti (265-290) may have attempted to regain control. But the Chinese remained to weak until the T'ang emperors (619 AD). During 563-367 the empire of the Hephthalites was conquered by the Western Turks - Tu'chueh - under Khakan Istami. With the T'angs Chinese power grew again. They defeated the Northern Turks in 630 and then in 640 sent an imperial Chinese army across the desert to Turfan. The Chinese set up a protectorate centered at An-hsi. Chinese campaigns continued until by 658 they succeeded in control over the4 entire Tarim basin and across the mountains as far as Kabul and to Persia, though client kingdoms. In 658 the control center was moved from An-hsi to Kucha. But Chinese control was still maintained through local rulers who offered submission to the emperor. For instance there was an embassy from Su-le - the king of Kashgar - in 635 and another in 639 offering submission. Other embassies were recorded from Kara-shahr, Karghalik, Turfan, Kucha, Khotan, and even Sarikol.
But in 662 there was a rebellion of the Western Turks against which the Chinese sent an army but they ran into Tibetans who soundly defeated them. The Tibetans had conquered the Kuku-Nor region in 663 became serious rivals of the Chinese in Central Asia. In 665 they attacked Khotan which was relieved by the Chinese. In 670 the Chinese suffered a major defeat by the Tibetans north of Kuku-Nor. The Tibetans went on to capture Kashgar in 676-78 and held the whole region until 692. That year the Chinese general Wang Hsioa-chieh regained control over An-hsi and Kucha with a garrison of 30,000 men. For the next 50 years theChinese continued to attempt control over all the Tarim to west and south-west. But now they faced also the Arabs to the west and the Tibetans to the south. The struggle continued during the reign of Emperor Hsuan-tsung (713-762). Then Chinese policy became defensive. For one thing they faced renewed attack from the Northern Turks, and for another there was internal governmental weakness. Between 705 and 715 the Arab armies of Qutayba iben Muslim swept over Sogdiana and the Oxus region and to Farghana. Arab and Chinese conflict resulted in Farghana. Another rebellion of the Western Turkish tribe, Turgash, in 717 brought Chinese conflict with both Arabs and Tibetans who even besieged Turan and Ak-su. At t he same time the Tibetans expanded control into the southern border of the Tarim such as at Endere. Nevertheless Emperor Hsuan-tsung was able to regain much control and even expand diplomatic and military campaigns westward over the mountains. The threat of the Arab (Muslim) invasion actually caused local rulers Kashmir to the Caspian Sea to seek Chinese protection (719-751). A key Chinese strategic policy was to block the Tibetans from joining the Arabs to their west. As noted in the chapter on Gilgit, the Tibetan route west lay through Gilgit and Yasin to the upper oxus River. In 722 the commander at Kashgar led 4,000 troops to rescue the king of Little P'o-lu from the Tibetans. Kao Hsien-chih in 747 again pushed the Tibetans out of Little P'o-lu. Kao Hsien-chih again was successful in 750 after crossing the Hindu Kush. But in 751 he interfered in Tashkend and executed the ruler. This brought about an uprising in which the Arabs were called in. Kao Hsien-chih suffered a huge defeat by the Arabs in 751 near Talis (Auliata) from which the Chinese never recovered.
The Chinese were also suffering defeat in the south, in Yun-nan and having internal problems. Emperors Hsuan-tsung and Su-tsung could not spare troops for western campaigns. Western troops even helped regain the capital at Ch'ang'an. But the Tibetans lost no time to take advantage. In 758-759 they took control of Kan-su and western Shan-si. This effectively ended Chinese communication with the Tarim at An-hsi. Even so, as the records found by Stein show, the local Chinese governors in some of the Tarim towns managed to retain local control for some time. The Tibetans maintained control over all or most of the Tarim until the years 860-873 when they were replaced by the Uigurs. They in turn were replaced by Turks who in following centuries converted to Islam.

Section IV - Notices of Kashgar during the T'ang Period

Having described the general history of the Tarim, Stein wants to focus in more detail on T'ang accounts of Kashgar. There follows a series of specific incidents with their dates that record a variety of events: some are from T'ang archives and some from memoirs of passing Buddhist monks. Stein comments that no only Buddhism profited in its spread into China but Zoroastrian and Nestorian Christianity did also..

Chapter IV - The Ancient Remains of Kashgar and the Oases of Yarkand and Karghalik

Section I - The Stupas of Kurghan-Tim and Kizil-Debe

Stein took some time during his stay at Kashgar to visit any 'ancient' remains in the vicinity that came to his attention. Two of these were the stupas at Kurghan-tim and Kizil-debe. Both these are in such ruined condition as to make detailed measurements of their original size inconclusive. Stein several times mentions the city wall of Kashgar, but does not describe or photograph it. Unfortunately now the Chinese are leveling even what remains of the old city.

Section II - The Ruins near Khan-ui

Here Stein discusses his visit to another ancient site, Khan-ui, about 20 miles from Kashgar. He visited the place on 4 September. Nearby was the shrine, Bibi Khanum the tomb of the daughter of Satok Beghra Khan. He found another decayed stupa nearby. There were yet more expansive areas in which remains of pottery and ancient debris lay on the surface, but nothing of major interest. The ruin of a place called Mauri-Tim stood out with another stupa the remains of which were fully 38 feet high and on a plateau 30-40 feet above the surrounding area.
Stein drew a plan and section of the stupa. The entire stupa originally was coated with a hard white plaster. All in all the stupa was in remarkable condition. But there was also clear evidence that the stupa had been dug into in search of 'treasure'. The interior contained a shaft which Stein presumed was designed to hold a relic. Here is looks ahead and notes that this design was similar to that of the stupas he found later at Niya, Endere and Rawak and also in India. Near the stupa was another ruin which Stein recognized had originally contained cells in which images of the Buddha and Bodhisattvas were displayed. There were also other ruins and scattered artifacts on the ground.

Section III - The Oasis and City of Yarkand

Stein finally departed Kashgar for Yarkand on 11 September. As always he stayed off the normal commercial caravan route in order to explore in the desert. This route led him to the oasis at Kizil. Along the way he was told about other ancient sites that were too far away to visit. But he did make it to Ordam-Padshah, a desolate shrine in the desert. This place commemorated (at least the legend) of the death of Sultan Arslan Boghra Khan who was attacked there by surprise.
He continues with description of Yarkand. He writes that during the Mongol era the town was named Yeh-li-ch'ien, or Yar-erh - k'an. Marco Polo also mentioned the place as 'Province of Yarcan', then inhabited by Nestorian and Jacobite Christians. Stein continues with discussion of various mentions of Yarkand in ancient or medieval Chinese documents. But he did not stay long. He does note that the town is very favorably situated at the junction of trade routes south and west into India and Iran with those north to Kashgar and on to the northern edge of the desert.

Section IV - Karghalik in Chinese Records

Stein reverts to his guide, Hsuan-tsang, whose journals he follows closely. The monk called the town Che-chu-chia. He also references T'ang Annals and journals of Fa-hsien and Sung Yun. The monk evidently visited Karghalik between Kashgar and Khotan. This is the largest oasis watered by the flow of streams off the mountains, especially the Tiznaf river that flow into the desert or to the Yarkand River. Stein discusses the languages - spoken and written - in use in Karghalik in medieval times. Stein believes the population then was related to that of Sarikol.

