he Geographical30utnal,Vol. 165, No. 3, November 1999, pp. 306-320 The 'Mysterious and Terrible Karatash Gorges': Notes and Documents on the Explorations by Stein and Skrine

DANIEL C. WAUGH Depart of HistogyandXackson School of Inte7nationalStudiesn Universi of Wsshington,Seattlen Wsshing:ton 98195 USA E-mail:dwaugh@,u.wsshington.edu 7haspaperwss aueZbtedforpublicationinXune 1998

This paperis a discussionof the explorationof the KaratashRiver valley in the 'Kongur Alps'of (), south of .The paperquotes at lengthunpublished doc- uments concerningthe explorationsby Aurel Stein in 1913 and ClarmontSkrine in 1922-23, the firstEuropeans to passthrough the forbiddingKaratash Gorge. This mater- ial is juxtaposedto contemporaryobservations from August 1996, which illustratehow poorlyknown the regionstill is and how in manyways the life of the indigenousKyrgyz herdershas little changed. Several photographs are provided, among them two previously unpublishedones takenby Skrine.

KEYWORDS: China, KaratashRiver, Xinjiang, Stein (Aurel),Skrine (Clarmont), Kyrgyz.

A FEW HOURS' D1OIE south of Kashgaron and narrowvalleys east of Kongurand MustaghAta the KarakoramHighway is LakeKarakul (Fig. and within two days hike from Lake .In the 1).On a typicalsummer day at the lake,tourist 1950s,the Sovietethnographer S.M. Abramzonrelied minibusescome andgo, stoppinglong enough for their heavily on emigre informationfor his study of the patronsto eat at the restaurantand perhapsbe pho- Kyrgyz in Xinjiang (Abramzon,1959). In 1980 the tographedsitting on a camel,against the backdropof Britishexpedition planning a first ascent on Kongur 7546-metreMustagh Ata. The Kyrgyz who live in could not obtain, in advance, detailed topographic Subaxand the othernearby villages enjoy the benefits information(Bonington, 1982; Ward, 1983). Even of electricity(one house even sports a satelliteTV more recently, ethnographersstudying the Kazakh antenna),may try to sell polishedrocks or other trin- and Kyrgyz nomadic pastoralistsin Xinjiang have ketsto the tourists7and may even knowa few wordsin been ableto gainonly limited access for fieldwork,and variousforeign languages. Their primary occupation is then not alwaysin the areasthat had been visitedby largelytraditional - animalhusbandry which involves Europeansearlier in the twentiethcentury (Hoppe, generallyshort migrations to mountainpastures above 1988; 1995).As a result,we stillknow too little about the lake and on the flanksof the greatmountain - but possiblechanges over time in patternsof settlement, some of them clearlymake a good livingtransporting economiclife and communicationamong the herders. climbersand theirgear to the base camp on its north- The authordiscovered, in 1996,how difficultit is to west ridge. Accessible as it is, this area north of obtainsufficient information even to plan a trek into MustaghAta was visitedby Thomas Hoppe in 1990, the mountains drained by the Karatash River whenhe was collectingfield data on the Kyrgyzfor his (Koshan,Ku'shan Ho) east of Kongurand Mustagh study of the ethnic minoritiesof Xinjiang (Hoppe, Ata (Waugh, 1998). In response to initial queries 1995). The Chinese have even produceda superb, aboutthe route, the China XinjiangMountaineering detailed topographical map with inscriptions in Association,from whom a permitwas obtained,was English,encompassing the regionfrom Mt Kongur,on uncertainas to whether it was possibleto cross the the one side of LakeKarakul, to MustaghAta on the KaratashPass to reachLake Karakul. They informed other(Mi Desheng). In short,the areait coverswould the authoronly apera l 9-daytrek that in recentmem- now seemto be well-knownand easily accessible. ory they had abandonedas too dangerousa guided However,to generalizefrom such an impressioncan trekup the Karatash.The authorfound himselfrely- be misleading.In fact thereis surprisinglylimited out- ing primarilyon materialpublished by ClarmontP. side knowledgeof many long-inhabitedmountain Skrine, who as British Consul-Generalin Kashgar regionsin Xinjiang,among them the jumbledpeaks had exploredand mappedin this area in 1922-1924.

0016-7389/99/00034306/$00.20/0 t 1999 The Royal GeographicalSociety 'MYSTERIOUS AND TERRIBLE KARATASH GORGES' 307

gHRG

Fig.1. 771e[Minter and Summer routesfrom TasAkur,g,han toKasAlgar

Ininstances such as this,indeed early explorers' obser- In a sense, the Karatashwas very well-knownhis- vations- not all of them published- often are the torically,but only in its upper reaches.These lie on most detailed source of information,although even whatseems to have been a standardtrade route from retracingtheir routes can be difficult. Yarkandto LakeKarakul and beyond. Probablythe 308 'MYSTERIOUSAND TERRIBLE KARATASH GORGES' first Europeanto follow this route was the British Royal GeographicalSociety on the KongurAlps; his explorerand agent Ney Eliasin 1885,who crossedthe descriptionof both his and Stein'sexploration of the GhijakPass to Chat (Qat) on the Karatash,followed Karatash appeared in Ehe GeographicaljFournal of the riverto its source near the KaratashPass, and November 1925 (Skrine, 1925a). His book, Chinese wenton to the lake (Fig.2). It was the unexploredand CentralAsia, contains additionalmaterial on his ex- treacllerouslower reaches of the Karatash which plorations,and he helped revise descriptionsin the interested Sir Aurel Stein. He became the first General Staff book on routes in Xinjiang Europeanto traversethat part of the river in 1913, (Skrine, 1926; General Staff, 1926). In 1996, the probablyone of his moredangerous undertakings, but author took with him copies of Skrine'sarticle and when he first published about his extended and its accompanyingmap. rewardingThird Expedition,he merely summarized Subsequently,an examinationhas been made of the feat in a paragraph(Stein, 1916; 1923).His length- Skrine'sunpublished correspondence (Skrine, 1922; ier description15 years after the event admittedthe 1923) and uncatalogued photographs, and more difficultyof the undertakingand includeda numberof importantlyStein's correspondence and fielddiary for photograplls(Stein, 1928). With Stein's encourage- 1913 (Stein, 1913aand 1913b),where there is consid- ment,in 1923 Skrinebecame the secondEuropean to erable detail that he did not publish on his venture completethe traverse,while undertakingto extendhis into the Karatash.In order to providea sense of the predecessor'ssurveys into the heart of the 'Kongur way they learned about the Karatashand help pro- Alps'.On his returnto England,Skrine lectured to the vide a base-line for further,systematic study of the

