O <(^3NVS01^ V oe. i: < ,5MEUNIVERS'//. % % ^\^f' /aaiv# :5 ex? 5ME m. #• 5 ^ "^ffi # ^l-UBRARYOr, ^lUBRARY(?/r ■A. 5 1 ir^ % I- ^lOSANCElfx^ o ■%jaAiNa-3WV ^OFCAIIFO/?^ 4.1 '^^c»Aaviiaiii^ ;^ 4.0FCAIIFO% §? ^ ,—.'1 V c:^ ^lOSANCEU-f \ >;,QFCAtlFO% I ITTh AN EXPEDITION INTO THE CENTRAL TIAN-SHAN MOUNTAINS CARRIED OUT IN THE YEARS 1902-1903 M l*».-(K.S..KMAII.lOT a3,0oj Kr.), TAKI-.N FROM A l-OINF (AIlnUT 15,000 1 T ) ON ITS lif /^ ' TMI: Ufl-r.R INVUIIKK riLACIER (MUCH lORESIIORTKNED). SOUTHERN AND S ' WC»ir.RN SLOI'l. THE CENTRAL TIAN-SHAN MOUNTAINS 1902-1903 BY DR. GOTTFRIED MERZBACHER PUBLISHED UNDER THE AUTHORITY OF THE ROYAL GEOGRAPHICAL SOCIETY 4745^ LONDON rIN MURRAY, ALBEMARLE STREET, W. 1905 PRINTED BY HAZELL, WATSON AND VINEY, LD., LONDON AND AYLESBURY. i| CONTENTS IS PAGE ^ INTRODUCTION i CHAPTER I _ FROM PRZHEVALSK TO NARYNKOL AND THROUGH THE H MUKUR-MUTU VALLEYS lO CHAPTER II THE VALLEY OF BAYUMKOL I9 "^ ^ CHAPTER III '"--J THE SARY-JASS VALLEY AND THE SEMENOFF GLACIER . 36 ^^ CHAPTER IV TO THE INYLCHEK GLACIER AND FARTHER SOUTH . . 60 CHAPTER V FROM THE KAPKAK VALLEY TO THE GREAT MUSART VALLEY 76 CHAPTER VI NORTHERN MUSART VALLEY, MUSART PASS, AND SOUTHERN MUSART VALLEY 82 CHAPTER VII FROM THE MUSART VALLEY TO KASHGAR . . . -99 vi CONTENTS CHAPTER VIII PAGE EXCURSIONS ON THE SOUTHERN EDGE OF TIAN-SHAN TO COLLECT PALAEONTOLOGICAL MATERL-^LS . . . I03 CHAPTER IX THE SOUTHERN MARGIN OF TIAN-SHAN BETWEEN KASHGAR AND UCH-TURFAN Ill CHAPTER X TO THE KHALYK-TAU AND BACK TO UCH-TURFAN . . I26 CHAPTER XI THE SOUTHERN TRANSVERSE VALLEYS — THE ALLEGED AND THE ACTUAL BREACH, FORCED BY THE NORTHERN WATERS 139 CHAPTER XII THE SABAVCHY GLACIER 151 CHAPTER XIII THE KUKURTUK VALLEY 158 CHAPTER XIV FROM IHF, KUKURTUK VALLEY TO THE BEDEL VALLEY AND OVER THE BEDEL PASS TO THE NORTH . . 1 66 CHAPTER XV THE WATERSHED BETWEEN NARYN AND SARY-JASS AND THE PASSAGE OF THE SOUKA PASS TO LAKE ISSYK KUL . 1 76 CHAPTER XVI SURVEYING ON THE SEMENOFF AND MUSHKETOFF GLACIERS 185 CHAPTER XVII SECOND VISIT TO THE INYLCHEK GLACIER, AND DISCOVERY OF THE TRUE POSITION OF KHAN-TENGRI . . . 193 CONTENTS vii CHAPTER XVIII PAGE OVER THE ACHAILO PASS TO THE KAYNDV GLACIER . . 217 CHAPTER XIX FROM THE KAYNDY VALLEY OVER THE UCH-SHAT PLATEAU TO THE KOI-KAF 225 CHAPTER XX THROUGH THE KAYNDY VALLEY TO THE SARY-JASS AND INYLCHEK AND BACK TO THE TEKES .... 237 CHAPTER XXI TO THE BAYUMKOL AGAIN AND THEN TO THE LITTLE MUSART VALLEY 245 CHAPTER XXII VISIT TO SOME ALPINE LAKES 254 CHAPTER XXIII THE DONDUKOL VALLEY AND THE NORTHERN MUSART VALLEY AGAIN 263 CHAPTER XXIV OVER THE TEMURLIK-TAU TO KULJA 272 CHAPTER XXV SUMMARY 276 NOTES CONCERNING THE MAP . . . .283 INDEX 287 LIST OF ILLUSTRATIONS KHAN-TENGRI, SOUTHERN AND SOUTH-WESTERN SLOPE TELEPHOTOGRAPHIC VIEW OF KHAN-TENGRI . THE "MARBLE WALL," OR PEAK NICHOLAS MIKHAILOVICH TELEPHOTOGRAPHIC VIEW OF THE SUMMIT OF KHAN-TENGRI TELEPHOTOGRAPHIC VIEW OF KULU-TAU VIEW FROM A PLATEAU BETWEEN THE SARY-JASS AND TYS-ASHU VALLEYS ...... PEAK IN THE SOUTHERN BORDER-RANGE OF THE INYLCHEK JIPARLIK GLACIER ..... IN THE UPPER TEREK VALLEY (KHALYK-TAU) PEAK EAST OF JANART PASS .... EASTERN MARGIN OF JANART PASS ISSUE OF KUM-ARYK FROM ITS GORGE . SABAVCHY GLACIER, DEBOUCHING OF A LATERAL GLACIER NORTH-EAST ...... MOUNT CATHERINE, IN THE BORKOLDAI RANGE CHAIN BORDERING MUSHKETOFF GLACIER ON ITS SOUTH LAKE KARA-KUL-SAY ..... PARTING OF INYLCHEK GLACIER SUMMIT OF KHAN-TENGRI, NORTH-WESTERN SLOPE HEAD OF KAYNDY GLACIER .... VIEW TO THE HEAD OF SAIKAL VALLEY, FROM A RIDGE EASTERN BAYUMKOL GLACIER Frontispiece Facing page 26 28 52 62 64 66 90 134 142 144 148 156 176 190 198 198 222 222 246 AN EXPEDITION INTO THE CENTRAL TIAN-SHAN, IN THE YEARS 1902 AND 1903 PRELIMINARY NARRATIVE INTRODUCTION When in 1892, on a journey into Central Asia, I first made acquaintance with a small portion of the Central Tian-Shan, I received, even by a mere flying visit, abiding impressions of its magnificent mountain chains. Later on, these impressions were renewed through reading the masterly descriptions of the celebrated Tian- Shan pioneer, P. P. SemenofF, and through study of the reports of his successors, N. A. Severzoff and J. W. MushketofF, who have earned for themselves liigh honour by their researches. The desire was accordingly kindled in me to gain more accurate insight into the highest regions of this mountain chain and its glaciers, and also to con- tribute somewhat to their exploration. Extensive travels, however, in other mountain 1 2 INTRODUCTION lands and labours of large compass in other fields barred the way for ten years to the gratification of my desire. At length in January, 1902, during my stay in the Russian capital, the initiatory steps were taken. There, encouraged by the assured support of the Imperial Russian Geographical Society, especially its President, his Imperial Hiirhness the Grand Duke Nicholas Mikhailovich, and its Acting President, Senator P. P. SemenofF, I determined to start that year on my travels into the Tian-Shan. SemenofF's valued counsels and the study of the rich Russian literature on that mountain chain, handed over to me most oblig- ingly by the Secretary of the Imperial Russian Geographical Society, Prof Grigorieff, confirmed me in my opinion that one summer would not suffice for the accomplishment of any substantial result in the high regions of the Central Tian- Shan, so extensive and so difficult of access, but that, first and foremost, experience would ha^'e to be gathered respecting the technical difficulties, awaiting the explorer in ice-and-snow regions of quite a unique type. From the first, therefore, I was resolved to devote at least two years to tlic enterprise. Our knowledge of the orographical and geo- logical structure and of the flora and fauna of the Tian-Shan lias been enriched by many eminent Russian explorers. Its highest regions, however, buried in snow and ice, had hitherto remained but very imperfectly known. A more thorough ex[)loration was needed to answer the many THREE PRIME REQUISITES 3 questions, respecting the structure of the central parts, which a glance into the existing maps at once called forth, and to throw light on many dark points in the later history of the mighty configuration of the chain. Our knowledge, again, of the glaciers of Central Tian-Shan has been greatly increased, more particularly by the explorations of A. W. von Kaulbars, and by the expedition of I. W. IgnatiefF and A. M. Krassnoff, fraught in many respects as it was with important results. There still, however, remained much, particularly in respect of the largest glaciers, craving elucidation. In order to explore extensive glacier regions and their environment, and to unravel the complex structure of the parts not easily surveyable, it is necessary to follow up the glacier valleys to their head and to climb peaks of great elevation with a view to obtaining a comprehensive plan and orientation. For such a task three prime requisites were wanting to my predecessors : practice, experience, and outfit. It seemed to me therefore imperative to enlist " Alpinism " in the service of geographical science in the Tian-Shan, in accordance with the example of so many travellers of brilliant accomplishment in other regions of lofty mountains. I accord- ingly invited to join me in my enterprise one of the best approved of modern Alpinists, the engineer Hans Pfann, of INIunich, a truly valu- able aid, and further engaged a young and vigorous Tyrolese mountain guide, to whom, the following year, was added a second guide. 4 INTRODUCTION By way of special provision for the geological service of the expedition, and with a view to amassing a palaeontological collection, I secured the assistance of a young and energetic geologist, not to be daunted by difficulties in the ground to be surmounted. Prof. Steinmann, of Freiburg, Baden, was good enough to recommend to me one of his pupils and assistants, the young geologist Herr Hans Keidel, who on my invita- tion joined the expedition. With the trusty support of such assistants I might hope to obtain some data of value in the service of science. Unfortunately the time assigned to the pre- paration of an undertaking of such compass, extending over so long a period, for procuring and testing the indispensable instruments, apparatus, and manifold outfit, was all too scantily allotted. Only by dint of feverish exertion and the active help of self-sacrificing friends, among whom I will name only the celebrated mountain photographer, Cavaliere Vittorio Sella of Biella and tlie Caucasian explorer, M. von Dechy in Odessa, Mas the expedition for 1902 set on foot in tolerable time, though, indeed, several weeks later than desirable. In this report, written in Tashkent, immediately after the rcluiii of the expedition from the mountains,' it is of course impossible for me to ' 'I'liiM report was .icspatdied from Tashkent on April 18th, IIKH, loiij^ Id-foro llic publication of the narratives of Dr. Fried- rir-liMMi (" l-orschiuiffsreiHen in den Central Tian-schan u. Dsunga- riwhcn Ala-tiu " : MiftfiPi/inufrn ilvr (.'iw/rapfii.srhm dcscllschaft iu IIuniburK, IJd. XX. August l'J04) and ISigr. Ciiulio Brochcrel ("In Asia PURPOSE OF THIS REPORT 5 render an exact account of all the work done throughout this long and toilsome journey, or to communicate all the observations of scientific interest. The purpose of this report is rather to give particulars of the itinerary of the expedition and a general narrative of its experiences, especially those new or hitherto unknown. The more elaborate digest, embody- ing comprehensive deductions, must be reserved to a later date after the rich collections, amassed by the expedition, have been scientifically exam- ined and arranged ; this latter task will, however, presumably claim a lengthened period of time. Accordingly, the more detailed report of the journey, for which Herr Keidel has undertaken the elaboration of the geological and geotectonic part, can hardly be published till a somewhat remote date. It seemed therefore advisable to give in the preliminary report rather more than a bare enumeration of data and to render it at least a provisional picture of the districts traversed. In this report I have endeavoured more particularly to embody observations on the present and past glacier conditions of the Tian-Shan and on peculiarities in the physical features of its valley Ceutrale " : BoUettino deJla Societa Geogrufica Italiana : July 1904), when I had not the slig^htest knowledge of the results of the former or even of the routes followed by the Italian expedition. Hence no reference to either publication will be found in the present report, the appearance of which would have been too long delayed, had I attempted to incorporate comparative notes after my return. Moreover, I was quite unaware, wlien writing my report, that several valleys and localities, to whicli I had every reason to suppose mine to be the first visit, had been previously reached by the Italian expedition. 