Chapter V - The Route from Karghalik to Khotan: its Ancient Topography and Remains

Section I - By the Desert Edge to Khotan

Stein then used the direct road from Karghalik to Khotan from 2 to 12 October. This route passes to the southeast over barren terrain on the edge of the desert. It is made useful by a series of small oases where it crosses small streams and springs fed from the high southern mountains. Stein notes that these oases, sustained by irrigation must be over 1000 years old, although archeological remains are few. This is in contrast with the oases he found now deep in desert east of Khotan. He ascribes this difference to the different relationship of the area to its southern mountains. Here there are foothills but east of Khotan the high mountains are much closer to the desert and this caused a difference in water supply, east of Khotan the large rivers provide more than do the small streams west of Khotan. Stein refers to the usual memoirs of passing monks for historical details.

Section II - The Oasis of Guma

One day east from Karghalik Stein came to a rest house at Kosh-Langar near which he found a ruin, a mound abut 75 feet in circumference and 30 feet high. Continuing east he reached the largest oasis on the route, Guma, on 4 October, where he remained on the 5^th to explore. This oasis relies on the water of the Kilan River. Here Stein began his direct investigation to discredit the forged documents 'found' and sold by one clever Islam Akhun (to be discussed in detail in the last chapter). At Guma Stein quizzed the local head men and inspected the sites claimed by Islam Akhun. Stein was unable even to purchase 'antiquities' at Guma. Stein consideres that Guma was part of the medieval kingdom of Pe-shan, mentioned in the Han Annals as half way between Khotan and Yarkand and directly on the south to India.

Section III - The Tati of Kakshal

On 6 October Stein left Guma toward Moji and Khotan. Much of the barren ground was covered with pottery debris. Presently he came upon another ruined stupa - Topa-Tim, some 47 feet square at base and 29.5 feet high, which he dated to the 10^th century. He found more acres of potsherds on the hard ground around Kakshal Tati.

Section IV - Ancient Remains at Moji

Kakshal Tati was the easternmost area of the Kilan River region. On 7 October Stein halted at Moji Bazar. He was rewarded the following day by being presented with bags of copper coins collected by local headmen. He found more at the nearby original site, Togujai. He identified them as Muhammadan plus some 23 Chinese coins of Emperor Su-tsung (758-760). In addition he found worthwhile examples of pottery, glass and metal. But he found no examples of structural remains. From Togujai Stein moved to a Muhammadan burial-ground called Hasa.

Section V - From Moji to the Khotan Oasis

Stein moved next to the oasis at Zanguya. He again found nondescript potsherds. Moving on his next oasis was Pialma. He found another ruined stupa that measured 65 feet square at base and 22 feet high. Finally, on 10 October Stein reached the eastern edge of the Khotan oasis at another shrine, Kaptar-Mazar. This is the famous place of the 'sacred pigeons' and also the 'sacred rats'. Stein recounts the several legends commonly told here. In General they relate to the first conquest of the area by Muhammadans from the Buddhist residents.

Chapter VI - The Khotan Oasis: Its Geography and People

Section I - The Oasis in its Geographical Features

Khotan has from ancient times been the largest and most important territory on the south side of the Tarim. It is frequently mentioned in Chinese Annals. As usual Stein begins with a lesson about the geographical features and this history but he has reserved much of this information for inclusion in Ruins of Khotan. Stein notes that M. Grenard and Sven Hedin spent long periods in Khotan and have already described it in detail.
He writes that Khotan owes its prominence to its geographic position. The cultivated area here is 40 miles long and has ample irrigation from the Yurung-kash and Kara-kash rivers which flow from the high Kun-lun mountains. And these two rivers between them have an immense watershed. (Stein was to explore and map this area during his second expedition). During this first effort Stein was only able to approach the headwaters by going south from Khotan until he was blocked by impassable gorges. The mountain range reaches 23,000 feet elevation with permanent glaciers. The huge volume of water released by summer sunlight enables the two rivers when they join north of Khotan to form the only river that is able to flow from the south clear across the Taklamakan desert to the Tarim River on the north side (some 300 miles). Before joining to the north, that is from well south of Khotan to its environs the two rivers form many channels to which are added canals thus enabling the cultivation of a wide area. Nevertheless much of the flood water that flows from June to August is lost. But the cultivated area extends from 8 to 20 miles in width west to east. At Stein's visit the chief restriction on irrigation and cultivation was not water but lack of manpower. The cultivated area is also limited by the nearby and encroaching high sand dunes.
The mountains south of Khotan are exceptionally barren. And they are rugged with sharp ridges and deep gorges. They do not offer much grazing ground for sheep or yaks.

Section II - Agriculture and Industries in Khotan

Stein turns to examination of the agricultural and industrial economy of Khotan oasis. The fertile soil enables large crops of wheat, millet, rice, oats and Indian corn (which can produce a second crop). There is also much harvest of lucerne, cotton, mulberry trees and fruit trees (apricot, peach, olive, apple, plus almonds, walnuts, melons and figs),. Khotan also has grapes (rasins). Irrigation water is distributed to the villages according to ancient established detailed custom. The political administration of the entire area is organized according to the irrigation scheme. Stein describes this organization in detail and provides a table depicting the population of each canton.
Khotan, Stein notes, is also the major industrial center of Eastern Turkestan. First of all is mining of jade (white and green). Trade in jade was very important due to its precious value to the Chinese and to its relative ease of transport over long distance. Next in importance was silk of which Khotan was the main supplier. Khotan's cotton crop enables the manufacture of cotton goods. Next comes wool from the large flocks which is raw material for carpets and felt. All these are family, home industries. Then comes paper from the bark of the mulberry trees which grow almost exclusively around Khotan. Stein notes that he later found paper manuscripts in several locations dating from mid 8^th century. Next, Stein mentions ceramic art that was much better long ago. By 1900 glass making had disappeared. Metal work proceeds in brass and copper and also gold and silver.

Section III - The Population of Khotan: Its Distribution and Character

Stein next examines the population mainly as to culture. Stein includes the population estimates of various European visitors and gives his own at about 220,000. Khotan is divided into three population centers none of which equal Kashgar or Yarkand, but together with the populations of the settled areas between and around them they become significant. Khotan's location with mountains to south and desert to north and east and west reduced the migration there of significant ethnic groups. Thus most large scale migration recorded in historical times moved along the T'ien shan and northern areas beyond the Tarim basin. Only the conquest from the west by Satok Boghra Khan could have brought significant ethnic change. Up until then travelers such as Hsuan-tsang noted the population was refined Buddhist in culture. The Buddhists of Khotan resisted the conversion to Islam more strenuously than inhabitants of other parts of the Tarim, but once converted the basic culture has remained much the same - that is easy-going, indolent, somewhat decadent and good-natured. They are fond of feast and entertainments, music, singing and dancing. Many men are adventurous and seek buried treasure rumored to abound all around. There is very little of crimes of violence.