rig. 2. nle KaratashRiver azld the KongurAlps 'MYSTERIOUS AND TERRIBLE KARATASH GORGES' 309

area,portions of the documentsare reproducedhere, below, the present author was alone and largely and commentis made on theirobservations. It is strik- unsupported.The Journeynearly met a tragicend a ing how little conditions of travel and, in many second time, in an incautiousattempt to ford the respects,patterns of settlementand economicactivity Kaying at a point where a bridge had been washed have changedin the more than 80 years since Stein out. Shakenfrom having been sweptagainst the boul- firstentered the KaratashValley. ders of that glacial stream,the author arrivedwith Where Stein tended to understatedifficulties he misgivingsat itsjunction with the unfordabletorrent overcame, in contrast Skrine did not hesitate to of the Karatashwhere he hoped to head up into the emphasizethat the gorgesof the lowerKaratash were narrowestpart of the 'mysteriousand terriblegorges'. 'mysteriousand terrible'. In a letter home dated Passagewas to be from north to south, the opposite 3 October 1922, he provideda vivid impressionof directionto the traversesof Stein and Skrine.No visit the obstacles to the passage of the river (heading was made to the lower Karatash,but, unlikeon their upstream)on his initial,unsuccessful attempt: journeys, the river was followed practicallyto its sourcenear the KaratashPass. Using Steinas the pri- Ournext march was to KayingAghzi ('mouth of the Kaying[glen] '), marysource, his routefrom south to northwill now be not long but distinctlydifficult. The road lay almost continuously followed,beginning with his preparationfor the trip. over the stonesand bouldersof the riverbanks & the swiftrelentless Stein's accomplishmentsrested to a considerable green flood had to be crossed8 or 9 times. The reasonwhy the KaratashRiver is practicallyunknown is that while in winterand degree on meticulousadvance planning.As he was springit is blockedwith ice and snow,in summerthe riveris swollen preparingto return to Central Asia for his Third by melting snows & the increasedvolume of the glacier streams Expedition, on 8 July 1913, he wrote to George makesthe passageof baggageanimals up or downimpossible. Only Macartney,the Britishconsul in Kashgar: the Kirghizfind their way alongappalling 'paths' along the cliffface, trackswhich make you dizzy merelyto look at them from below. The otherpoints about which I shouldbe veryglad to obtainprelim- Once I noticed, high up on an apparentlysheer, smooth-faced inaryinformation through your friendly help are the following:Will precipicea tree-trunkwhich looked as if it had beenglued to the rock it be practicableto traveldown the Kara-tashor BeshkanValley face. I was wonderinghow on earthit had got there,when the Beg from the side of Chichiklik,by [?] Kara-kulL. by the middle of who accompaniedus told me that it had been placed there by September?The routedoes not appearto have ever been properly Kirghizto helpthem over a badbit in summer!The lastfew milesof surveyed,but would suit me particularlywell. In June 1906 Rarn the march were throughthe Tugene Tar, a nasty bit of winding Singhfound it closedlower down by the summerflood. The diffi- gorgedown which an icy windswept continually. The riverwas only culty is probably in the lowermost portion, not higher up. just fordable,it beingmid-October - a fortnightearlier we wouldnot Consequentlyreliable information is morelikely to be obtainedfrom havegot as faras we did.As it was,we had to giveup ouridea of pen- Yangi Hissarthan from the Kirghiz.Of courseI should send all etratingright up to Chimgan,10,250 ft. up & a flourishingKirghiz heavybaggage via Ighizyar... settlement...The marchbetween Kaying Aghzi and Chimganwas Stein,1913a:Fol. 151v foundto be be quiteirnpossible even for riddenhorses, such was the forceof the stream.We weretoo early,the Kirghiztold us. It was not In fact Stein seemsto possiblefor animalsto get up to Chimgantill Novemberor early have been less interesjedin the December.The riveris crossed16 timesin thissection, and the walls Karatashthan he was in two potentialarchaeological of the gorges,many thousandfeet high, almostmeet overheadin sites west and north-westof Yangi Hissar that had someplaces. been mentionedby Forsyth. Skrine,1922 What Stein planned was to explore a new route from Tashkurghanto Kashgar.Although there were Imagessuch as thesewere in the author'smind in late some variants,the normalSummer route (Figs 1 and July 1996 when crossingthe Gez River on a foot- 3) traversedseveral passes, crossing the ShindiRiver bridgenext to the KarakoramHighway and heading valleysouth of MustaghAta to the ChichiklikPlateau, up into the mountains.The firstsection of the route past the settlement of Toile-bulun and on to the largelyfollowed Skrine's on the trailsfrequented by KinkolRiver. The routeemerged from the mountains the Kyrgyz in the Tigarmansu, Chopkana and above Ighiz Yar and met the Kashgar-Yarkandroad Kaying Valleys (Fig. 2). At one point in the early at the regionaladministrative centre of Yangi Hissar. going,owing to a sillymistake in interpretingthe map, The Winter route was that followed today by the the author was in a precariousposition above an KarakoramHighway, west and northof MustaghAta impassiblecliff- it was a sharpreminder of the dan- and then north of Kongur through the Gez River gers of venturingalone. In fact in some ways distin- gorge, where high water in Summerprevented pas- guishedpredecessors had an easier tirne of it. They sage. It was on his way back to India in 1924 that travelledwith a caravanof pack animalsand atten- Skrinediscovered alternative routes, known only to dants and used local guidesall along the way. Not a the local Kyrgyz,which circumvented the Gez during believerin 'toothbrushtravel', as he called it, Skrine high-water season (General Staff, 1926; Skrine, even included in his camp kit 'a tin-lined wooden 1925a;1925b; 1926). bath-tubof immensesolidity' with hot and cold water On 2 August1913, Macartney responded to Stein's tanks(Skrine, 1926). Except for a shortstretch noted requestfor information: I

310 'MYSTERIOUS AND TERRIBLE KARATASH GORGES'

---- Stein 1913 ------Skrine1923 (thcll follows :W Stein 1913)

BEK TARGHAK

Old Karatash Ka>aS< Pass Jr 5 SHARGUT

sMUSTAGH ATA

9er,=, \.