6 INTRODUCTION formations ; subjects to which, throughout the expedition, my attention was specially directed. On the other hand, in order not to give to the report a compass which would retard its publication, botanical, zoological, and chmatological observations will have to be almost wholly omitted. The quoted figures incorporated m the report will of course be taken as of only approximate vahdity, and are stated in round numbers, seeing it ^^^ll require some considerable time till the different calculations in question have been com- pleted. The various heights given will be accepted as ha^'ing only at most an approximate vahdity relatively to one another. On May loth I left JMunich, accompanied by Herren Hans Pfann and Hans Keidel, joined at X'^ienna by the previously engaged mountain guide, Franz Kostner, of Corvara ; we thence repaired to Odessa, whither had been despatched the larger part of the luggage. Here we were detained a few days in complying with the Custom-house formalities and taking over the provisions of con- ser\cs, biscuit, etc., which, thanks to friendly assistance, were already awaiting us. Owing to the fact that the Imperial Russian Ministry of Finance had obligingly granted free entry to my outfit, instruments, apparatus, etc., the trans- actions witii tlie Custom-house were rapidly settled. 'JMie stay, moreover, in Odessa till the departure of the steamer was made grateful to us by the ainial)lc hospitalities of the noted explorer, M. von Dc'chy, and of the Crimea-Caucasian Mountain ASSISTANCE FROM RUSSIA 7 Club and its most courteous and helpful President, Prof. Ilovaisky. On May 25th we landed at Batum, and thence proceeded to Tiflis, where we were again delayed a few days. There I received maps, most kindly left for me by the chief of the Topographical Department of the Ordnance staff in St. Peters- burg, Lieut. -General von Stubendorf. There, too, all my instruments were retested at the Observatory. At Tiflis I had the high honour of a reception from the President of the Imperial Russian Geo- graphical Society, his Imperial Highness the Grand Duke Nicholas Mikhailovich. As his Imperial Highness had in Petersburg facilitated for me the introductory steps to the expedition, he took a warm interest also in its development and assured me of his further assistance. After being joined at Tiflis by the Preparator, E. Russel, of Piatigorsk, the expedition next pro- ceeded, via Baku, to Krasnovodsk, and then, by the Trans-Caspian railway, to Tashkent. There, in consequence of letters of recommendation from the Imperial Russian Ministry of Foreign Affairs and War, and thanks to the letter of accreditation of the Imperial Russian Geographical Society, I met with the friendliest reception from his Excellency the Governor- General of Turkestan, Lieut.- General Ivanoff. In the most handsome manner official papers were given me, ensuring me the support of all authorities in the Russian lands to be traversed by me. Seeing the expedition was to extend over two 8 INTRODUCTION years, the provisions and materials of different kinds had to be divided, and the part intended for the second year packed and forwarded to Kashgar. Thanks to the active support of my revered friend Herr R. Schubert in Taslikent, this and other affairs were happily disposed of On June 9th, therefore, the expedition of five persons, now heavily encumbered with lugoage, was enabled to enter on its lumbering tarantass-passage through the Central Asiatic steppes. \A'hile 1 was on my way alone from Pishpek, which I left on June 18th, to Vernoie, there to present myself to the District Governor of Scmircchcnsk, his Excellency Lieut. -General lonofF, and to receive from him special letters of introduction to the authorities under his a(hninistration, Hcrren Pfann and Keidel made an excursion to tlic Alexander Mountains, climbing one of the highest peaks. Meanwhile, under charge of Kostner and Russel, the heavy luggage was farther carried by Dunganian carters to Pr/licvalsk. On June 24th I rejoined my fellow- travellers in 'J'okmak, whence our marcli lay along the north bank of Lake Issyk Kul to Przhevalsk. There I had the pleasure of en- countering the expedition of Prof Saposhnikoff, of 'I'onisk, and its mcnil)ers, among them Dr. M. I''ric(h'icliscn, of Hamburg. Friendly greetings were exchanged. At first I was under some apprehensions, lest tlie Russian expedition and mine might clash in their respective routes through the high niounhiins -a contretemps which in the TWO EXPEDITIONS 9 interest of science would be all the more regret- table, inasmuch as in the extensive and little- explored Tian-Shan there was ample room for the exploratory enterprise of more than one expedition. At once Prof. Saposhnikoff frankly communi- cated to me his programme, and we found that our routes would only cross eacli other in the Sary-jass valley at the foot of the SemenofF glacier. As, moreover, the Russian expedition had assigned to that strip of the mountain chain no more than a few days — quite insufficient for a thorough exploration of the SemenofF glacier, included as an essential item in my programme- — my appre- hensions were happily shown to be groundless. CHAPTER I FR03I TRZHEVALSK TO NARYNKOL AND THROUGH THE MUKUR-MUTU VALLEYS A SERIOUS drawback to the progress of our expedition was the delay for nearly a week of the arrival of our luggage at Przhevalsk. Not till July 2nd was our advance by the San-tash pass to Karkara begun. In crossing the pass (7,200 ft. ; 2,155 m.) made know^n to the world by Scnienolf and ServerzofF, we had the opportunity of gathering in our first carboniferous fossils in the Tian-Shan. On tlie descent of the pass, which leads through extensive tertiary deposits, one conies up on the first indications of glacier forma- tion in this region : porphyry granite and syenite blocks transported thither by ice from the heights of Kungeu and Kuuluk-Tau. Soon after, on the desfcnt from tlie tertiary sandstone heights at Taidy-biilak, the wide, green-mantled, ancient lake- floor of Karkara ((5,000 ft.) is seen below, encom- passed on the south by a long, many-peaked, calcareous chain (Rash-ogly-tagh), bearing on it sniaJi glaciers, and towering above the lake-floor to a height of 4,500 ft. (1,200 m.). On its A TEMPORARY TOWN 11 margin the ancient lake-terraces are in good preservation. On the north and north-west the \^^de basin is enclosed by low, flat tertiary ridges, offshoots of the Chiil-adyr, behind which crop up here and there the much more considerable heights of Ketmen-Tau. On the south-west edge of the basin, in these tertiary deposits, considered hitherto deprived of fossils, Herr Keidel had the rare good fortune to collect a small fauna, which may prove of great importance for the determination of a part at least of these tertiary deposits. Alpine meadows, dressed in flora of surpassing beauty, adorn the wide, high-lying floor, strewn with debris. In its midst there rises up every year, from INIay to October, a spacious town of metal houses and wooden booths — the famous yearly market, that is of such great importance to the extraordinarily numerous Kirghiz population of the Tekes, Chalkody-su, Kegen, and Charyn region. Thousands of Kirghiz tents collect in a wide circle around the wooden town. This is the mart where the Kirghiz exchange their products of wool, hides, sheep, and horses for the manufactiu'cd goods, exposed for sale by dealers mostly of Tartar race. Here, in a secluded green Alpine bottom, completely withdrawn from the highways of the world, walled in by mountain chains glittering in glacier snow, the traveller may admire a lively tumult of business and contemplate modes of commerce, belonging to an epoch of 12 FROM PRZHEVALSK TO NARYNKOL culture, unkno^^^l for centuries past in Europe. Here he may study scenes of a picturesque charm, not to be easily surpassed anywhere else. During the four months of the yearly market the seat of the administrative authorities of the regions in question is transferred to this spot. It thus happened that the head of the Narynkol district, J. I. LikhanofF, on whom depended the future fate of the expedition, had his official seat at this odd market-town. Here the greater part of the riding- and packhorses, as also the saddles, covers, headgear, etc., appertaining thereto, had to be bought. Here, too, a number of Kirghiz " Jigits " (troops serving for escort) familiar \^'ith the moun- tain routes and some porters from the discharged Narynkol Kossacks had to be engaged. The safe transport of later supplies of provisions to the high mountain regions and many other matters had to be settled. Thanks to the energetic support of Herr Likhanoff", these affairs were satisfactorily disposed of in a few days. On July 7th I was able to continue my march to tlie Kossack village of Narynkol (Okhotnichi). 