Section IV - The Population of Khotan: Its Physical Characteristics and Racial Origin

Now comes ethnography and racial study. In his time ethnic questions were of great interest and studies of racial makeup of populations were in vogue. Stein was always eager to conduct anthropometrical studies on the local populations where ever he went. He writes that his data from Khotan and Keriya is limited and there is no similar information from the Chinese Annals. Thus he notes his estimates are based on his own physical measurements of a limited number of subjects. He includes photos of several groups as illustrations. He turned his data over to an expert, Mr T. A. Joyce, whose subsequent analytical paper Stein quotes at length. One conclusion is that the people do not display Mongolian characteristics but rather those of the Galchas, that is Alpine Turks, a group largely resident in the high valleys between the Hindukush and the Altai and who speak an eastern Iranian dialect. These include the Wakhis and Sarikolis who Stein already encountered. Some differences, Joyce, attributes to slight admixture of Turki and Tibetan blood. Stein refers again to the T'ang Archives in which it was noted that the peoples of Sarikol and Khotan have t he same appearance. Furthermore, Stein notes, the documents he subsequently uncovered at Dandan-Uiliq were written in Brahmi characters of the 8^th century in an Indo-Iranian dialect similar to the Galach dialect of the Pamir region. Further, the admixture of Turki blood must have occurred after the Mohammedan conquest and conversion of the Buddhists. Next, Stein expounds at length on the peoples of Tibet and potential for some influence from that quarter on the population of Khotan.

Chapter VII - Historical Notices of Khotan

Section I - The Early Records and Names of Khotan

In this section Stein focuses on the historical record and the names ascribed to Khotan. For this effort he relies on Chinese official annals and the reports of Chinese travelers plus some references in Tibetan documents. He notes that the relations between China and Khotan extended over 1000 years prior to the Moslem conquest. He acknowledges the work of Mr A. Remusat who published back in 1820 a history of Khotan culled from translation of all Chinese notes on the topic to be found in the Pien I tien. Stein remarks on a difficulty that arises from the fact that rulers of Khotan appear in Chinese records with Chinese names while in Tibetan records with Indian names. He writes that the very name, Khotan, is connected with ancient legends. His Buddhist monk - mentor (patron saint) hsuan-tsang naturally mentions Khotan under the name Ch'u-sa-tan-na or Che-chu-chia which is a rendition ofthe Sanskrit Kustana - and both mean 'breast of the earth'. The official Chinese records, however, use the name, Yu-t'ien or Yu-tun. Stein continues for several pages about the name of Khotan.

Section II - The Legendary Traditions of Khotan

Then he turns to legendary traditions about Khotan. He writes that Khotan first entered official Chinese records during the reign of Emperor Wu ti (140 - 87 BC). The legendary origins of Khotan are much older and he repeats several. One ascribes the foundation to the Indian god Vaisravana or Kubera, the Hindu god of wealth and of demons, who is also in Buddhist mythology. There he is worshiped as one of the Lokapals, ruler of the North. The legendary foundation of Khotan specifies colonies from both northwest India and China. This discussion continues for many pages.

Section III - Khotan in Chinese Records, from the Han to the Sui Dynasty

Stein turns again to Chinese records, mentioning again those of the Emperor Wu-ti. These begin with mention of the first embassy from Yu-t'ien to Wu-ti's court from which the kings of Khotan received tokens of investiture from the Chinese emperor on during following centuries. The population then was given as 3,300 families or 19,300 people including 2,400 soldiers. Later, during the reign of Emperor Kuang-wu ti (25-57 AD) the King u-lin of Khotan became a subject of the powerful king of So-ch'e (that is of Yarkand). Stein continues with extracts from the Han Annals in which the relative power of Khotan and such neighbors as Yarkand and Kashgar waxes and wains. It was in 73 AD that Chinese general Pan Ch'ao launched his major offensive into the Tarim. In following years Chinese power also waxed and wained throughout the Tarim Basin.

Section IV - Khotan during the T'ang Period

Stein moves on to the T'ang Dynasty records. The second T'ang emperor, T'ai tsung, (627-650) reasserted Chinese power in Eastern Turkistan. Khotan then was one of the 'four garrisons' from 648 or 649 (including Kucha, Kashgar and Tokmak). The records continue until around 790 AD.

Section V - Later Chinese Records of Khotan

Khotan continued to be noticed in later Chinese documents after a break of 150 years or so. During this period first the Tibetans and then the Uigurs took control of parts of the Tarim Basin. However, the Chinese records mention an embassy from Khotan in 938 seeking assistance against the Tibetans. The Chinese in response sent a mission to Khotan. Another embassy from Khotan arrived in 942 and others in 966,n 969, 971. Khotan was then conquered by the Moslem Turks by 1006. Stein continues with the scattered mentions of Khotan. For instance Marco Polo passed through between 1271-1275 at which time Khotan was on a flourishing caravan route.

Chapter VIII - Ancient Sites of the Khotan Oasis

Section I - The hill of Gosringa

Stein resumes his archeological efforts to identify the ancient sites around Khotan from physical evidence. But again, he relies on Hsuan-tsang for clues to locations. He started with a visit to Kohmari hill above the Kara-kash river SW of Khotan. By 11 November Stein reached Ukat at the entrance to the Kara-kash gorge. The following day he went to Kohmari hill. He found the location exactly as Hsuan-tsang had described it and provides photos for illustrations in the text. There is a Ziarat or Mazar (shrine) still there, but now Islamic rather than Buddhist of course. Stein paid a fee to explore inside the sacred cave.

Section II - The Culture-strata of Yotkan

Stein refers to accounts of other recent European travelers such as M. Grenard, who first learned that the village, Yoktan, was the source now of numerous antiques entering the market. He recognized that this must lie over the remains of ancient Khotan. Stein never content with second hand acquisition of ancient artifacts promptly went to Yoktan on 15 October and began purchase of whatever was now available. He returned to Yoktan on 25 November to conduct his own exploration. He made a detailed survey including a map. He discovered that a deep ravine in the village showed multiple layers of archeological interest - clearly remains from former occupation. The locals indicated that the ravine itself began in the 1860's with construction of an irrigation canal that turned into a channel which uncovered a part of the underlying remains. But when some residents found bits of gold the result then was annual digging efforts precisely with the object of finding salable artifacts, especially of course gold and this greatly expanded the excavation area. The work was an annual affair during the period, July to September, in which sufficient flood water was available to wash the dirt. The proceeds if any were shared by the land owners and the digger teams. Since as is so often the case legal ownership of all gold found in the ground is the Chinese government, naturally what is found is usually kept secret. Stein notes that the original capital at the Yoktan site was occupied by a rich city for centuries, so it is understandable that much valuable material should be found eventually. The entire area of the ancient city now is buried under many feet of soil that has accumulated during the centuries of irrigated agriculture.

Section III - The Site of the Ancient Capital

Stein turns to the archeological remains and what they may reveal. His map provides potential details about the extent of the town. Stein again turns to descriptions of the ancient capital in Chinese official annals and the memoir of his favorite monk. The location of Yoktan corresponds to its description in the Annals. Important for dating the former capital were the many coins that Stein purchased. They included Chinese coins with Indian inscriptions dated from after 73 AD. But there were earlier coins as well including one from near 1 AD. Stein describes many more coins with various dates. T'ang era coins dating as late as 779 were found as well. And, after a break in dates, coins from the 11^th century were numerous. But the coins provide no evidence related to the cause for the abandonment of the city.