b %,

t.}

Km 0 15

Fig.3. T71gh4Rnta.VAlRivewMantlils.volllAlewelaluluwoacllet 'MYSTERIOUS AND TERRIBLEKARATASH GORGES'

311 KaratashRoute. So far I have been unableto get preciseinforma- largebend to W.' (Stein, 1913b:Fol. 83). He reached tion as to whetheryou can penetrateto the plainsby followingthe the Karatashin another3.3 kilometres,where 'water valleyof the KaratashRiver, down. One of my late dakmen,Rozi appar. less than that united uramsal-] Merki AkhunLadakhi, tells me that he once went to Sarikolby the follow- ing marches: drainage'(Stein, 1913b:Fol. 83). At that intersection there were two yurts,presumably on the spot where 1 KitchikKaraul to GhajikPass. todaythere is a singlemud-brick hut. 2 CrossGhajik Pass to KaratashRiver. Stein considered the route from there up the 3 Cross Karatashriver, follow it up, cross KaratashPass to Karatashto its sourceto be 'wellknown'. In fact, the Shwakte. most detaileddescription today is in the little-known 4 Shwakteto Tagharma. 1926 route book (GeneralStaff, 1926),which leaves some confusionabout the KaratashPass. Before fol- There is evidentlysome difficultyin followingthe KaratashRiver lowingStein down the Karatashthen, comment will be lowerdown; but what that difficultyis, I don't know.According to Rozi Akhun,there is a placecalled Khoja Pakhlan Khoja Mazar sit- made regardingthe route upit to the pass which the uated at the spot (presumablyhalf in the hills & half in the plain?) locals call KaratashDavan (Fig. 2). The entranceto where the KaratashRiver emergesfrom the hills. I have just sent the uppervalley is deceptive,for the Karatashindeed Rosi Akhunto thatMazar to inquireif, by the beginningof Sept.,it contains less water than the Buramsal-Merkiand is possible to travel all the way from KaratashPass down the seems to lead directlyto an impassiblewall. There, KaratashValley to the Mazar. I shall let you know furtherwhen unexpectedly,the Karatashturns sharply south, rising RoziAkhun returns in 5 or 6 days... graduallyin a narrowbut flat valley.After the initial P.S. (dated8 August):...RoziAkhun has just returnedfrom Khoja bend to the south,within one kilometreit turnswest PakhlanKhojam. He tells me that he saw there MohamadKhalik, (as Stein observed)and thence continueslargely in a the KirghizBeg of AktuBazar & got the followinginformation from westerly him: Owingto the volumeof water& the narrownessof the Valley, direction.Along the way in the lowersections it is not possiblejust now to follow the KaratashRiver down to arefields protected by stonewalls, and thenoccasional Khoja PakhlanKhojam. But this can be done after40 days (15th herderhuts scatteredin the starkerand rockysections Sept.?);but his ponieswill have to be unloaded.The valleyis moreor higherin the valley.At least in August 1996, none of lesscultivated by Kirghiz.Mohamad Khalik has his own cattlein it, the bridge crossingsand fords posed particularob- & knowsthe road.The accompanyingsketch [now lost-DVS1 was staclesto a person on foot. A day's hike from Chat madeby him. bringsone to a large Kyrgyz encampment(about a Stein,1913a: Fols. 167-171v dozen yurtsand huts),apparently the Shargutof the 1926route book (General Staff, 1926). Such informationoffered encouragement, but it was The maps, descriptions,and currentobservations not untilhe approachedthe Karatashthat Steincould about the route have what at firstappears to be con- learnwhether conditions would permit passage. After tradictoryinformation regarding the KaratashPass. exploring previouslyuncharted valleys on his way Just above this settlement,the valleyforks. It is likely north from India to the MintakaPass, he arrivedin that the routestraight ahead leads to the difficultpass Tashkurghanon 12 September.The next day he (labelledon Figuresin this paper as 'Old Karatash crossed two passes to Tar-bashi Karaul on the Pass')described by the RussianColonel Kornilov: ChichiklikPlateau but noted he was unableto obtain informationthere about the Karatashroute. By previ- Itsheight is 16,500'and the steepnessof its slopesmakes it one of the ous arrangement,fresh yaks from Buramsalwere mostdifficult passes in the Kashgarianhills. The ascentfrom the side waiting for him at Toile-bulun,where there was a of Karakulis especiallydifficult, being covered with rubble and frag- mentsof rocks. Chineseresthouse. There he partedwith his assistants GeneralStaff, 1926: 169 Lal Singh andJasvant Singh, who were to take the standardroute to the east on through the Kinkol. The way to the Karatash(Kalatax) Pass of the modern Stein turned north, and below the 4420-metre map (Mi Desheng)is different:via the lush pastures BuramsalPass at the ak-oi (yurt) of a local beg, abovethe left fork one headssouth-west and then swings Yuzbashi,he was able to obtain a guide to Chimgan west again, traversingold glacial moraine (the trail (Qimgan),the upperterminus of the KaratashGorge. becomesindistinct on the hardground) and then once Five days out of Tashkurghan,Stein reached the moreturning south-west just before the pass.The mod- Karatash. On 17 September his party descended ern routeinvolves a gradualascent and smoothfooting, along the Buramsal,passing at 12.8 kilometresthe exceptfor a bit of rubbleright at the top (PlateI, andits KizmakJilgha,which leads to the pass of that name altitudeis some300 metreslower than that recorded by and beyondto the KinkolValley. When Skrinefinally Kornilovand Skrinefor what the authorassumes is managed to retrace Stein's journey in 19237 he indeeda differentcrossing. The descentto the Ygebali crossedthe Kizmak(the first European to do so) to get valleyis steeperbut with good footing on easyscree. The to the Karatash.About six kilometresbelow Kizmak, Kyrgyz at Shargut,who apparentlywinter in Chat, Stein noted 'mouth of Kara-tashJ[ilgha] in view & werepasturing some 50 of theiryaks in the Ygebaliand behindsnow covered range which forces river to make crossthe passon a regularbasis to checktheir herds. A 'MYSTERIOUSAND TERRIBLE KARATASH GORGES'