'J'hc way thither leads from the luxuriant grass- plains of the ancient lake-basin, over the adjacent undulating high plains, into a spacious green landscape, the configuration of which is throughout due to former ice-action. The peaks of the long, niiich curved mountain chains, Bash-ogly-tagh and Kapyl-Tau, shooting up in the south are intersected by wide, trough-shaped, high-lying valleys, each of which is occupied by a small A GREAT VALLEY 13 field of nev^ and a small glacier. As may be distinctly perceived, these are only the remains of former ice-cm-rents of considerable extent, the course of which may be satisfactorily traced by the ground, lateral and terminal moraines re- maining intact and now coated over with vegeta- tion. All superficial forms, characteristic of a landscape vacated by the ice, even drumlins, may here be observed. On a second visit to this district, the way led me into a larger lateral valley (Bash-kara-bulak), where I had the oppor- tunity of more closely examining these typical forms of a vanished glacial epoch, and to follow them into the cauldron-shaped hollows of the mountain chain, where great masses of glacier snow once rested. Beyond Sary-jass-tuty station you leave the river-bed of the Chalkody-su, and, crossing the mountain chain through the glen of Tute, enter the uppermost valley of the Tekes. On the way the traveller, viemng the circular wall of the mountain chain, is already impressed by a feature characteristic of the Central Asiatic mountains and especially the Tian-Shan. The mouths of the great transverse valleys of older origin are always wide and their floor at the same level as that of the principal valley. This is due to the enormous amount of debris piled on the latter in a region poorly drained, and covering the base of the margin of the mountain chain. The farther march to Narynkol is confined almost entirely to the region of the tertiary 14 FROxAl PRZHEVALSK TO NARYNKOL formation and of the younger river and lake deposits. Only a stage distant, however, in the above-mentioned glen of Tute, you cross a zone of quartz-porphyries and hornblend-por- phjTies, at the foot of which lie the tertiary deposits. On July 9th the expedition entered Staniza Narynkol (G,200 ft. ; 1,760 m.), lying near the northern foot of the first lower chain of the Central Tian-Shan, hard by the Chinese frontier. This place served for a length of time as headquarters for the explorations in the high mountains. Seeing our arrival took place three weeks later than had been planned, there was no time to lose if any results were yet to be harvested in what was left of the short summer. Herr Keidel took in hand the in\'estigation of the tertiary formation of the Tekes plain, and of the carboniferous limestones, towering up behind it. On July 10th I made my first mountain excursion, and, with Herr Pfann, the Tyrolese Ivostncr, and a Kossack from the Tekes valley, rode some twenty versts (thirteen English miles) down to the mouth of a transverse valley, cutting in a southward direction into the mountain chain, a valley known by the name of Mukur-mutu. l}et^vccn the great transverse valleys of the Great and the Little Musart rivers, which, in an approxi- mately southern course, cut into its northern slope, the great chain is again divided, principally by three short transverse valleys, filled with exceed- ingly dense pine-forest— the JMukur-mutu valleys— KUTINGY 15 which, after short courses, terminate in an extensive high plateau region. The Kahnuck population of the Tekes valley understand in general, under the name of the Mukur-mutu, the whole slope of the mountain chain between Little and Great Musart with all the transverse valleys, intersecting it. According to this acceptation, Mukur-mutu w^ould therefore designate the region which to the east and the west of the great valleys named is bounded on the south and south-east by the valleys of Maralty and Dondukol, and on the south-west by Uertenty valley : valleys of which, later on, there will be much to say. The district is known by the name also of Kutingy. I was here able at once to assure myself that the delineation of the whole of this strip of land in the forty-verst map does not suggest even a remote idea of the reality. Of the JNIukur-mutu valleys, for example, only one is shown, and that, too, just three times longer than its actual course. In the high plateau region in which the Mukur-mutu valleys originate erosion has caused only broad channels of little depth. The many-peaked chains, walling in the head waters of all above-mentioned valleys, form likewise the verge of the plateau mass, which on its turn swells up into some dome-shaped heights. By the forty-verst map it looks, as though Khan-Tengri towered up here in the southern enclosing wall of the plateau, and to make sure of the fact was the motive to this excursion. We wandered but a short distance through the most western of the Mukur-mutu valleys — their mouth about 16 FROM PRZHEVALSK TO :N"ARYNK0L 6,400 ft. (1,850 m.) high — and soon turning sharp eastwards, we made a very steep cUmb over slopes wooded and carpeted with Alpine meadows of unsurpassable luxuriance, displaying a marvellous Alpine flora over old ground moraine-deposits. After some time craggy ranges of slabby rose- coloured granite break through the steep, abraded beds of schist and the soft forms of the green- clad diluvial deposits covering them, and you mount to a stage of the plateau, where we en- camped at a height of about 7,700 ft. Thence we turned southward and ascended to a far higher stage of the plateau, and there soon reached a zone of dark, richly fossiliferous, dense limestones. AVithout undergoing a high degree of crystal- lisation, these limestones, along with the stratified granites cropping up between them, have been subjected to enormous pressure, so that most of the organic inclusions were crushed beyond identi- fication and also transformed into silicates, of which, too, but very little is to be obtained. The booty gathered was therefore small. On a second visit, tlic following year, to the valley, we were fortunate enough to pick up, at another spot, a somewhat better collection, from which the age of the lime- stones was determined to be lower carboniferous. Tlicse dense, dark limestones alternate with light- coloured and somewhat granular calcareous slates, and, farllicr on, with red argillaceous-calcareous slates. The whole series follows the dip of the granites (average direction N. by 35^ E.), which in their turn, follow, farther to the south-east, on TELEPHOTOGRAPHIC VIEWS 17 the limestones. The series varies in its course, however, very much, and higher up, passes into an almost opposite direction. There you find yourself in a region of dislocation. A beautiful cauldron caused by subsidence, with a little lake at its bottom, still lies on the boundary between the granites and limestones in this latter formation. Higher up, a part of the calcareous mass, composing the plateau seems to have subsided to a consider- able length, in a southward direction towards a trenchlike depression, the axis of which, directed east -south-east, is the axis followed by the high valley of Maralty, cutting transversely through the plateau. A more detailed description of this interesting region would exceed the limits of this preliminary report. Be it only added that the spot where the better preserved fossils are to be found, lies exactly in a plain of fracture. We mounted one of the highest dome- shaped protuberances of the plateau (about 11,000 ft. ; 3,400 m.), there photographed the magnificent peaks of the Uertenty valley, and took telephotographic views of the high-peaked, ice-covered chain which is planted in front of and parallel to the main ridge here trending east-south-east. Of this main ridge only a few elevations could be seen, towering up behind the parallel chain. Did Khan-Tengri rise at the spot, where in the forty-verst map and in all other maps, it is represented, its pyramid must inevitably have been seen from our standpoint. All we learned by our excursion was therefore only 2 18 FROM PRZHEVALSK TO NARYNKOL the confirmation of the opinion, previously sug- gested, namely, that in this cardinal point the maps were all of them at fault. The task therefore devolved on us to determine the actual situation of Khan-Tengri. CHAPTER II THE VALLEY OF BAYUMKOL The first advance made towards the solution of this problem led us into the large transverse valley of Bayumkol (wrongly named by some travellers Karakol and Biankol). The axial direction of this valley, some forty miles long, changes many times from the spot, where the river bursts forth from the mountain chain ; the valley cleaves its way through the high land in an approximately southern direction, but afterwards bends round to the south-east, then to the east- south-east. It next again takes a southern direction and, at its termination, bifurcates into two branches ; one trending south and south-west, the other south- east, both of them occupied by considerable glaciers and surrounded by chains wholly covered with glaciers. The peaks of these chains are among the highest of the Central Tian-Shan, rising to 20,000 ft. and more. These chains form part of the central watershed of the Tian-Shan. The river, rushing in large volume down the valley, takes, on issuing from the mountains into the vast basin-shaped expansion of the Tekes valley, at first an eastern direction, flowing through the capacious basins of 19 20 THE VALLEY OF BAYUMKOL two ancient border-lakes, once connected. Of one of these lakes of late tertiary age the margins, com- prised of sandstones, sandy clay and slate beds, are in excellent preservation. The stream next strikes north-north-east by the Staniza Narynkol, and at last, taking a northerly course, reaches the Tekes. Our way, therefore, into the mountain valley led us through a depression beginning some twelve miles up the river. This depression, in the neighbour- hood of the Staniza Narynkol, is marshy and hedged round by a broad belt of dense high-grown bush. In this thicket, through which our road led us, there were buzzing millions of gadflies. These set upon my horses, which had just been brought down from the cool mountain meadows, with such fierceness that, becoming restless, they dis- placed their burdens, and so, getting scared, some of them took to flight. In a twinkling the others all followed their example. In less than a minute all the twelve packhorses, throwing off" their loads and dragging their girths after them, bolted in