Section IV - Antiques acquired from Yotkan and in Khotan

This section describes other antiques Stein found or bought in either Yoktan or Khotan and which he displays in extensive illustrations. Typically these are mostly ceramic pottery and terra cotta pieces. He also provides extensive text descriptions. Many are Buddhist and show the Graeco-Buddhist artistic style of Gandhara. But there also are Roman items. There follow 12 pages of detailed descriptions of each item.

Section V - Buddhist Sites described by Hsuan-tsang

Stein devotes this section to descriptions made by Hsuan-tsang of individual Buddhist places around Khotan. The first of these was a convent with stupa west of the capital. Stein searched for it on 28 November. At a place called Somiya he found, as usual, a Mohammedan Mazar and cemetery with shrine located at the site taken over from a Buddhist shrine of forgotten age but avoided by the locals for superstitious reasons. He continued on to locate another site mentioned by Hsuan-tsang - a Buddhist convent called Sangharama southwest of the capital and now occupied by a Ziarat, cemetery and tomb of a Mohammedan worthy. Stein returned to Khotan on 29 November and visited more similar locations. Stein recounts the legends and stories related to these.

Chapter IX - The Ruins of Dandan-Uiliq

Section I - The Desert March to the Site

Stein engaged the local Ak-sakal of Afghan traders ,one Badruddin Khan, to send parties of 'treasure seekers' into the desert to locate potential ancient sites worth excavating. When Stein returned to Khotan in November the results were offered. The most promising result was offered by Turdi Khwaja, a resident of Tawakkel and a professional explorer with long experience in the desert, who already had visited such places. Turdi turned out to be one of Stein's best assistants in these explorations. He advised a visi to Dandan-uliq (houses of ivory). It turned out that this location had been discovered by Dr Hedin by accident in 1896 and his two guides were found. Moreover, the samples produced by Turdi convinced the local Chinese Amban - Pan ta'jen - of the potential of such explorations in the desert so that adequate labor was made available. Stein spent a week after 29 November writing reports of results to date. He also had to reorganize his baggage in order to take the minimum essential into the desert on camels. He left Khotan on 7 December for Dandan-uliq via Tawakkel. They crossed the Yurung-kash to Tawakkel, a recent colony in reclaimed desert. There he hired laborers for excavation teams. Stein spent 11 December at At-bashi organizing the 30 laborers and collecting a month's food supply. Eight camels were insufficient for the load so a dozen donkeys were hired as well and the ponies were sent back to Khotan. They started into the desert on 12 December. Meanwhile Stein had sent Ram Singh on another surveying route through the mountains east of Khotan before turning north to rejoin the party. The march to Dandan- uiliq took 5 days through empty sand dunes. Water was found on this route by digging wells to a depth of 7 feet. The daytime temperature remained below freezing and at night reached -10 degrees. On 18 December they reached the remains. The area containing the ruined buildings was 1.5 miles from north to south and about a mile east to west. immediately two Chinese copper coins of 713-741 were found. Once unloaded at the site the camels were sent on to the Keriya Darya for rest, water and fodder. The donkeys were sent back to Tawakkel.

Section II - First Excavation of Buddhist Shrines

Stein got to work promptly on December 19th with the assistance of Turdi Khwaja who could indicate the best specific places to excavate starting with a Buddhist shrine. This consisted of an inner room surrounded by an outer corridor. Both inner and outer walls consisted of a wood frame with reed matting covered with rough clay plaster. These were covered with plaster of Paris fresco decorations of Buddhas and Bodhisattvas. Remarkably the color of many decorations was preserved. The uniformity of many stucco items indicated that they were made in moulds. Other decorations were produced by using stencils. Stein includes these in the illustrations. After describing both the excavation methods and results of work at D1 he continues to DII a much better preserved shrine. Because that building was more deeply buried in sand the decorations on walls were better preserved. Unfortunately a huge statue of the Buddha was not so protected which resulted in only the feet remaining.

Section III - Art Relics of Shrine D ii

Excavations on 20 and 21 December revealed another shrine containing many artistic items which Stein describes as usual. Again, many were small relief Buddhas or Bodhisattvas. Stein notes the similarity of the artistic style here to that of Gandhara. One relief, he notes, depicts two soldiers wearing coats of mail the look like Greeks with lower armor plates in horizontal rows and upper armor resembling a lorica with scales in vertical order exactly like armor worn by figures in the museum at Lahore. Stein considers that there is a representation of Kubera - Vaisravana - the most prominent of the four Yaksa kings or Lokapalas who was worshiped at Khotan. Vaisravana is depicted on one of the pillars. Stein continues with detailed interpretations of many items based on Buddhist mythology.

Section IV - First Finds of Ancient Manuscripts

In addition to Buddhist art Stein found many ancient documents, among them some of the most significant finds. On December 22 he proceeded to another ruined building DIII expecting to find a former dwelling. This structure measured 23 feet east to west and near 20 feet north to south. In this one Stein found Buddhist texts in Proto-Tibetan language. Other documents were in Brahmi alphabet and Eastern Iranian language. Sanscrit documents of the 7^th to 8^th century were also uncovered. Stein uncovered a kitchen complete with many utensils and supports. He also found more shrines among the ruined buildings. He found also painted wooden blocks. Among the representations Stein identifies Vajrapani, Manjusri, Maitreya and Avalokiesvara.

Section V - Discovery of Dated Documents

Stein next shifted work to ruins east north east of the camp, Div and Dv. They yielded some wood paintings and tablets and Brahmi documents. A Chinese document was also found. It was a private petition to recover a donkey and dated from 781 AD. It remarkably corresponded with other documents Turdi had recovered and sold some years past. Stein continues with discussion of many more documents. From these Stein dates the closing of the site to shortly after 781- 790. And he believes that the Brahmi script documents in Iranian language then date the same as the Chinese documents, indicating that the Iranian language was still spoken there at the end of the 8^th century.

Section VI - Records from the Hu-kuo Convent

Stein recovered a significant harvest of relics from another groups of buildings designated Dvi and Dviii. The lower part of the walls still had their fresco decorations. He found manuscripts as well. These related to a Buddhist monastery.

Section VII - Other Ruins of Dandan-Uiliq and General Observations on Site

This section deals with various other ruined buildings that Sten excavated. All are described in detail. He notes in this section that Ram Singh arrived on 24 December from his extensive survey route from the east - the Keriya River route. He records his general conclusions. He again proposes that Dandan-uiliq was abandoned near the end of the 8^th century, shortly after 790. The documents also indicate that there was a Chinese garrison with Chinese control at that time. Stein connects this abandonment with the major political changes that took place in T'and China at that time. This was the period of the Tibetan invasion and collapse of T'ang power in the Tarim. The loss of population and effective political control would result in the desert gaining an upper hand due to failure to maintain irrigation canals and reduced water supply. Stein notes that Dandan is on the highest ground, the watershed between the Keriay Darya and the Yurung-kash and that this would be the logical location for a line of canal coming from the south. Thus the permanent abandonment of the location centuries later is likely due to the change over that time period in the climate and reduction of the amount of water available at the oases due south.