312

PLATE I Approachingthe Karatash (Kalatax) Passfirom the north-east, August, 1996 well-definedpath leads down the Ygebalito its intersec- Whilesome of the buildingshave windows, the typical tion near the upperend of the Kengxuwar(Konsivar) houseis stillthe traditionalwindowless structure, with River (note, GeneralStafE, 1926, calls this the Gez). dirtfloors and a raisedsleeping platform on one side of Thereis frequenttraffic on thistrail, which leads to the the main room. The town is surroundedby wheat settlementsin thevicinity of LakeKarakul. fieldsand on its outskirts,in the directionof Chimgan, Since his goal was not to follow this 'well known' are tombs decoratedwith yak tails on poles, presum- route up over the pass, Stein turned down the ably the 'mazar'noted by Stein.While these observa- Karatash;his diary describesthe approach to and tionsare not systematic,it wouldseem thatthe current routebeyond what is now the 'modern'town of Chat town consolidatessome of the settlementsto which (i.e.'Ghijak' Chat): Steingave separatenames and whichare inscribedon the Surveyof India maps. Most notably,in his time Valleycontinues easy with ample pastures to Kalmak-mazar,5 1/2 m., therehad been two separate'Chats', on eitherside of wherea considerablearea is cultivatedin spasmodicfashion. Below town (the name is a generic Kyrgyz one route crossesa lourspur to long-stretchedplain, Karugach-Kechu, the current wherefields abound. Reached junction with valley leading to Ghijak meaninga place above the intersectionof two rivers Pass at 17 1/2 m. Sarai built by Karim-shahBeg whom I met in [Kirgizsko-russkii,1965]). Probably what is seen here 1900 at Kara-kul& who now lies buried at Ghijaknillgchati is analogousto the process that can be seen in the [=GhijakChat]. Valley turns sharply to W. tow. Chimghan. in ,where to the outsider's Reached Yalghuz-KayingMazar beyond junction with Ghijak eye (but not the locals' consciousness)in modern streamat 18 1/2 m. To NE of it tombs, incl. Karim Shah Beg's Karimabadthere is no separationbetween what had Gumbaz.Ground open & widelycultivated. Marched down through once been distinctlyseparate villages of Altit, Baltit fields to fan, 19 1/2 m., where successionof defilesextending to and so on. Chimghan-aghzi clearly visible... Steill, 1913b: Fol. 84 When Skrine reached Chat in April 1923, he elected to spend an extra day there photographing Located at the intersectionwith the route over the and surveyingfrom one of the ridgesabove the town. GhijakPass, the modernChat (PlateII) is the largest This workprovided him with importanttriangulation village on the upper Karatash.It contains perhaps for the mappingof the Kongur range and produced threedozen mud-brick houses, a schooland a sizeable the strikingpanorama which he laterpublished in his 'club'whose veranda is decoratedwith wall paintings. book (Skrine,1926) and which providesan excellent 'MYSTERIOUSAND TERRIBLE KARATASH GORGES' 313

PLATE II Thevillage of Chat('Ghyak' Chat) on the Karatash River, August, 1996 sense of the topographyof the region.Included here until 7:30. Loadingcompleted by 8:30. WhileI lookedafter it Y. K. (Plate III) is a previouslyunpublished telephoto he [? (Muhammad)Yakub Khan-DW] sent to fix position.Mistakes took at this time, whichis the only photographknown ChimghanValley for our routeof last nighthaving asked for direc- to the author that shows with some detail Kongur tion of Tash-ktlrghan!Rectification of sillyerror took me an hour. fromthe south-east.In his book he alsopublished sev- OpenChimghan V. extendsfor at least5 m. withwidth of 1-1 1/2 m. Branchto W. calledBash-Chimghan, other due S. or S.W.Tersoze. eral photographsof the 'Kirghiz of the Karatash', In bothgrazing grounds and glaciers. includingtwo of a weddingparty to which he and his Lattermust drainShiwakte Range 8: accountfor abundanceof wife Doris were invited en route downstreamfrom water.Cultiv[atio]n in mainvalley as faras eye reaches. Chatto Chimgan. Stein, 1913b: Fol. 85 Travelin this partof the KaratashValley is not dif- ficult.The valley is quite wide, and the riveris easily The ChimganRiver, which flows out of the glaciers fordablejust above Chat. Belowthe town, en routeto on the yet unexplored south-east side of Kongur, Chimgan,the trail in places clingsto clifEsabove the indeedis broaderthan the Karatashand mighteasily riveron the traditionalrafiks of brushand stones.At be mistakenfor the main route in the directionof one such precariousstretch in 1996, the path was Tashkurghan.Even in early morning, water in the crumblingaway, forcingriders to dismountand lead Chimganat the fordnear the intersectionof the rivers their horses across the dangeroussection. Unlike in comesup to a horse'sbelly. Familiar as the crossingis, the day of Stein and Skrinethough, the Karatashhad the Kyrgyzgauge the depthof the waterby throwing to be crossedjust once beforeChimgan, and that on a rocksand listeningfor the sound of their strikingthe bridge whose abutmentshave been reinforcedwith bottom. concrete. WhereStein, Skrineand even Bonington(as late as Steincontinues: 1980) noted relativelyfew permanenthouses in the Summer encampmentsof Arrivedat cultivationof Chimghan-aghzi25 1/2 m. A mud-built the Kyrgyz or in their house with two Ak-ois reachedat 8:15 P.M. Usual troubleabout Winterresidences, today the numberof stoneor adobe- transport.Quartered in dirty Kirgha,relieved only by fresh air 8: brickhuts seemsto have grownsubstantially. Many of clearmoonlight through roof. Chimghansaid to containfifty oiliks, the Kyrgyzstill live in theiryurts in Summer,but others all awayon pastures.8500 ft. move from one permanenthut to anotheras they go Sept. 18. Up by 5:30A.M. but animalsdid not turnup in batches higherinto the mountains.Thus Chimganitself now 314 'MYSTERIOUS AND TERRIBLE KARATASH GORGES'