mad gallop in all directions all over the wide steppe and through its thickets, continually kicking at the packages with their hind legs. Instruments, apparatus, provisions — everything was flung to the winds. Speechless with horror, I looked on at the spectacle. Should the altogether indispensable articles of outfit, more particularly the instruments and apparatus, be smashed, many months would be needed to make up for the loss. The ex- pedition would be wrecked at its threshold. The cases of a number of packages were burst open A STAMPEDE ' 21 under the horses' hoofs, and their contents, especially the boxes of preserves, flung higgledy- piggledy among the tall grass of the steppe. While some of the ^' Jigits" and Kossacks hurried after the runaway animals, the others searched among bush and grass for the packages. After a time it appeared that the scare was worse then the scath, and that I had got rather cheaply out of the disaster. The inost valuable articles were found to be all of them undamaged. Help came from Narynkol, whither I had despatched a messenger. The horses were caught and brought back ; the damaged cases, straps, etc. were hastily mended. After the loss of five hours the caravan was again ready for the march. It was some time, however, before I recovered from the shock. So soon as the enclosed basin just mentioned, about four and half miles long, had been left in our rear by a narrow passage in its enclosing wall, we entered another basin, much more extensive still, Avhose northern enxdronment is formed by a moderately high calcareous range. The tertiary lacustrean deposits of the enclosing wall of the basin we had just left are continued along the foot of the calcareous range in a series of terraces. In this calcareous wall, exactly opposite the mouth of the Bayumkol valley, and at the north end of the lake-basin, three and one-third miles broad, there is noticeable a gatelike breach, through which there now flows, in a straight course north- wards to the Tekes, the inconsiderable streamlet 22 THE VALLEY OF BAYUMKOL Ukurchy. On the other hand, instead of con- tinuing its northern course, to which in the wide plain there is no opposing obstacle, and thereby reaching the rocky gate in the north and so making a direct passage to the Tekes, the Bayumkol river, on issuing from the mountain chain, all at once bends to the east. It there- upon forthwith encounters a calcareous cliff, Tas-tepe, barring its path, and which it is bound to break through. It has sawn out a deep bed in the calcareous rocks at the edge of the mountain chain in order to be able to continue its farther course east, north-east, and north, till at last it reaches the Tekes. What could induce the river to make this comphcated journey ? Evidently in former times it had taken a straight course to the north across the plain and through the breach, which it had once itself effected. This continued to be its course till in the ice age either masses of ice or boulder-deposits blocked the passage and compelled it to take an eastern course. To the importance of the former glacier age the ancient masses of moraines lying on the skirts of the mountain ranges in the Tekes valley give testimony. In their form and arrangement I was able to read that the ice masses in the past, pushed forward from the mountain chain, liad flooded the crest of the first border chain. The mouth of the Bayumkol valley is about fbiir-fil'ths of a mile wide; the bottom lies at the same level {v'uk p. 13) with that of the princi- pal valley (about 7,000 ft), and, owing to the ANCIENT TERMINAL MORAINES 23 enormous masses of deposit, piled on the ancient floor, it rises at a quite moderate incline (about 174 ft. per mile). The valley spreads into basins as much as a mile in width, and separated from one another by contractions of no more than 1,100 ft. Most of these expansions con- tained lakes, dammed by the ancient terminal moraines, which, in the period of the successive retreat of the earlier glacier, got thrown up, one behind the other. Only in the case of two of these expansions could I make out other causes for their origin. One, near the mouth of the valley of the Ak-kul, has without doubt been formed, or at least developed, by lateral erosion of the valley river. Another, at the mouth of the lateral valley Tyr-asha, arose in consequence of a fault between limestones and chloritic slates. Of most of the ancient terminal moraines, only inconsiderable remains are preserved. Only two of them still block up the valley as enormous walls. One is at the mouth of the lateral valley of Alai-aigyr, which, running eastwards, affords access to the Saikal valley (Little Musart). The other is at the mouth of the Kenem-begu valley, which leads west to a col, giving access into the Ashu-tyr valley. Both moraines, each of them over the third of a mile broad, owe their preservation to mighty mountain-slips, rolled dow^n and cover- ing to a great extent the moraine walls, one of the mountain-slips being of granite, the other of phyllitic rock. Where these vast masses of 24 THE VALLEY OF BAYUMKOL blocks rest, atmospheric influences and the strength of waters, endeavouring to clear away the morainic walls were spent in vain. The river was obliged to force a passage at both places in a deep ravinelike glen, where, to all appearance, it resumed and deepened the bed it had occupied before the ice age. Besides these two monumental witnesses to the once mighty glacier conditions of the valley, evidence to the same fact is to be found also in the form of high- Ijnng pohshed rocks and in the heaps of moraine debris or terraces of glacial rubble along the walls of the valley, preserved everywhere, where the slope is not too steep. These debris heaps form higli terraces many miles wide, now on the right, now on the left bank. In many places moraine debris may be seen towering more than 800 ft. above the level of the river. At the mouths of many lateral valleys, especially that of the Ashu- tyr Aalley, the moraine walls of very considerable magnitude, formed of the debris, are in excellent preservation, at the mouths of others they have got washed away and shifted. iVt the entrance of the Bayumkol valley the enclosing walls are formed of granite, to which, a little higher up, succeed fossiliferous limestones and calcareous slates, as well as dark argillaceous slates, to which, in turn, again granite succeeds. Gianites of very various character, limestones, cHlcarcoiJs slates, argillaceous slates, also gneiss ajul otiier crystalline slates, alternate along the whole length of the valley in unintermittent GEOLOGICAL FORMATION 25 sequence and in very peculiar conditions of stratification. Into this matter, however, there is the less need to enter here, as Herr Keidel has taken a geological profile of the valley, which he will publish and elucidate in the geological part of the more complete report. I may, how- ever, here call attention to the fact, that granite and gneiss take the prominent part in the structure of the enclosing walls, that the sedi- mentary matter always reappears, pressed between the granites, without, however, showing any sign of contact-metamorphosis, and that the granites appear to have been vehemently ground. This points to folding processes which have affected both kinds of rocks in common. We further note the embedding of diabasic rocks, more especially diabasic slates. Lastly, here too attention must be drawn to the important fact, first established in the Bayumkol valley and since confirmed in all the Tian-Shan valleys, leading to the principal ridge, which were visited by the expedition — the fact, namely, that in every case the crystalline rocks reach no farther than proximity, nearer or more remote, to the main watershed. This latter is itself built up exclu- sively of sedimentary rocks, which have undergone transformation through dynamo-metamorphic pro- cesses, in part also in consequence of the eruption of diabasic rock. In the structure of the most central and highest region of the Central Tian- Shan, not only limestones of different kinds have taken part, but also dense, dark, argillaceous 26 THE \^ALLEY OF BAYUMKOL slates of very various formation, dark slates, having the character of roofing-slate, prepon- derating, and marbles of different colours, mostly- white, grey or with Hght streaks. The valley presents the character of a northern Alpine valley, showing excellent Alpine meadows and extensive and very dense pine forest {Picea Shrenkeana), with which are here and there com- bined deciduous trees, such as sorbus, willow, cornus, mountain ash. The somewhat auriferous alluvium of the river was, more than forty-five years ago, when the district still belonged to China, exploited by the Chinese. Later, the attempts to find gold were prosecuted with the aid of extensive plant by Russian speculators. It would appear, however, that they did not pay, seeing the constructions are no longer worked and are fallen into decay. The river carries an uncommon volume of water, and, in the warm hours of the day, rushes in raging flood down its bed. It is therefore dangerous to cross, as I learned to my cost. One of the packhorses, slipping, was at once carried away into a whirlpool, whence it was rescued only with the greatest exertion. Of its load a package was lost, containing all my personal belongings. Just before reaching tlie mouth of the side valley of Ashu-tyr, and behind a belt of wood whicli stretches diagonally across the main valley, the magnificent pyramid of Khan-Tengri suddenly comes into view. The mountain looks so near TELEP110T..,.KAIinC VII, W UF KllA.N-I£.Nt.lU (aKOL'T 23,0^ II.), TAKEN FKOil .NOKlil, h Ko.M THE MIUDLE COURSE OF THE BAYUMKOL VALLEY. DISTANCE ABOUT 24 MILES. [To face p. 26. KHAN-TENGRT 27 as to convey the illusive impression that it is planted in the background of the Bayumkol valley. Arriving at the end of the great granite rockslip, lying on the top of the first ancient terminal moraine at the debouchement of the lateral valley of Alai-aigyr, one sees, far below, the middle course of the Bayumkol valley as a forest-encircled basin with a quite level bottom ; here, on the other hand, the view is magnificently closed by Khan-Tengri, and it seemed again as if at the end of the Bayumkol valley we should reach the foot of the giant mountain. There we found indeed the head of a valley, exhibiting magnificent glaciers and a circle of very high mountains clothed with ice from their feet to their summits, but Khan-Tengri was not among them. Owing to the fact that the mountain has no rival and overtops the highest summits of all the neighbouring ranges by over 3,000 ft., it can be identified from points of sufficient altitude and at sufficient distance in any direction whatever. The determination of its position, with a careful exploration and topographical survey of the Bayumkol glacier, and the geological investigation of the ranges bounding the valley, would shortly form our task. Our camp was pitched at the end of the main valley at an altitude of about 10,500 ft. (3,200 m.), near the spot where the glacier arms, approaching one another from the south-east and south-west, unite in one common terminal tongue which ends about 10,660 ft. (3,250 m.) in altitude. 