Section VIII - List of Objects excavated or found at Dandan-Uiliq

An extensive and detailed 16-page description of the artifacts recovered at Dandan-uiliq

Chapter X - From Dandan-Uiliq to the Niya River

Section I - The Rawak Site

Stein finished work at Dandan on 3 January and set out north to Rawak on the 4^th. He paid the Tawakkel workers and sent them back home. Turdi Khwaja showed Stein the ruin which consisted of two badly ruined stupas. They found some broken glass and stucco laying about. However the very tall sand dunes covered whatever else may have remained of buildings. They did find a pair of long tablets containing Kharosthi script in an Eastern Iranian dialect. In addition they found six Chinese coins, 4 were Han Dynasty and one T'ang.

Section II - A Judaeo-Persian Document

Stein was surprised when several of the Tawakkel workers returned. Rather than return home immediately they had continued to search at Dandan and had found a remarkable document that they thought Stein would want. In addition to a small stucco on which were written some Chinese characters and the other was a lump of paper that was sensational. After the paper was opened back in London it was determined to be a fragment of a Judaeo-Persian document. It turned out that this is the most ancient such document,. It is a letter written by a Persian speaking Jew relating to business affairs and was likely written around 718 AD.

Section III - Keriya, Niya, and Iman Ja'far Sadiq

From Dandan Stein next moved on to the Niya River and was shown another extensive ruined village abandoned in the desert. He set out from Rawak on 6 January for the Keriya River. By evening of January 8^th they reached the frozen Keriya. After for long days they reached Keriya town. At Keriya on 12 January Stein met with the Aqmban, Huang Daloi, who eagerly assisted with obtaining more workers and supplies. He remained there for 5 days. On 18 January he set out by way of the Mazar of Iman Ja'far Sadiq for a rumored site near Niya. He reached Niya on the 22^nd. Niya corresponds to the Ni-jang described by Hsuan-tsang. There he hired another group of laborers and obtained supplies. Stein was excited when a local man brought him several ancient tablets with Kharosthi writing which were claimed to have been found at a deseerted site to the north.
He left Niya on 23 January headed north along the river for Iman Jas'far Sadiq with 20 workers and additional camels. They reached the shrine after 3 days. From there on the course would be across desert sand. The shrine is an active site visited by pilgrims from throughout Turkestan. It has mosques and Madrashas and houses for the shaikhs. An imposing tomb houses the saint who it is believed died here in battle with the remaining infidels (Buddhists). Stein measured the temperature at 44 below zero. At last he found the remains of buildings mostly buried in the sand.

Chapter XI - The Ancient Site Beyond the Niya River

Section I - The Ruin N i and the first Finds of Inscribed Tablets

Stein immediately headed on 28 January for the building from which his guide, Ibrahim, claimed to have taken the two wooden tablets. He was anxious as always that his finds not be disturbed before he could render scientific archeological descriptions of their exact settings. As a precaution he even had this individual carefully watched less he go ahead and grab more tablets to offer. The building Stein marked as Ni and it is illustrated in the report. Stein was delighted to find the mass of documents just as Ibrahim had claimed. The latter was no doubt chagrined that he had not realized their great value and extracted more of them when he had the chance. The building was constructed of timber with massive squared beams as foundations upon which were square wooden posts that framed the walls and supported the now gone roof. The walls were composed of matting made of thin tamarisk branches woven diagonally. This was covered on each side with white plaster, making the walls 6 to 8 inches thick. Lesser buildings such as sheds were constructed less elaborately. Stein numbered 114 tablets recovered in the main room. The documents were composed of two wood tablets carefully tied together with writing on the inside surfaces and clay seals on the outside. Stein describes the physical characteristics of the tablets in detail. Stein determined that the form of the Kharosthi writing of an early Prakrit language corresponded to that of North-West India during the Kusana or Indo-Scythian era - the first three centuries AD. On January 29 Stein began clearing again. The building is shown on Plate XXVIII. Immediately more and more documents (over 145) were recovered. These included not only more of the double tablets with writing on inside surfaces but also many various single wooden pieces of varied length and width with a wide variety of text there on. Stein decided these were records and accounts and there were also official letters. Stein was able to translate much of the Sanscrit writing. These appeared to be the random remains from some official's residence or office. Stein also remarks on the total absence of paper. He also noted the severe erosion caused by wind and wind-blown sand.

Section II - Excavation of Ancient Residences, Nii, Niii, Niv

On 30 January Stein walked to several areas in which his men had located other ruins which he described in detail. Some buildings were empty, but in others he found more wooden tablets, but most of them were too badly eroded for recovery. In one building he discovered an ancient ice house. The method for preserving ice there for summer use was recognized by his Keriya locals as identical to that still in use. But he soon abandoned further excavation as the chance for finding documents was clearly too small.
By 31 January the minimum temperature had risen to minus 3 but daytime maximum was already up to 43 degrees, indicating the rapid approach of warm spring and hot summer. The critical immediate issue was not the heat for working conditions or even the pending sand storms, but the continued preservation of the vital ice supply on which they depended for water. He humidly concentrated on several imposing ruined structures. In one of these his helpers recognized the organization of a kitchen similar to their own. In another room he found a bow and part of a shield and staff and other implements. The walls of this ruin remained to a height of 6 to 8 feet thus preserving interesting fresco decoration. The part of an ancient rug and the ornamental carving on a chair showed Stein the influence of Indian art seen in Gandara. More documents were uncovered in adjacent ruins. Included there were Chinese texts on wood as well as more in Kharosthi. He also found more chairs and woolen carpet. Outside these buildings Stein found the remains of gardens and orchards including peach, apricot, mulberry, and oleaster trees. Stein finished clearing building Viv on 5 February..

Section III - Discoveries in an Ancient Rubbish - heap, Nxv

On February 6^th Stein shifted the camp site north to be closer to more ruins. One of these was a refuse dump from which Stein recovered a trove. He also excavated a stupa that measured 20- feet 6 inches tall and with lowest base 19 feet 6 inches square. It was built of sun-dried bricks. In the center was a shaft 1 foot square. The refuse heap yielded over 250 documents on wood and leather as well as broken pottery, matting, wood, straw, rags and woven material. It was so extensive that 3 days were required for its excavation.

Section IV - Ancient Documents on Wood and Leather

Stein turns from his description of the physical characteristics of the buildings and rubbish heap to discussion of the wood and leather (sheepskin) documents found there in. Stein notes that these leather documents are among the oldest ever found in an Indian language. He describes the methods for creating completed documents from two pieces of wood tied together with string and sealed cover the tie of the string with a clay stamped seal and provides with the text clear drawings showing how this was accomplished. The desired result was to create a secure message written on the inner side of the wood tablets thus sealed together. On the outside surface of the covering tablet the name and title of the addressee was written. On the reverse of the under tablet the name of the messenger or person referred to in the text was written. Unfortunately at the time of the writing of this report Stein did not posses the professional translations being accomplished at the British Museum. He is left to guess at the content and purpose of these documents, considering that they may be contracts or agreements. He turns to description of some of the clay seals found either still in place or separately. Many are too eroded to be decoded. Bu some show Greek or Roman designs, such as an Eros of Pallas Athene. Some are bust portraits. Still others are Oriental and likely made locally. Stein remarks on the evidence of such wide spread influence of Western art into Central Asia.