PLATE III Ihe eastridge of Kongur, telephotofrom above Chat. Photob} C.P.,Skline, Apail 1923. Collectionof theRoyal Geographical Sociefy with the Institute of British Geographers (tempowa number Skwine JVo. 2). ,Sesnote on photograph.s. consistsof moret]lull bI dozcIl mud-plasteredhouses of Althoughotherwise the authorcarried all his food somesubstance. As in Stein'sday, there is extensivecul- and gear,both at Chimganand Chat he negotiatedto tivation.The village is surroundedby wheat fields, hire a pack animaland guide, in the firstinstance to watered by small irrigationchannels and protected ensuresurviving the ford and in the secondbecause of fromanimals by stone or mud-brickwalls or by fences concernsabout the vaguenessat the edge of Skrine's of brushcuttings. map. Both Stein and Skrinehired their animalsand It was with obviouspride that the author'shosts in guidesin stagesand frequentlycomplained about how Chimgan showed off a new house that was nearing difficultit was to get awayon schedulein the morning. completion.It had a decorativccalved friezearound Just as in their day, the author discovered that the ceiling boardsin one room, and in the adjoining European/Americanimpatience does little to ensure room there were severalpillars with muqarnas-style a prompt, early departure.The local Kyrgyz obvi- capitals.Frames for the windows were being trans- ouslyknow the routesall along the Karatash,but still portedfrom Chat on donkey-back.The standardcon- seem willing to hire for only short segments of the structionof roofs is to lay a few large beams, which journey. The owner of the horse ridden to Chat normallyare not carvedor shaped and not covered insisted that his son, who was to accompany the with boards, place across them numerouspoles of author,could go no further,since he had to returnand smalldiameter, lay branchesor a reedmat on top, and head downstreamthrough the gorgesto Khan-terek. then put down a layer of mud mixed with straw. The guide from Chat to thc KaratashPass was ini- Timber is broughtdown from the higher mountain tially reluctantto commit himself to more than the valleys on donkeys,but it seems likely that the reed day'sjourney to Shargut,although with time on his mats come from the bazaars in the market towns hands(and probably feeling flush from havingnegoti- below. The relativeprosperity reflected in this new ated the sum he did for his services)he then seemed house in Chimganseems to have been the exception: more than happyto travelwith his friendsor relatives in the adjoiIliIlghouse, where the authorwas invited from there over into the next valley to help herding to breakfast,those present squattedin the dirt in a the yaks. windowlessroom, and the wok containingthe rice The easyfords and broadvalleys were soon behind, (mixedwith a little meat and greens)rested on some as Stein's expedition headed north from Chimgan emberson the floornext to the door. into the narrowKaratash Gorge (Fig. 2): 'MYSTERIOUS AND TERRIBLE KARATASH GORGES' 315

Startedfor Khan-terek at 9:40AM. Cultivation& arboursfor 1 1/2 m. some cases for hundredsof metres along the valley. T hen beyondhamlet of Tring-doweroute ascends cliffs to ca. 200 ft. The clusterof huts standsright at the point wherethe aboveriver. Fixing at 3 m. Kara-yulghun. valley narrows; the modern trail drops to the river In narrowgorge with patchesof vegetation(Suget, myrtle, roses) and at one rockoutcrop disappears briefly in a moder- to Ortang,a cultivatedfan, 4 1/2 m. Then path dropsto riverbed& currentof water up to the hips. It continuesto a overboulders. For footmen a calltileer bridgeof ate skirtsdifficult Parri perhapsat the samelocation as the 2 poplartrullks at 6 m. Whole knownas Bakuchak. Beforefixing cantileverbridge, firstgreeted by by distantdust haze in gap of valley. one Stein describesbut rathermore substantial,being Crossillgto r. ballk about 1/2 m. belosvbridge. River c. 40 yds. reinforcedwith steel rods and concrete.Once across wide and 2-3' deep. Got thoroughlywet in feet. Aftersome hulldred the riverat thatpoint, the trailremains on the eastside yards1. bank regainedfor about 1 m. Difficulttrack under banks of until the next bridge, below Kara-tumshuk(see conglomerate.Projecting spur forced us to crossthird time to r. bank below).The authorcamped near some sizeabletrees at 8 m. Waterup to saddleflaps. Haltedto dly myself& for hasty in a narrowcultivated arca where therewas one hut lunchat 2 P. M. andstarted down by Maidan-tal.Small patch of culti- (Stein'sMaidan-tal?). Such patches of cultivation(in vation followedby steep cliffsalong the foot of which the reduced the given instanceprobably no more than two-tenths waterlevel has left an extremelydifficult track over rock ledges & boul- is a ders. The rock-coulissesseemed to close passagecompletely when of a hectare)generally are foundonly wherethere camelsappeared ahead moing unloadedover a realRafik. rhis spur side streambringing clear water down into the river. knownas Ara-sundepl-oved most difficult bit of march.Photos. [Stein, Unlikein Stcin'sday, therewerc few placeswhere the 1916,fig. 4, reproducesone of these,which provides an excellentsense patllwas OI1 a rafikattached to the sideof a sheercliff. of the difficultiesof the gorge-DW.] Ourerrant gtlide had proved so There were tracesof that kind of constructionin one ullreliablethat I greetedsight of baggagewith relief. or two sectionswhere the trailhad washedout at river Remarkableskill in constluctionof Rafikswith tree trunks& level, but otherwise,much of the trail was a solidly brushwood,strong enough to carryc.amels. How the lattermallaged beaten path well above water level and on relatively to get along slippely boulders,heaped up irl conftlsion,seems a levelground. mystery.Also risks of quicksands. Stein, 1913b:Fols. 85-86 About midwaybetween Chimgan and the junction of the Karatashwith the Kaying,the trailclimbs high This section of the route seems to be much easier abovethe river;one can lookacross to a lushfarmstead today. As one approachesthe first,very small,settle- witha mud-walleddwelling complex and fencedwheat ment north of Chimgan (presumablyStein's Tring- fieldsnext to the intersectingstream, the Kara-tumshuk dowe),there are smallirrigation channels extending in (PlateIV). In low water,the riveris probablyfordable