28 THE VALLEY OF BAYUMKOL While the south-western glacier (the longer one) forms a rather compact, not A^ery steeply inclined, ice-field, about eight miles (twelve versts) long, which has its origin between snow-clad peaks on a lofty snow-clad ridge (first trodden by me the following year), the south-eastern glacier is somewhat shorter, but much steeper and more rugged ; it is formed by the union of three ice-streams which, breaking through gorges in the ice-clad ramparts of the valley, unite in a circus-shaped basin. On the surface of the ice a number of funnel-shaped lakes are hollowed out. The wide ice-basin is immediately overhung by a mountain which for height, massiveness, and boldness of form is the most commanding of the giant peaks which rise round the Bayumkol glacier. From its ice-clad shoulder on the north-west side a per- pendicular wall about 6,500 ft. (2,000 m.) in height, on which, of course, neither snow nor ice can remain, falls straight to the rugged broken ice of the glacier-floor. This precipice is of white, grey and streaked marble, and for this reason we called the mountain the " Marble Wall" ('Marmorwand'). Like Khan-Tengri, this commanding mountain stands conspicuous as a landmark of the Central Tian-Shan, an orientation point. It can be recognised by its remarkable height and by the fact that it towers up just at the point of union of the main ridge with its branches, visible far and wide from every point on the high ground. Seen from the Tekes plain, it is known by its remarkable form and METEOROLOGICAL CONDITIONS 29 its precipitous marble side. It was not till later that we proved what an important part it plays in the formation of the Tian-Shan. During two weeks which we spent in the Bayumkol valley, we were busy with the investi- gation of the glaciers and their surrounding hills, Herr Pfann in addition taking measurements and making a survey, while Herr Keidel prepared a geological section of the valley and collected the material necessary for its corroboration. These labours were, however, often interrupted and sometimes hindered by two causes — inclement weather, and the refusal of the porters to work on difficult ground. The summer of 1902 was on the whole distinguished by unsettled weather. Moreover, in the high valleys of the Central Tian-Shan this variability was affected in a conspicuous manner by local conditions. As was often proved in the course of the journey, and as could be established from the meteorological observations, which were recorded regularly twice a day, each separate valley has its own meteoro- logical character, which depends on the direction of the axis of the valley. For the Bayumkol valley the determining factor is that, being wide open to the north, it debouches immediately into the wide Tekes plain. The layers of air stagnating there and rapidly cooled during the night are, towards midday, set in violent commotion, owing to the uncommonly effective insolation of the floor of the steppe ; they rush in storm towards the mountain range and penetrate the wide chasm 30 THE VALLEY OF BAYUMKOL of the Bayumkol valley to its upper stretches, where, rapidly decreasing in temperature on the comparatively cold slopes, which extend to the north and north-east, they condense their vapour. The weather in the upper valley was, as a rule, good in the forenoon, but the force of the current of air, which regularly ascends from the plain in the middle of the day, is so great that it displaces that which, earlier in the day, prevails in the upper valley, and the latter does not regain its ascendency and restore calm till evening. With great regularity towards noon the air became dull ; about two or three o'clock torrents of rain or snowstorms began, and after- wards, during the evening and night, fine clear weather prevailed. These winds, however, con- dense their moisture on the middle heights, and the highest ridges receive but little of it. At our headquarters, about 10,500 ft. (3,200 m.) high, the Aveather was always worse than at our camps, 3,300—6,600 ft. (1,000—2,000 m.) higher, where our work was mainly carried on. In the valley the precipitation was more continuous and more copious. The dry, loose condition of the snow on the extreme heights of the Tian-Shan (of which more hereafter) receives from these facts at least a partial explanation, though doubtless other circumstances are also partly accountable for it. As for the porters, one-half of the Kirghiz deserted in the night, and the other half refused to serve, if they should have to climb on foot over INDIFFERENT PORTERS 31 glacier ice at the higher elevations and carry on their back loads of quite moderate weight. Our discharged Kossacks were somewhat better, but they would not undertake what an Alpine porter of only average strength accomplishes with ease, to say nothing of the loads, carried by the natives in Sikkim and Kashmir. They usually showed the greatest aversion to snow at a great altitude, though I had equipped them all with Tyrolese shoes nailed for mountain wear, as well as crampons and ice-axes. If to the unfavourable factors already mentioned there is added the bad condition of the high snow, which, especially on the northern and eastern slopes, lay dry and powdery on a surface of ice, one can easily imagine the wretched difficulties which opposed our investigations. I soon perceived that the extreme heights of the Tian-Shan are no proper field for the gratification of the Alpine craze. Our incipient purpose of climbing the " Marble Wall " had to be given up, since the porters could not be induced to carry the baggage, indispensable for a prolonged sojourn over heights of 16,000 ft. (5,000 m.), to a saddle at the foot of the north-western arete of the mountain. We had pitched our little mummery-tent at a spot free from ice, about 12,500 ft. (3,800 m.) high, on a de- pression in the north-eastern rampart of the eastern glacier. Thence we made excursions to the lofty granite peaks in the north-west, from 14,000 to 14,800 ft. (4,300—4,500 m.) high, wi'eathed with small glaciers. The granite is there altered in an 32 THE VALLEY OF BAYUMKOL unusual and multifarious manner in consequence of mountain pressure. Next we went to the com- pletely snow-clad schist summits, 16,500 to 18,000 ft. (5,000 — 5,500 m.) high, to the south-east of the high camp, to get from these heights an insight into the formation of the surrounding chains, and into the course of the valleys, which separate them, as well as to take photographic (especially tele- photographic) panoramic ^dews. These would be of great value for the completion of the topographic surveys, in which, moreover, the detail was obtained by photography. Of these excursions the following was of special interest: On July 28th, soon after midnight, we left a bivouac at 14,000 ft. (4,300 m.) on the north- eastern rampart of the eastern glacier, and out- flanked that obstacle by traversing in the dark, over dangerous ground, the south-west flank of a lofty snow-clad summit. We then chmbed the next peak, about 16,500 ft. (5,000 m.), and descended several hundred metres to a snow-saddle, and again worked our way up to a similar dome- shaped ridge about 15,800 ft. (4,800 m.) high. Thence we descended towards the east, and thus reached the head of a hitherto unknown valley, quite filled with glacier ice. The course of this valley was first north-east, then east, and finally south-east, debouching in the neighbourhood of the Musart pass, and having thus a length of about twenty-six miles (forty versts). From the quite level ice-floor of the valley-head we turned towards the south-west, ascended about 1,300 ft. (400 m.) THE "MARBLE WALL" 33 over nevd-covered slopes and, passing over a wide snow-clad ridge, reached the foot of the north- west arete of the " Marble Wall." There a magni- ficent view opened out, towards the west over the wild glacier region of the Bayumkol, and towards the east over the far-stretching ice-field of the newly discovered valley. This is bounded on its south side by a commanding ice-clad range of splendid peaks, stretching away towards the Musart pass. Tn the deep bays between these peaks lie exceedingly rugged and picturesque many-terraced glaciers, which descend steeply to the main glacier. This range, branching off from the " Marble Wall," without doubt forms the main watershed between the northern and southern slopes of the Central Tian-Shan, as was proved to a certainty by subsequent observations from various points of view. I estimate the average height of the ridge of this chain at about 16,400 ft. (5,000 m.), and that of the peaks at more than 19,500 ft. (6,000 m.). There is only one deep depression in this mountain rampart. My expectation of seeing Khan-Tengri, towering up in it, was disappointed, and the question as to its position became ever more mysterious. It could not be far off, but in which of the valleys, lying behind this range, could it rise ? Once more the inaccuracy of all the maps