Section V - Chinese Documents from Nxv and the Writing on Wood

In this section Stein focuses on the physical description of the documents. The examples of documents written in Chinese were fewer in number than those in Kharosthi. But Stein describes them in detail. He again references M. Chavannes for descriptions of Chinese methods of writing and document construction found in Chinese literature. Stein found physical documents of which no examples existed in China. This section will be of interest to readers who study ancient Chinese methods of writing.

Section VI - Decipherment of Ancient Documents, Kharosthi and Chinese

In this section Stein turns to translation (decipherment) of the documents, both those in Kharosthi and Chinese. While he is fluent in Sanskrit, he regularly deplores his lack of time to learn Chinese. Thus translation on the spot depends on his Chinese secretary and later on experts back in Europe. He mentions the three fold importance of the Kharosthi documents - the nature of the script, the relation of the language to others, and the enlightenment on Asian history from the texts. He notes that the full translation and commentary on the texts has not been accomplished by the experts as of the date of his report. However, Professor Rapson has already confirmed that the main language for the Kharosthi documents is an early form of Prakrit, with mix of Sanskrit terms. Most are official documents and records such as reports and orders to local officials dealing with police, complaints, summonses, supply and transport and the like. The wedge-shaped double tablets appear to concern the bearer. Some of the documents relate to private matters. Some are only drafts. Some of the proper names written in the documents are Indian but others are not. Various non-Indian titles are included. The documents are from the first centuries BC and AD and generally correspond to the Kusana era and the region around Taxila. Stein wonders how such documents containing language and script not found from such an early date in India itself (apart from inscriptions) should have been in such use in Turkestan. The Kusana era in north west India was the first three centuries AD. The coins found at Niya were all from the Later Han era.
Stein turns to the Chinese language documents. These also were mostly official texts in script associated with the Later Han dynasty. One of these is explicitly dated to the 5^th year of the reign of T'ai-shis of the Emperor Wu ti (265-289 AD), founder of the Western Chin and relating to the year 269 AD. The documents confirm that the Chinese established administrative centers over the local rulers throughout the Tarim basin at that time. Documents also refer to Tun-huang as a center of administrative control. Some of the documents appear to be identifications of individuals providing name, age and appearance - evidently as passes for travel. They also list the material being transported. The manner in which the Chinese and Kharosthi documents were mixed indicates that the local officials and Chinese worked simultaneously.
The other artifacts, lacquer wear and pottery are also Chinese. But the glass wear was imported from the West, The Yueh-chin - or Indo-Scythians.

Section VII - Exploration of Ruins Nvi - xii and General Observations on Site

In this section Stein describes his excavation of remaining buildings at Niya, which continjued into February, and his general conclusions about the site. One building was a shrine. Relatively few objects were recovered from these buildings. The Niya site measured 7 miles north to south and 2 miles west to east. Stein indicates that the substantial wooden buildings he found scattered throughout the area most probably belonged to the Begs and other head-men who lived on their own property, but that the common folk likely lived in mud huts in between and that these would have decayed completely. He estimated the extent of the original cultivated area and noted that the site must have been abandoned when the Niya River ceased to flow far enough north.

Section VIII - List of Antiques excavated at Niya Site

Again, Stein provides detailed descriptions of each item recovered at Niya with reference numbers indicating where it was found. This fills about 30 pages in folio size. A typical entry is N.i. 24. Wedge cov. tablet (belongs to N.i 180) Obv. 2.5 inches from sq. end, seal cav. Remains of string; cav 2 5/8th by 1 1/4 inches. No writing visible. Rev blank 13 7/8th by 2.5 by 3/4 inches Much perished by damp.

Chapter XII - The Endere Ruins

Section I - The March to the Endere River

Stein describes his move from Niya to Endere River. He notes that the temperature while at Niya had a minimum of 6 - 9 degrees below zero. On 13 February Sten started back to the Mazar Iman Ja'far Sadiq. He was reluctant to leave, but realized the winter was ending and little time remained before the spring would bring sand storms. There he paid the exhausted workers. He hired a fresh team from Niya village. They set out eastward on 15 February. On 17 Febrary they followed the course of the Yartunguz River. They stopped at a new oasis village, Kala-sulaghi. On 18 February they continued east over sand dunes reaching to 180 feet high.

Section II - Excavation of the Endere Temple

On February 21 Stein sighted the stupa at Endere ruin. The new team from Niya arrived at the same time, having crossed 120 miles of desert. Immediately Stein found something new - an imposing fortress wall encircling the main ruin shown in plan on Plate XXXVI. It was a circular clay rampart nearly 320 feet in diameter. Stein immediately put his entire team to work excavating a building within the fortress walls. It turned out to be a well decorated temple. It was ab out 18 feet 4 inches on each side, with walls made of timber and plaster. The main timbers were very large. Remains of statues occupied each corner of the central room. Although most of the plaster figures and decorations were broken and extremely fragile, Stein managed to take a few samples including a piece of otherwise destroyed wall fresco to London. He describes the building and its contents in detail and includes photos. His hopes for documents were fulfilled when he found a nearly complete 46 page folio sized Buddhist canonical work in Sanskrit from the 7th or 8^th centuries. Other documents were in Brahmi or Gupta in possibly proto-Tibetan language. There were other fragments in Tibetan language. Dr Barnett confirmed that these were part of a Pothi containing about half of the canonical text of the Salistamba-sutra. Analysis of the paper indicates that the document was imported from Tibet. Stein writes that these are the earliest specimens of Tibetan language yet known. On one wall he found a written text with a specific date of 791 AD. It was in 791 that T'ang control finally fell to a Tibetan invasion. On other walls there were grafiti in Tibetan script. Stein also found pieces of textiles and drawings.

Section III - The Ruined Fort and Stupa of the Endere Site

After thoroughly excavating the temple on 23 February Stein focused on several smaller rooms. In some of these the walls retained the lower portion of excellent frescos and on one wall appeared a complete panel depicting Ganesa or Vinayaka, the four armed deity with elephant head from India. Stein found another, very large, building with massive walls constructed of large bricks. The remains of large roof beams were lying in the sand. He identified this as the residence of the commander.
Stein then examined the fortification. The wall was a solid rampart of stamped clay probably built up inside rectangular forms. Sections of the wall were destroyed by erosion but other sections retained some height up to 17.5 feet . The width of the rampart was about 30 feet at the base and where the top remained there was a parapet 5.5 feet high and 3 feet thick. Where the original gate was located there was only a wide gap 18 feet wide with small square bastions on either side. Stein then describes the stupa, which is well illustrated with the text. It was rather large, a dome 16 feet in diameter and 14 feet tall despite a broken off top. This is on top of the standard three level square base.
For more illustrations please see Endere.

Section IV - List of Antiques from the Endere Ruins

This is an extensive 5-page listing of the items recovered at Endere.