PLATE IV Thefarnzsteadat Kara-tumshak, viewfrom the south, August, 1996 'MYSTERIOUSAND TERRIBLE KA1RATASH GORGES' 316 here, as it was in Stein's day, although just beyond, the the river,was somewhatanalogous. In coming along rapids once again become deep and treacherous. Stein the very dry plateauto its north,back away from the describedthis section of the route as follows: gorgeof the river,the bridgeon which to crossto the east bankwas neverseen (hadit been therein Stein's BeyolldAra-sullde spur, 9 m., routeascended a seriesof steep-edged day, he would not have had to backtrack).So the plateatlsfrom which view opened tow. easierportion of defiles.The authorfollowed the well-troddentrail up behind the camelshad crossed to 1.bank 8 did not returnto our sideuntil below the isolated cultiv[atio]n.8 farrnhouseby debouchureof Kara- pinnacle(see Skrine'spreviously unpublished photo- tumshukJ[ilgha], 9 1/2 m. Here fixing.The look of valley below graph,Plate V) and descendedto the farmsteadacross raisedhope that worstlay behindus. Butwe had scarcelyproceeded the side-stream.There, fortunately,two young infor- for a mile along easy slope when we were met by camelsreturning. mantswere found.With his few wordsof Kyrgyz(the Two loads had been takenacross to 1.bank, but the fordproved so key one being kEipik,bridge) and by sketchingon the riskythat Musa 8 the intelligentYuz-bashi of Khan-terekwho fortu- ground,the authorlearned from the boys that it was natelyhad met us lastnight & agreedto guideus, decidedto takethe necessary to backtrackto the bridge and cross. other camels back to fold at Kara-tumshtlk.Our own half-crazy Conversationwith them established that the bridgewas Kirghiz guide hacl clecamped& when discovered by Khabil, attackedwith stolle and kllife. one of fourbetween the Kayingand Chimgan.Three Marcheclback to Kara-tumshukford, 12 m., & foundit c. 50 yds. of the four are constructedpurely in traditionalcan- wide & 2 1/2' deep at average.Crossed small stream of side valley, tileverfashion, using logs, rocks,and brush;only the descendingover steep talus fan, & climbedplateau with striking rock bridgenearest Chimgan had modern reinforcement. pinnacleof Kara-tumshuk,13 1/2 m. Descent to fan where the Stein was obviouslymistaken in thinkingthat the camelswere waiting provedtroublesome & when at 14 m. fresh worstwas overonce past Kara-tumshuk: Parrisfaced us ahead,decided to pitch camp. Foundpleasant tree- ringedterrace 8 haltedby 6 PM. Yuzbashioffers to go ahead for Sept. 19. Lightrain fell all night;but when I awakenedmy men at freshtransport, but rain& darknessprevent start. Elev[atio]n. circ. 5 A.M. Tordukol,the energeticYuzbashi, & Kabil had already 7400 ft. [in margin:c. 7150R1 . First evening with comfortabletem- startedfor Khan-terek.The baggagemoved off by 8:15 A.M. but peraturesince Yasin. troublewas encounteredalready after a few millutesill descending Stein, 1913b: Fols. 86-87 to riverbankowing to rocksoverhanging path. After less thana mile camelshad to crossto r. bank& then to regainus oppos.to Chatkir The author's experience in approaching the rock pin- cliffs(photo). nacle of Kara-tumshuk (Kyrgyz, 'black beak' or Some cultivationpassed belourPitlik-aghzi. But after 1 1/2 m. 'snout') from the opposite direction, on the west side of fromcamp track took to rockledges & Rafiks& we had to partagain

PLATE V TheKaratosh Rivergorge, with the rockspire of Kara-tumshuk, viewfrom the north. Photo by C.P. Skrine, April 1923. Collectionof theRoyal GeographicalSociey with the Institute of British Geographers (temporaty number Skrine;No. 1). Seenote on photographs 'MYSTERIOUS AND TERRIBLE KARATASH GORGES' 317

PLATE VI Cantileverbridge, Karatash River valley, south of confuence with Kapring, August, 1996 with baggage.The riverhere was too deep & rapidto be forded.So who continueddownriver. The author had entered the ponieswere left to swimacross while we had to ascendthe steep the KaratashValley by descendingthe Kaying,which spurdividing us fromKaying-aghzi. Up to 2 m. trackled alongpre- Skrinehad ascendedin 1922 when unableto cipitouscliffs of slateyrock, 50-100' above river. Ill placesRafiks just proceed as bad as in Hunza.Then followedto 3 m. a very steepzigzag climb furtherup the Karatash.Nowadays, the cultivation along slopes of decomposedrock & shale, riskyin places. Height Stein and Skrinereported at Kaying-aghzi,the inter- aboveriver prob. 700-800'. section of the rivers,seems to have been moved two The descentto N was fortunatelysafer, leading over earth slopes. kilometresup the Kaying,where there is a clusterof Reached trianFlar plateau of cultivationat mouth of Kaying huts on its north side. The lower Kaying shows few J[ilgha]after 3 1/2 m., by 11:30A.M. Elev.6850 ft. Riverc. 15 yds. signsof recentSummer traflic though, and it is possi- wide, I-1/2' deep,much clearer water than that of mainriver. ble that the settlementhas been abandonedowing to Steill, 1913b: IFols. 87-88 the significantgulleying that has occurredfrom flood- ing of the streamwhere it is located.The main access The modern trail in this section only hints at Stein's dif- to the upper Kaying Valley is principallyover the ficulties. Much of the time it is on broad terraces along ChopkanaPass from the ChopkanaValley, the route the west bank and only once or twice drops to near which Stein's Kyrgyz guides took to Khan-terekin water level on loose scree. About the only reminder of order to avoid difficultiesin the Karatash below how dangerous the route may once have been is a rafik Kaying-aghzi(see below). clinging precariously to a clifTabove the rapids on the Khan-terek,at the mouth of the Chopkana,is now opposite bank. A bridge that apparently did not exist the Winter residencefor many of the Kyrgyzwho early in this century (PlateVI) now enables one to avoid summerin the Karatash,Kaying and Chopkana.As climbing the major spur Stein describes. Another two seemsto be the case at Chat, it may now consolidate kilometresor so of hiking on the east bank brings one to what had been separate Winter settlementsat the the intersection with the Kaying, just below which is a mouth of each of the main side valleys.If the author bridge at the location noted by Skrine (General Staff, understoodcorrectly from one of the Kyrgyz passed 1926). The narrowest sections of the Karatash Valley in the Karatash,the administrationat Khan-terekhas were now behind (PlateVII), although the challenges of the responsibilityof keeping the trail through the the route ahead were still substantial. Karatashin repair.Skrine published an excellentpho- Here the author's route diverged from that of Stein, tographof the canteliverbridge at Khan-terek,which 'MYSTERIOUSAND TERRIBLE KARATASH GORGES'