of this region was proved. There, where, according to the maps, Khan-Tengri should be, rises the *' Marble Wall." The northern rampart of the ice-valley, though not so lofty as the southern, 3 34 THE VALLEY OF BAYUMKOL is imposing enough ; through the indentations of its crest we could see an ocean of peaks, many of which had never before been looked on by human eye. They belong, in part, to the mountain range, bounding the unexplored valleys to the north-east and east of our position, some of which, at least, I was able to traverse in the following year. Owing to the nearly complete covering of snow and ice on these lofty ranges, one could see but little or nothing of their geological structure. That diabase must be represented in them was shown by blocks among the meagre debris at the head of the valley. In the following year I was able to determine their composition, which is identical Avith tliat of the range at the head of the Bayumkol valley. In viewing these mighty pro- tuberances of the ground, rising round us, one could not but perceive that the broad masses of the mountain ranges east and west of my position are cleft only by the courses of a few deep valleys, evidently of very old formation, and are thus di\ided into single groups {massifs), whose roofs are in most instances furrowed only by elevated troughs or not very deep channels, separating narrow crests and numerous peaks which rise out of the plateaux. The mouths of these smaller elevated valleys, retaining snow and tiny glaciers, almost always lie very high above the level of the main valley. Without discussing this interesting subject, I sliall only mention that at the time when the channels of the main valleys were still BAD WEATHER 35 filled high with ice, the small contributory glaciers in the upper valleys disembogued at the level of the surface of the glaciers in the main valleys. As the glaciers, both below and above, retreated (the tributaries far faster than the main glaciers), erosion by the action of flowing water was, in consequence of the rapidly increasing dryness of the climate, insufficient to contribute materially to the development of these newer valleys, while, on the other hand, in consequence of the enhanced destruction of the mountain crests, the filling up of the hollows with debris began and continued till these were again partially filled with snow and ice, o^ving to a renewed but less copious glacial period. In the con- figuration of the roof of these ranges, accordingly, we see the result of erosion and excavation no longer in vigorous action, while, in all the deep channels, especially during the interglacial periods, both continued and still continue to act very powerfully. The continuously unfavourable condition of the weather in the Bayumkol valley caused me, though my labours were not yet ended, to leave it for the time and not return till autumn, when, with less conflict between the thermal conditions of plain and mountain, more settled weather might be expected. I wished to try whether better weather would not fa\'our exploration in one of the larger valleys, the Sary-jass valley. CHAPTER III THE SARY-JASS VALLEY AND THE SEMENOFF GLACIER We withdrew about sixteen miles (twenty-five versts) from the head of the (Bayumkol) valley, and then turned southwards into the Ashu-tyr side valley, already mentioned, which possesses great wealth of water, alpine meadow, and pine forests. The valley has an approximate length of sixteen miles (twenty-five versts), and, ascending steeply in three stages, stretches with many turnings, but on the whole in a south-south-westerly direction, following the strike of the gneiss ; this often passing into granite, and alternating w^ith limestones, phyl- lites, metamorphosed schists and especially marble schists forms the boundaries of the valley which especially in its lower course, are ruggedly peaked. INIarbles and marble schists show, particularly at the head of tlie valley, great disturbance and extraordin- ary clca\'age, due to fractures. The valley every- where displays traces of its former ice-covering, not only in the deposits of drift, but also in the grinding and rounding of the lateral cliffs, noticeable hi a high degree in the upper portion of the valley. Its supply of glacier-ice is no longer great, yet some of the many side-valleys, which open into it, 36 KARAKOL 37 contain small glaciers, while two of them have glaciers of more considerable size, which, however, are also in a period of rapid retreat. Everywhere snow and ice (to be seen especially on the slopes of some rugged, massive peaks) are limited to northern and eastern exposure. At the head of the valley we made a very steep ascent o\^er marshy meadow-land (water being everywhere in the valley surprisingly abundant), an old gi-ound moraine, covering the mountain slope and reached a glacier, the crossing of which was very difficult for the horses on account of its covering of soft snow, and owing to crevasses, which the snow concealed. Crossing the snowy ridge about 12,800 ft. (3,900 m.) high, we reached the Karakol valley, which opens into that of Sary-jass. I must here remark that the Kirghiz know no other name for this side-valley than " Karakol." This, after many inquiries, I was able to establish just as surely as that nowhere in the Tekes valley do the Kirghiz population or the Kossacks of Narynkol or the constituted authorities for the Bayumkol valley, apply to it the name " Karakol." Hence Herr IgnatiefF is wrong, I think, in re- naming the real Karakol valley after his Kirghiz guide Bektur-bulak. Geographical names cannot be dealt with too carefully if one would avoid confusion. Dr. Friedrichsen, wlio, with the SaposhnikofF expedition, crossed by the same route as we did, but two weeks earlier and in the opposite direction, in his " Reisebriefen " calls the valley Ashu-tyi', a name which belongs 38 THE SARY-JASS VALLEY only to the valley, running from the pass north- wards to the Bayumkol valley. Dr. Friedrichsen took this tributary for the main valley, though the far greater volume of the main river is alone sufficient proof that the main valley must lead at its southern end to extensive glaciers. The pass itself he calls Narynkol pass, evidently on the probably erroneous assumption that it is the identical pass, crossed by IgnatiefF and by him called Narynkol. If the pass must bear a name, Ashu-tyr would be a more suitable one. We made a steep descent from the pass in a south-westerly direction following the course of the Karakol glacier, which comes from dome-shaped, snow-clad summits, and is very little laden with drift. Our path lay in a hollow between the margin of the glacier-tongue, a wall of ice 100 ft. (30 m.) high, and the mountain side. The range here consists of phyllitic schists, stratified porphyry, granite, hme- stones, and extraordinarily riven marbles, as well as conglomerates and breccias, which are con- nected with the outcropping of the porphyry. In the limestones Herr Keidel found badly preserved fossils. The rock walls on both sides of the valley are polished to a great height by ice, and the lower part of the valley may be regarded as a type of a valley to a considerable extent shaped by, if not entirely due to, the action of ice. Besides the main glacier, which, after a course of about three miles (four to five versts), ends abruptly in its own debris about 3,700 m., there are two other considerable glaciers, which come in from the left, but their GREAT GLACIERS 39 tongues cling to the black slate walls, no longer reaching the main glacier ; in a similar condition are a number of smaller ones, which lie in holes and corners of the rock sides of the valley. The lower course of the valley, much widened in consequence of faults (one is especially finely exposed), has been eroded into the form of a kettle, owing not only to the action of the main glacier, but also to the convergent action of the numerous secondary glaciers, which formerly forced their way beyond their actual boundaries concentrically into this valley, thus presenting a true object-lesson on the corrosive action of ice. Here also, in conse- quence of ruptures as well as of the grinding force of the ice and the weathering, which is peculiarly active, owing to the valley's being open to the south and west, an illustration of advanced destruction of the mountain range is presented, such as I had seldom seen in the Tian-Shan, rich as it is in phenomena of this sort. The southern and western exposure, favourable to an extraordinary insolation of the dark slate cliffs, and the consequent intensity of reflected heat, are the cause of a more conspicuous retreat of the main and the secondary glaciers than I observed in any other valley of similar altitude in the northern Tian-Shan. The main glacier at one time de- bouched seven miles (ten versts) below its present termination to join the giant glacier which formerly filled the Sary-jass valley. On a green terrace of old morainic drift, near the spot where the Karakol stream now flows into the Sary-jass river, I had 40 THE SARY-JASS VALLEY my headquarters at about 11,500 ft. (3,500 m.), whence excursions were made for the exploration of the SemenofF glacier and the ranges, sur- rounding it. According to the publications of certain travellers, who \isited the Sary-jass valley and penetrated some versts upwards on the ice of the SemenofF glacier, this ice-stream would appear to be fed from the snow-fields of Khan-Tengri. If this were the case, the mountain would be situated in the background of the ice- valley ; but the course of the valley winds about, and even from elevated positions its back- gi'ound cannot be recognised with sufficient cer- tainty, and the less so that broad side-valleys, which themselves have branches, debouch near the valley-head. From many points on the margin of the Sary-jass valley Khan-Tengri is seen, how- ever, always in such position, that one is constrained to believe it can only rise at the head of the Semenoff glacier. Yet, since in the Bayumkol valley I liad settled how far the SemenofF glacier stretches northwards ; I was doubtful of this