Chapter XIII - Kara-dong and the Search for Hsuan-tsang's P'i-mo

Section I - Expedition to the Kara-dong Ruins

Stein completed work at Endere on 26 February. With the spring sand storms rapidly approaching Stein started back westward to explore more ruins around Keriya and Khotan. He returned to Niya by 2 March, thus completing a loop over 300 miles long. On 3 and 4 March he rapidly moved to Keriya, 80 miles. Two days there sufficed for Stein to obtain additional labor, camels and supplies thanks to the gracious assistance of the Amban, Huang-Daloi. Stein's next objective was the ruin at Kara-dong, about 150 miles north of Keriya in desolate desert. This required a 6-day trip down the Keriya River. At the Mazar of Burhanuddin-Padshahim Stein hired more local laborers. But finding the ruin turned out to be difficult due to the hazy memory if the local guides. The ruin consisted of a quadrangle within a mud rampart about 30 feet thick around an enclosure 126 feet square. Much of it was covered by deep sand. There was much pottery, metal objects and shreds of fabrics lying about in the open. Stein did recover two bronze arrow-heads five Chinese copper coins and a few other items. He found the 21 foot square gateway on the eastern side was well preserved owing to protective walls and sand. Stein as always meticulously recorded the measurements of each room and building and the meager quantity of artifacts. Kara-dong was a much more primitive site than Endere or Niya and was located in a more desolate area with less water available. Stein guessed it was designed as a way station for travelers along the Keriya river route. In the Middle Ages the Keriya river reached as far north across the desert as the Tarim River. This route is the most direct and shortest between Keriya and Khotan and the north around Kucha.
(We learn later that Stein in fact during his second exploration tested this idea, but not by marching north along the river in relative safety but rather south from Kucha across the empty space between the Tarim and the lost end of the Keriya in the desert).
Stein remarks that the medieval history of the area shows that the governments of Khotan and Kucha had a common border in the desert and close contacts. Stein found no datable items from which to establish exactly the period during which Khara-dong was occupied.

Section II - Hsuan-tsang's Notice of P'i-mo and Marco Polo's Pein

Stein finished his work at Kara-dong on 17 March and departed on the 18^th . Sand storms were already significant. He was set on finding another ancient place - P'i-mo because it was mentioned by Hsuan-tsang and was likely the place Marco Polo named Pein. Therefore he moved South-westward from Kara-dong into the desert. Once again local guides supplied by authorities were too afraid to admit their actual ignorance, so led Stein into the desert from the Keriya River starting on 23 March. Eventually they found some pottery debris in the desert near Arish-Mazar. Stein digresses to recount the legend told to Hsuan-tsang.

Section III - The Sites of Uzun-Tati and Ulugh-Ziarat

Stein continued his search around Mazar Lachin-ata. He questioned local farmers and shepherds at length about any ruins seen in the desert. So they set out again on 25 March with out reliance on the guides. They passed through now- abandoned cultivated areas of 'old Domoko'. Eventually after days of wandering they found ruins called Uzun-tati and Ulugh-Ziarat. Stein indicates that they wandered about and finally found these places with a couple miles of their tracks but hidden by sand dunes. These were fields strewn with pottery and other debris and a few low mud walls. But the area showed clear evidence of previous digging by 'fortune seekers' from local villages. However, a coin from the period 1038-39 AD and some ceramic porcelain from late Chinese production served to show the remains dated much later than Niya or Endere. Another coin was from Muhammad Arslan Khan - 11^th century. Ulug-Ziarat proved to contain the ruin of a small oval fort 480 feet north south and 348 feet east to west. The remaining wall was 11 feet thick a base and 9 feet high composed of stamped dirt with layers of rushes. There was a parapet about 5 feet high and 3 feet thick. Behind it was a platform 5 feet wide.. There were no visible remains of buildings within the enclosure. Even though Stein found precious little at these locations, he was determined to connect them with Husuan-tsang's P'i-mo. He places strong reliance on the measured distances between various towns and also in the name, Ulugh-Ziarat. Ziarat means a holy shrine and the medieval Buddhist travelers wrote that there was a very famous and large statue of the Buddha at this place. Stein always considers significant that Moslems take over Buddhist holy places and then continue to worship there to different saints.
By March 28 the temperature at mid day was already 88 degrees in the shade while the minimum at night was still 27-28 degrees. Moving back south they soon reached the flourishing oasis at Gulakhama.

Section IV - The Tuga-dong Mounds and the Keriya-Khotan Route

Stein allowed for a day's halt at Gulakhama, but was pressed for time to beat the coming summer. During the break Stein visited ruins at nearby Chira. There he found mounds and uncovered human bones. So on March 30 Stein set his main party under conrol of Lal Singh on to Khotan while he went back to Keriya for the courtesy of proper leave-taking of the Amban. This he accomplished on 1 April with Huang-Daloi. Stein made sure to heap high praise on his Darogha, Ibrahim Akhun, which secured the latter recognition and a promotion. Stein started back to Khotan on 2 April observing along the way the extent of the spring flood. He reached the edge of the Khotan oasis at Lop village.

Chapter XIV - The Ruins of Ak-sipil and Rawak

Section I - Hanguya Tati and the Site of Tam-oghil

On April 5 Stein moved from Lop Bazar north, directly to the ancient Tati of Hanguya. After crossing cultivated areas and the desert he came to Arka-kuduk-Tim (the Mound of the Back Well), which was another ruined stupa. He then returned through extensive cultivated areas to camp at Yurung-kash. Stein stayed at Yurung-kash on 6 April, but left on the 7^th for Ak-sipil (the White Wall).

Section II - The Remains of Ak-sipil and Kighillik

Stein camped at Ak-sipil on 8 and 9 April in order to complete a full survey of the area. This was another ruin of a fort. The rampart and parapet of the an ancient fort was among low sand dunes about 8 to 15 feet above the original ground level. But most of it was destroyed by erosion. The better preserved northern part was an arc about 360 feet long. Stein estimated the original fort at 800 feet in diameter. The rampart was built of stamped loess and was about 50 feet thick at base and 11 feet high. The parapet on top was made of sun-dried bricks on a platform 2 feet high that projected 3 feet outside the wall and 2 feet inside. The parapet was 8 feet thick and no where more than 7 feet high. The bricks measured 20x15x4 inches. There were loopholes about 3 inches square on the lower level and 6 inches wide by 8 inches high in the upper level, which were 16 inches and 5 feet above the base. Stein found two areas along the wall that were strengthened by solid brick platforms projecting 3 feet on either side of the base. They likely served as bases for watch towers. One had a stairway near by.
Stein found only one coin at Ak-sipil, but was brought 13 old copper coins while at Khotan by sellers who alleged they came from Ak-sipil. Stein as always was very cautious about crediting such information. Turdi guided Stein to another ruined Buddhist shrine 1.5 miles to the south-west, called Kighillik (the dunghill - for evident reasons) where despite the destruction already created by 'treasure seekers' he uncovered plaster and terra-cotta fragments. Over the course of the three expeditions Stein found much of interest in similar 'refuse' dumps. The mound proved to be the remains also of a Buddhist shrine. Many of the fragments were formerly part of wall decorations, small standing Buddhas.