318

PLATEVII TheKarataJh RivergorgeJ;om thenol th, taken 3-4 kiloletresabove the confZuence with Laying, August, 1996 is the more interestin£;because of the water pipe Scantysupply of uraterfrom Chopkana J[ilgha], head of whiclldimly which is stlsperldedabove it acrossthat narrow part of visiblethrough dust haze. Wide alluvial fan, 10 1/2 m. the river(Skrine, I 926: facing 146).As noted in one of Changeof sceneryvery striking after leaving Tugine-tar. Instead of bold scarps& pinnacles,much-eroded valleys on both sides.The the neighbouringvalleys, the Kyrgyzstill very skilfully slopesstreaked red & greyshow sameexcessive erosion as I remem- engineer their svatersupplies in traditionalfashion, beredfrom Tokuz Davan & lowerGez-darra. It seemedas if the very althoughusing modern tubing instead of the hollowed skeletonof thesehills had been laidbare. Instead of slate& a reddish logsof Skrine'sday to bridgegulleys. stoneresembling trachite, sandstone takes now the chiefplace. For the last of the Karatashtraverse, beginning Tordukol& Khabil had come to Khan-terekby a pass which fromthe Kaying,Stein relates: leadsfrom KayingJ. to top of Chopkanavalley (Khan-terek). Thus the worstpart of the rivergorge avoided.It appearsthat thereis a continuationof this upperroute also to Cllimghanby a pass,practi- After 4 m. alollg leEtballk passecl Bash-ktipruk[='BIidge head'], a cablefor yaks.... well-madecalltileszel- bl-idge, 2 I yds. long, of 4-5 trunks.It spallsa Sept. 20...paidoff bravecamels & poniesfrom Chimghan,with nalrowpal-tof bed & givesaccess to Terek-kichikJ[ilglla],4 1/2 m., rewardsto KirghizAbdul Kadir, Abdu Saiyid,Yomush & Tila Bai, throwlghhich largegrazing grounds & a pass tow. Yangihissarare jolly fellows.... reached.Here we fordeda widerstretch of riverand contilluedovel Stein, 1913b: Fols. 88-90 easy terraceson r. ballkto bold spur,5 m. which camelscould just skirt.Beyond the riverbedwidells to 1/4 m. Broughtup by another gorge at 6 m., with almostperpendicular cliffs rising to c. 3000' on A decade afterStein, Skrinestruggled with the lower 1.bank. Karatash, abandoned his attempt to go beyond Here the liZerhad to be crossed& recrossedwithin 6 1/2 m. Movingalong beds of botlldersoll 1.bank we clearedthe lastforbid- Kaying-aghzi)and headed insteadinto the then still dingrock gate of tlle valley,known as Twigine-tar,7 1/2 m. Marksof unexploredKaying Valley. When he finallymade it old Rafik(Yaskak) along impossible looking cliffs. through the KaratashGorge the followingApril, it At last,at 7 1/2 m., we emergedwith reliefinto opell valley where was from Chimgan.In a letter home, 19 April 1923, it isjoined by KalabelJ[ilglla] . We were aboutto do our fifthcross- he reported: ing whellfresh pollies were broughtup by Yuzbashi& Khabil.Ford 60 yds.wide, 3' deep.6450' elev. 3:30PM. This time there was only about half as much [water]as then [in Fulther passageeasy. Over flat broad riverineplateaus, partly October]:even so, we had to crossa riveraverag[illg] 40 ft. in width, cultivated,we got to our 6th crossing,oppos. to Khan-terek,10 m. 2 1/2 ft deep & flowingat 6 milesan hourover rough stones twenty- 'MYSTERIOUS AND TERRIBLE KARATASH GORGES' 319 six times in 2 days. Once it took us two hoursto do one mile, the Turdi Ali, probably in his sixties. He knew a bit of ponieshaving to be unloadedalld reloadedone by one four or five world geography - though he had to be reminded that tmes &the kitmanhandledalongledges or amongboulders.Itwas the Atlantic Ocean separated the continents-and hugefun & we bothenJoyed it thoroughlythough that day we didn't '. . could pltchour tents tlll nearly 8, whenlt wasqulte dark, & thenlt wason a clte the names of some of the Sovlet leaders. He natTowbit of landwith mightydark clifEs on one side & the foaming even llstened patlently as the author tried to explain river& even mightier & darkercliffs on the other! with diagrams and his few words in Kyrgyz how an Skrine,1923 e-mail message from home being carried in hard-copy had originally been transmitted. Just as information about the outside world must be His formal description of the Karatash was used to rare in the Karatash and presumably would not update the official route book (General Staff, 1926). always be comprehensible there, for those who would Just before he left Kashgar for good, in 1924, he spent wish to know its Kyrgyz, their world is as remote. Not more time in the Kaying and adjoining valleys and having any information in Kashgar, Macartney had to became the first European to cross the Kepek Pass, make a special inquiry of the Kyrgyz at the mouth of about which Stein had heard (although clearly it is not the Karatash concerning conditions along the river. A negotiable by yaks; see photographs in Skrine, 1925a). day out of Tashkurghan, Stein still could not learn When the author approached the Karatash in details about the lower Karatash, and he was greatly August 1996, fresh from a violent dunking in the relievedtohavetheKyrgyzfromKhan-terektoguide Kaying, he expected to find a dark gorge hemmed in him through the gorges. Eighty-three years later, the by forbidding cliffs. In all likelihood, steps would have author could best learn about the Karatash Pass by to have been retraced because of an impassible ford or having a local guide lead him to it. It was indeed 'huge washed-out bridge. The reality proved to be otherwise fun' to be following, albeit timidly, in the bold foot- - even at its narrowest, the gorge does get the sun, in steps of Stein and Skrine. More significantly, the trek many places there are trees and other vegetation, and offered an opportunity to step back in time in an area the bridges were all in place. Now during the Summer that has been visited by few outsiders since they were the Karatash seems to accommodate regular traffic there. between Khan-terek and Chimgan. The modern world is encroaching, even though to a considerable J%oteon photographs by C.P. Skrine degree the older patterns of settlement and pastoral The two unpublished photographs reproduced here economy have been preserved. Although they may with the permission of the Royal Geographical Society still be relatively isolated in the Summer when in their are from previously unidentified glass negatives in the