assumption. Taking advantage of favourable weather we forth- with ascended a mountain, towering immediately behind our camping-ground, on the north side of the valley. From this snow-crowned level, about 13,800 ft. (4,200 m.), there is an excellent view over the glacier ranges of the Central Tian-Shan. The favourable position of the point reached, the pellucid air, and the extraordinarily clear light made it possible to take a telephotographic pano- rama in twelve sheets of 8 by 10 in. which THE "MARBLE WALL" IDENTIFIED 41 will be of great value for the determination of positions as well as of the form and course of the mountain chains, which constitute the loftiest portion of the Tian-Shan. The view over these commanding ranges showed that Khan-Tengri has no rivals of even approximately equal nobility. Though many summits may reach a height of over 20,000 ft. (6,000 m.), and a few even 1,500 ft. (400 m.) higher, the slender pyramid of Khan-Tengri still overtops and dominates thein all. In these hurried notes all I can say concerning the rela- tive elevation of the Central Tian-Shan is, that the greatest altitudes are in the mountains surrounding the Bayumkol valley, especially those between it and the Semenoff glacier, though these may be surpassed by a few of the noble peaks to the south of the Adyr-tyr or INIushketofF glacier, but that these are all excelled by the mountains on the southern boundary of the Inylchek glacier, and that at all events the average crest and peak altitude of this range must be regarded as the highest in the Tian-Shan, a gradual slope towards the south beginning here. From our standpoint we could establish with certainty that the " JMarble Wall " is identical with the summit which on all the maps is marked as Khan-Tengri {vide p. 33,) and, though its whole importance as a centre of ramifica- tion was not completely proved till later, one could even now see that in its neighbourhood a parting of divergent ranges takes place. The grouping of the crests round the topmost pyramid of Khan- Tengri, however, as seen from this point, was such 42 THE SARY-JASS VALLEY that one could not say, even with remote confi- dence, from which of the valleys it rises, especially as in its neighbourhood, about north-east from it, there is a seemingly confused crowding together of mountain ranges, approaching from different directions. It could be conjectured, but not settled, that the base of Khan-Tengri, the lord of the Tian-Shan, lay in the Inylchek valley. Some hundred metres below the top of our plateau there extends, like a shoulder of the mountain, a terrace on which Herr Pfann marked out a base line and fixed its position by astronomical observations. From this he determined the position and height of Khan-Tengri and others of the most prominent points of the Central Tian-Shan, while I set about the exploration of the SemenofF glacier and its boundary ranges, and Herr Keidel busied himself with the investigation of the geological structure of the range, surrounding the Sary-jass valley below the camp, for which purpose he made excursions into the side-valleys on the right bank. Where extensive faults occur he found schists, phyllites, limestones, granites and diabases, which had fallen in small flakes to various levels. In the Kashka-su valley he was fortunate in finding I)e^'onian limestone. The geological stratification and composition of the valley ranges shows similarity to those of the Bayumkol valley, but in tlie Sary-jass valley diabases are more widely difliised tlian in the Bayumkol. More detailed notes are reserved for the special geological report. The Sary-jass valley is the most extensive and THE SEMENOFF GLACIER 43 the most important of all the valleys of the Central Tian-Shan, since it forms the great channel for the waters, flowing south to the Tarim. Its present configuration without doubt has been affected by a glacial period. The merit of having first pointed out the significance of the glacial deposits which are found in the valley, belongs to P. P. Semenoff; the wide diffusion of these deposits is, however, even more important, as this famous explorer has him- self acknowledged. I could observe these and other signs of the action of ice in the main valley and its secondary valleys up to 1,650 ft. (500 m.) above the present level of the river, to such heights, that one might infer that formerly the valley was almost completely filled with glacier ice. In comparison with the thickness of former glaciers, that of the beds of nev^ and ice, still found in the main valley and its tributaries, is insignificant ; nevertheless, these form one of the largest glacier regions in the whole of the Tian-Shan, and are, as will be proved by the results of my exploration, in any case much more important than has hitherto been believed. The largest glacier of the region is the Semenoff glacier, which hitherto was supposed to be the largest of the Tian-Shan. I had the good fortune in the course of the expedition to obtain proof that it is exceeded in length by other ice-streams, one being more than double its length. But also, the extent of the Semenoff glacier has been hitherto under-estimated. According to Ignatieff, who visited it in 1886, its 44 THE SARY-JASS VALLEY length is about seven miles (ten versts) whereas it is actually about three times that length. Of its breadth and the breadth of its tributaiy glaciers there was on the whole up till now no proper idea. From various causes, partly also as a result of the westerly direction of the axis of the upper course of the Sary-jass, we observe the rare phenomenon that the main glacier has retreated to a greater extent than the still existing side glaciers, which — at least, those debouching into the upper course of the valley — have almost kept their early horizontal length, if not their former thick- ness. This, however, is true only of those which debouch on the (orographic) left bank, since their axis is directed towards the north. Their tongues at the debouchement hang like patches of ice over ground-moraine debris 600—900 ft. (200—300 m.) above the present floor of the main valley, where tliis is free from ice. Of those which now terminate in tlie region of the main glacier as it at present exists, the tongues of the first three no longer reach it, but hang 300—500 ft. (100—150 m.) above the level of its surface. All the secondary glaciers farther east debouching on the main glacier, some of them of great extent, unite with the main stream, and the level of their floor as a whole lies in one plain with that of the latter. The unusually gentle slope of all these ice-streams (it is only 25 m. per verst or 126 ft. per mile in the middle and upper course of the main glacier) miglit, in my opinion, point to a considerable SIDE-VALLEYS 45 filling-up of the channels of the valley with mountain debris at some time when they were not yet covered with ice. The side-valleys, debouching on the right bank — at least, those debouching into the portion of the main valley, now free from ice — have, where their axis is directed towards the south, no longer any glacier ; only at the head of some of them small fields of neve he in recesses (Kare). The mouths of these side-valleys lie from 650 ft. to 1,000 ft. (200—300 m.) above the bottom of the main valley ; to reach them one has to climb up steep, green, marshy ground- moraines. While the chain on the left bank is cut into by numerous side-valleys, whose own boundary walls, also deeply indented, seem broken up into many steep and variously shaped peaks, the chain on the right bank is divided by relatively few side -valleys, and the walls, bound- ing these show far fewer broken crest-lines and more plateau-like tops, shattered crests with super- imposed tent-shaped peaks. The forces at present at work in the formation of mountains do not account for these facts, which rather point to the conclusion that, prior to the beginning of the actual ice-covering of the mountain range, erosion was more powerfully at work on the slope ex- posed to the north, and destruction (deflation) more on that exposed to the south, and con- sequently that the climatic conditions were then similar to those now prevailing, though they may have been less sharply accentuated. At the 46 THE SARY-JASS VALLEY same time, the high angle at which the strata composing these ranges are set must be taken into accomit. Several miles below the tongue of the SemenofF glacier the bottom of the valley is hollowed out to a basin about a mile and a half (two versts) wide, its level floor covered with shingle. At some former time the water from the glacier had been dammed up here by terminal moraines, so as to form a lake ; the basin still holds some small relics of lakes. The streams from the glacier are still busy, cutting up and carrying away the remains of the old moraine ice-drift, which still exists in considerable masses. To the climatic difference between the two banks is to be traced the fact that the tongue of the glacier runs along the south bank for more than a verst, where the north bank is free from ice. I subsequently observed like phenomena in other Tian-Shan glaciers with similar exposure. The tongue of the SemenofF glacier ends at about 12,000 ft. (3, GOO m.) altitude, according to observations taken in two consecutive years. The climatic difference is shown also in another way, the mountain chain along the bank facing south- wards having snow and ice only on its summits, which are but slightly divided, while its deep rocky precipices retain these only in ravines and cliannels ; whereas the chain looking north is wrapped in a bright garment of snow and ice, which seldom shows a rent. This range, much divided, stretches eastwards as a row of mighty GLACIAL PHENOMENA 47 snow-clad, round-headed mountains, horn-shaped peaks, and steep ice-ramparts, presenting a magni- ficent spectacle. In the middle and upper course of the glacier, where its axis is directed more to the north-east, the bounding range on the right bank also appears to a large extent covered with ice, though neither in this respect nor in imposing mountain shapes