Section III - The Rawak Stupa

Stein left Ak-sipil on 10 April to move to another site recommended by Turdi Khwaja - Rawak (the high mansion). He was surprised and rewarded to find a very large stupa instead of a simple dwelling. (Fig 60, PlateXIII) All about were remains including heads and upper parts of formerly massive statues. Stein called back to Yurung-kash village for more laborers and dug a well. The Rawak location was only 7 miles from the Yurung-kash River making the water table relatively high. But already the season of intense sand-storms had begun. Excavation of the Buddhist Vihara nevertheless proceeded rapidly despite the dense haze and increasing heat. The Vihara court was a rectangle 163 feet on south-western and north-eastern faces and141 feet on the other two sides. The enclosing wall was 3 feet 6v inches thick and built of sun-dried bricks. Were best preserved by the sand dunes on the south corner it remained 11 feet high. The bricks were 20x145 by 3.5 inches. The longer side walls were covered by so much sand they could not be uncovered with the workers available. There was a gate 8.5 feet wide in the south-east face. The large stupa was in the center of the court yard. Its three story base alone measured 22.5 feet high. The lowest story was 78 feet square and 7.5 feet high. The second story was 45.5 feet square and 9 feet high. Above that was a circular drum 3 feet high and above it another circular drum 32 feet in diameter as base for the ruined dome. Stein found that 'treasure seeking' operations had already created serious cuttings into the stupa. In addition to the standard construction of the base there were projections about 52 feet long on each face supporting flights of stairs from the court yard to the base of the dome. The lowest level of the ramp was 14 feet wide, the next level 9 feet and the third 6 feet wide. Remaining sections of the stupa protected by sand were covered with white plaster. Attached and nearby were many copper coins that had been votive offerings. Stein considered the many relief sculptures and decorations to be the most important finds. The larger than life statues represented the Buddha and Bodhisatvas in various poses with smaller statues of lesser divinities between them. In addition there were frescos and other colored art. The statues had been built with supports made of timber embedded back into the wall, but this had rotted due to the moisture in the soil. This made the statues extremely fragile and they were in danger of collapse once the sand cover was removed. Therefor he had to recover the statues once they were photographed. Already the statues had lost their heads due to exposure, fragile fragments of these were found scattered in the sand. Stein cited the evidence indicating that the entire shrine had already been covered by sand before the advent of Islam in the Khotan region.

Section IV - The Sculptures of the Rawak Vihara.

As usual, Stein separated his analysis of the art from his discussion of the mechanics of its find and excavation. This strenuous work continued all day and into the evening from 12 to 17 April. By then he had uncovered about 155 feet of wall or a quarter of the total length. In this length he had recovered 91large sculptures and numerous smaller reliefs. He carefully provided detailed descriptions and photographs despite the intense glare, heat and blowing sand. (Plan plate XL). The first statue by the west corner was a huge standing Buddha (fig 61, 69), but intact only up to the shoulders. The exposed right hand is in the standard "Abhayapanimudra" pose. Stein remarks on how closely these statues correspond to those of Gandhara - Grecian style in India. Unfortunately the reproduction of his carefully made photography in the book is very poor. The details of these statues can hardly be seen. His text describes each statue in great detail. Removal of the statues was impossible. To preserve them Stein reburried them as quickly as possible. For more photos of Rawak please see rawak.htm and the list of photos for Ancient Khotan.

Section V - The Date of the Rawak Remains: The Jumbe-kum Site

.Stein notes that the statues themselves offer no evidence of their date of construction, despite their close similarity to the Greco-Buddhist art of Gandhara, which means they are indeed old. No documents or other objects likely to suffer from decay were found. However, numerous Chinese copper coins with the characteristic wu-chu mark were found as votive offerings near the statues and around the stupa. This wu-chu coinage was used during the Former and Later Han dynasties. The second of these was from 25 to 220 AD. but similar coins continued to be minted into the 4^th century. Stein comments that the absence of coins of either an earlier or later era indicates that the site was abandoned between the 3^rd and 7^th centuries AD. Stein left the Rawak site on 19 April for Khotan.

Chapter XV - Departure from Khotan

Section I - Islam Akhun and his Forgeries.

This is a fascinating section of a much different kind than the accounts of archeology or surveying in high mountains. It seems that beginning in the 1890's strange documents written in unknown scripts and languages began to appear for sale at Khotan and Kashgar and in due course found their way to London and to India. Naturally learned scholars began to attempt decipherment and offer all sorts of theories about these 'finds', which were published in leading European scientific journals. Stein was skeptical. As one of his minor projects for this expedition he sought to find the source of these strange manuscripts. Stein's account of his investigation and 'judicial' proceeding is humorous and delightful in every way. And it reveals a great deal about the culture common in Chinese Turkestan in the 19^th and early 20^th centuries. Skipping much detail, we summarize that the whole forgery industry was started and expanded by one Islam Akhun, who recognized in the excitement that buying anything 'ancient' by passing Europeans caused them an opportunity to generate a small profit. From creating such forgeries by hand, using primitive methods he gradually expanded operations into a whole factory that flooded the market. But with this expansion carelessness caused the results to become more questionable. Stein asked the Amban to order the miscreant to appear. Islam Akhun duly arrived at the Yamen under orders on 25 April. Stein conducted intense cross examinations over several days while producing examples of Akhun's work that had already appeared in European publications. The forger was delighted to see that his efforts had seen such publication in scholarly journals but became off guard. Eventually Stein unraveled the whole mystery. He gradually elicited the whole story. But one has to read Stein's full account to enjoy this episode. I read it over and over with many laughs. Stein remarks several times that he was happy that "Eastern methods of judicial inquiry" were not employed. Stein was generous to his friends among the scholars who had insisted on the genuine nature of the forgeries.

Section II - Last Days in the Khotan Oasis

Stein departed Khotan on 28 April. He stopped briefly at Yotkan and acquired yet more artifacts. By 29 April he was at Kara-kash. There he met his former Darogha, Islam Beg, now appointed local Beg on Stein's recommendation. On the 30^th he stopped at another ruin, Kara-dobe at which he found more artifacts. All these days were spent passing through well watered and thriving agricultural territories. On 1 May he started the long trip to Kashgar. At Tarbugaz, the western border of the Khotan government Stein left Islam Beg and Badruddin Khan.

Section III - From Khotan to London

The journey to Yarkand required 6 days. He stopped there for 2 days during an intense rain storm that destroyed many a mud house. Then it was on to Kashgar which he reached on 12 May. The Indian Government had obtained permission from St. Petersburg for Stein to cross Russia on the Trans-Caspian Railroad and to ship his huge treasure in twelve large boxes also via Russia to London. The equipment and written records went back to India directly over the usual mountain passes under the care of Ram Singh and Mian Jasvant Singh. Stein started west again on 29 May, one year after having left Srinagar. He crossed the several passes in the Pamirs still deep with snow and reached Osh in 10 days. Via railroad he passed through Margilan and Samarkand, where he visited the medieval architecture of Tamerlane and the ancient ruin at Afrasiyab where, naturally, he obtained a few more ancient pieces of terra-cotta. He contined to Merv and then crossed the Caspian to Baku. He reached London on 2 July, 1901. The government graciously allowed Stein 6 weeks stay at the British Museum to arrange for division of the materials between that museum and those in Lahore and Calcutta.

We leave it to the reader to study the scholarly appendices.