mountain pastures, the Kyrgyz presumably no longer RGS collection. The view of Kongur from above Chat rely on entrepreneurial peddlars to bring them goods (Plate III is a telephoto version of the portion of his but can readily access urban markets from their panorama (Skrine, 1926: facing 116) under the in- Wintersettlements.Clothesarestore-bought;ifanyof scription 'Qungur II', taken 11 April 1923. In the the costumes Skrine photographed still exist, they panorama, the peak is quite indistinct. The author must be tucked away in family chests. A few of the knows of no other photograph showing this view ofthe younger men sported mesh hats inscribed 'World east ridge, which in the picture apparently hides the Cup',andtherewaseventheoccasionalcheapdigital true summit. See Bonington, 1982: facing 136, for wristwatch. Yet items as simple as stainless steel what is essentially the reverse of this view - looking spoons or plastic water bottles seem to be very much downfrom high on the south ridge of Kongur to what is in demand, and although medical care apparently is apparently the upper part of the Chimgan Glacier; in available, even those who undoubtedly needed them otherwords,approximatelyinthedirectionofChat. had no spectacles. A young wife could write; her older Skrine's photograph from within the Karatash husband apparently could not. Gorge (Plate V) looks south toward the rock pinnacle At the same time, there seems to be little active of Kara-tumshukand was taken apparentlyon 12 April curiosity about the world beyond the next pass. In 1923, the same day he had departed from Chat. Chat, one of the younger men did know about About six kilometres below Chat, at a place he called and its president Askar Akayev and Bek Targhak, he had been an honoured guest that requested that the author take Akayev a message to morning at a Kyrgyz wedding. Two of his published help arrange a visa to visit. The Kyrgyz of the photographs ofthe 'Kirghiz ofthe Karatash' (Skrine, Karatash are very insistent in reminding the outsider 1925a: facing 393, topmost picture; Skrine, 1926: fac- that they are Kyrgyz (i.e., not to be confused with ing 132, bottom picture) are apparently the ones he Chinese); few, if any, in Chat could read a permit in notes taking there. Chinese. They fingered the stamped seal (in Kyrgyz, Skrine's photographs are largely uncatalogued. The tamgha)assuming it did provide some assurance the Royal Geographical Society has most of his surviving author had official sanction to be there. The best negatives and many of the fine enlargements which informed of those met in Chat was a genial older man, he made himself in Kashgar and displayed at the RGS GORGES' 'MYSTERIOUSAND TERRIBLE KARATASH 320 India Office A the negativesare in the Oriental and the time of his lectureon the KongurAlps. Photo920/ 1. around a few of Collectionsof the BritishLibrary, largecollection of his CentralAsian prints and REFERENCES Library,Oxford, MS Stein, A. 1913a Correspondence.Bodleian S.M. 1959 Kirgizskoenaselenie Sin'tszian-uigurskoi Macartney, fols. 150-151v, 165- Abramzon, fnmdy Stein 96; for exchange with oblasti Kitaiskoi narodnoi respubliki. ofthe BodleianLibrary, Oxford. avtonomnoi Moscow: 17lv. Quoted by permission arkheologo-etnograficheskoiekspeditsii, Vol. II. 31-Nov. Bodleian Library, Krgizskoi as -, 1913b Field diary 1913, July nauk SSSR: 332-69. Translated are from the typed Izdatel'stvo Akademii Oxford, MSS Stein 207, 215. Quotations on theKtrgiz population in theSinkiang- of the Bodleian Ethnographicalinformation transcript,MSS Stein 215, by permission Oblastof the ChinesePeople's Republic, JPRS UigurAutonomous Library,Oxford. in Central Asia, 24,572/1964: 1-54. , 1916 A third journey of exploration Kongur:China's elusive summit. London: 192-229. Bonington,C. 1982 1913-1916. Geogrl. 48(2): 97-130; 48(3): andKansu (Records of Hodder and Stoughton. , 1923 Memoiron maps of Chinesefurkistan 1926 Routesin Sinkiang.Simla: Governrnent Survey GeneralStaff, India the Survey of India, 17. Dehra Dun: Trigonometrical India Press. of Range. In OfEce. T. 1988 Kazak pastoralismin the Bogda of explorationsin CentralAsia, Hoppe, onan , 1928 InnermostAsia. Detailed report Svanberg,I. (eds)Ehle Kazaks of China:essays underthe orders of Benson,L. and Ean-suand Eastern Iran carried out and dessnbed Studia Multiethnica Upsaliensia, 5. Uppsala: Press. ethnicminoriy. H.M. IndianGovernment. Vol. I. Oxford:Clarendon Research,Uppsala University: 201-39. Eurkistanand Centrefor Multiethnic of India, 1906-1908 Map ofportions of Chinese Xinjiangs:Kulturunterschiede undintereth- Survey C.I.E. andhis - ,1995 Die etnischenGruppen illustratethe explorations of Dr MA. Stein des Instituts fur Asienkunde Kansuto nischeBeziehungen. Mitteilungen R.B. Lal Singhand R.S. Ram Singh.Dehra Asienkunde. assistants,surreyors Hamburg,258. Hamburg:Institut fur Survey Office. Sheets 2 (Kashgar),3 slovar'.K.K. Iudakhin, Dun: Trigonometrical Kirgizsko-russkii,1965 Krgizsko-russkii 4 (TashkurghanRiver), 7 (Yangi-Hissar). entsikopediia. (Kara-TashRiver), comp. Moscow:Izdatel'stvo Sovetskaia furkzstanand Kansufom Me sunveys scaletopograph- , 1922 Maps of Chinese Desheng. Kongurfagh -MustagAta(1: 100 000 explorations1900-01, 1906-08, Mi House. madedunng Sir Aurel Stein's map).Lanzhou: Xi'an CartographicPublishing Survey Office. ical Library, 1913-1915. Dehra Dun: Trigonometrical C.P. 1922 Letters to his mother. British 3 (Sarikol). Skrine, 154/8. Sheets 2 (Kashgar,Muz-tagh-ata), India Office Collections, MSS Eur. F. TopographicalMap Oriental and , 1925 - 1926 Quarter-inch(1:253 440) Quoted by permissionof MrJ.W. Stewart. Survey Office. Sheets Oriental and Series. Dehra Dun: Trigonometrical Letters to his mother. British Library, (Mustagh-Ata), 420 , 1923 by per- 42M (1925) (Kashgar), 42N (1925) Oflice Collections,MSS Eur. F. 154/9. Quoted India (1926)(Tash-Kurghan). of MrJ.W. Stewart. southernSinkiang. Geogrl mission with Ward, M. 1983 The Kongur Massifin The Alps of Qungur. Geogrl. 66(5): 385-411, , 1925a 137-52. topographicalmap. i. 149(2): Alps' and the unknown appended 226-50. Waugh, D. 1998 Exploringthe 'Kongur roadsto Kashgar.i. Cent.Asian Soc. 12(3): ,1925b The side of MustaghAta. Himalayan. 54: 25-32. ,1926 ChineseCentral Asia. London: Methuen.