does it equal the chain on the left bank, which, besides, is loftier. This, and also the circumstance that the glacier bottom slopes towards the north, are to be traced to the gradual rise of the collective mountain mass towards the south. In consequence of the slope of the ice-bed towards the north, the water from the ice tends to flow to the north bank, and the main stream therefore springs, not from the end of the ice-tongue, but some versts above, from a cavity in the precipitous right side of the tongue. I was able to observe similar phenomena, due to similar causes, in the other great glaciers successively towards the south. The glacier has, near its tongue, a breadth of about 5,000 ft. (a verst and a half), but widens farther up and attains in its middle course a breadth of more than two miles (three versts). In its lower course its surface is free from snow, but is covered with some debris, though to a less extent than other great glaciers of the Tian-Shan ; here it is only furrowed by some deep troughs, owing to peculiar conditions of insolation, dependent on the form and exposure of the mountain walls, and also due to erosion by melting water ; elsewhere it is uneven, 48 THE SARY-JASS VALLEY in places wa\y, but not penetrated by crevices to any gTcat extent. Generally the cleavage of the surface is comparatively slight, partly in con- sequence of the gentle slope and the evenness of its base, which I have already mentioned ; partly on account of the absence of lateral pressure, since, apart from the immense size of the basin, the ice on both sides is separated from the rocky banks by deep chasms ; and lastly because, as already stated, most of the secondary glaciers join the main glacier v^^ithout any descent. The principal regions of cleavage are at the arched edges, mostly on the right side. There are only a few places at which seracs have been formed. In consequence of its great extent and its gentle slope the Semenoff glacier is fairly constant. I have visited it in two consecutive summers, have roamed over it in all directions, and altogether have spent more than two weeks on its surface, but neither at its terminal tongue nor at its edges could I find any indication of shrinkage in recent times. If a traveller, who paid a flying visit to the lowest part of the glacier some years ago, reported on its rapid and continual melting away, he was probably induced to take this view by the many rivulets, gushing over the ice, such as are formed in the case of every great glacier (even in the European Alps, though perhaps in less degree) in the midday hours of hot summer days. But for the melting away which takes place in the course of a brief Tian-Shan summer under the climatic conditions now prevalent, ample com- CHARACTERISTICS OF THE GLACIER 49 pensation is made by the extraordinary amount of snow and ice which the Semenoff glacier receives, especially from the very large secondary valleys of its upper course. So long as there exist such immense stores of snow as I have seen in the vast, hitherto untrodden interior of the Central Tian-Shan and so long as their masses, impelled by their own weight to lower altitudes, continue to deliver abundant material for the formation of neve and glaciers, there is, in my opinion, no danger of a complete drying up of the Tian-Shan, such as has been frequently mooted. These vast stores of snow, not only on account of the dry condition, peculiar to snow at great altitudes in the Tian- Shan {vide pp. 30, 31 ; more concerning this here- after), but also on account of the low temperature of the air at these elevations, undergo ^ ery trifling diminution through melting or sublimation, but on the contrary are increased by new falls of snow. On this interesting subject and on the phenomena connected with it, I must not further enlarge within the limits of this report. Of all the great glaciers of the Central Tian- Shan which I have visited, the SemenofF glacier, on the whole, shows in its general habit most resemblance to the great glaciers of the European Alps. Only in one point it is essentially different from them : With respect to its great wealth in ice-lakes, and with respect to their origin and disappearance, I shall state my opinion in a more detailed report. Most of them are funnel-shaped, and they are rather irregularly arranged on both 4 50 THE SARY-JASS VALLEY banks of the lower and middle course, but are more numerous on the right bank. Many of them have considerable length, 640—1,000 ft. (200— 300 m.), and present a magnificent spectacle when the ice-capped giants surrounding the glacier valley are mirrored in their green or blue waters. The difference in their colour, some being green, some blue, is a highly peculiar phenomenon. In the upper course of the glacier there are no ice- lakes, but in the moraines on the right bank there are numerous moraine-lakes, not inconsiderable in size. The snow-covering begins in the middle course, and is very thick in the upper course. A neve basin, resembling a lake, a mile broad, oval, trough-shaped, and rising in two stages, but elsewhere with only a very slight incline, forms the north-eastern and highest portion of the glacier, which is enclosed by the southern wall of the western Bayumkol glacier. In this range, in which some magnificent snow-peaks rise to a height of more than 20,000 ft. (6,000 m.) there is a deep depression, easily accessible from the uppermost snow trough, and to this, since it lies at the very head of the Semenoff glacier, I have given the name of the SemenofF pass. With a favourable condition of the snow-coA'ering of the Bayumkol glacier, one would probably be able to descend through this opening into the last-named valley. Tlie total length of the SemenofF glacier from its terminal tongue to this pass is about twenty miles (thirty versts). The masses of mountain debris, transported by THE MORAINES 51 the glacier are comparatively small ; the lateral moraines have become bank moraines ; the middle moraines (there are only two) receive but little material, since the great side-valleys, one of which, with an average breadth of 3,500 ft. (one verst), has an approximate length of seven miles (ten versts), are bounded by magnificent mountain chains, whose splendid snow and ice mantles show, however, but few rents. In the lower reaches of the lateral moraines, granites and limestones predominate and elsewhere chloritic schists and clay-slate, though limestone as a rule is found only in the left moraine, since the lime- stone crops out from a bed on the left bank, which runs north-east and does not again reach the right bank. The middle moraines consist almost entirely of granites of varied structure, and of granite porphyi'ies, pegmatite, and syenite, with some clay- slate ; but these rock fragments become scarcer the nearer one approaches to the upper course of the glacier. Here we meet with increasingly metamor- phosed limestones, slates, and white marbles, and also fragments of diabase and diabasic slates. This leads to the conclusion that the innermost boundary range consists only of this rock series. All accurate insight into stratigraphic relations is prevented by the thick snow mantle, covering the bordering ranges. On the right bank, where snow-free slopes occur here and there, the ground is covered with a chaos of blocks. Unfortunately, in Sary-jass also our work was little favoured by the weather, although it was not quite SQ unsettled as in the Bayumkol 52 THE SARY-JASS VALLEY valley. From a bivouac on the right margin of the moraine, about 12,800 ft. (3,900 m.), some ten miles up the glacier, Herr Pfann and I with the Tyrolese Kostner ascended a snowy pyramidal peak, rising to about 15,750 ft. (4,800 m.). From its summit was unfolded to our view in all its imposing magnificence the vast icefield with its ranges of peaks completely snow- and ice-clad, beyond which were visible the still loftier wonder- ful mountains of the MushketofF and of the Inylchek glaciers — altogether an Alpine prospect, such as is to be seen in few other parts of the globe. The pyramidal cone of Khan-Tengri was visible away to the south-south-east beyond a broad snowy summit, surrounded by several inter- crossing ridges, so that it seemed already evident enough that Khan-Tengri has no connection with the SemenofF glacier, although in the absence of any trustworthy topographic information it was impossible to say from what valley it rises. The northward view from our summit was specially instructive as to the conformation of the complex mountain system, stretching between the Bayum- kol, Karakol, and Kapkak valleys, as well as the trend of the upland valleys, ramifying through it. This was a welcome addition to the observations, which we had already made from the heights of the Bayumkol valley. There was just time to take a number of photographic pictures of the whole scene when a sudden snowstorm put an end to our observations. Being determined to solve the riddle of the V ,t "^ -'f ^ NV^ 1 TEI.I rlKlTdi.KAl'llli \li,\\ 1,1 liU. .SUMMIT OK K II A N -TE.Nl . RI I A 1;. )r r 2_;,6> o IT. I, lAKEN FROM NORTH-NORTH-WEST, IROM A PEAK (ABOUT l6,0O3 IT.) ON THE SOLTHEKN MARGIN OF THE SEMENOFF GLACIER. DISTANCE ABOUT 5 MILES. [To face p. 52, THE SEMENOFF PEAK 53 position of Khan-Tengri, we proposed the very- next day to make the ascent of the highest point at the head of the SemenofF glacier. This is a superb, broad, snowy peak, which is encircled by a system of wild seracs and crevasses, and rises some seven- teen miles above the tongue-end in the north-eastern extremity of the glacier, its highest point exceeding 19,685 ft. (6,000 m.) by a few hundred metres. This wonderful mountain, which dominates the whole basin of the SemenofF glacier, I have named the " Semenoff peak," in perpetual memory of the great services rendered by the energetic President of the Imperial Russian Geographical Society to our knowledge of the Tian-Shan. We started from our elevated station soon after midnight. AVith some difficulty we approached the right margin, threading our way in the dark through the system of lateral crevasses which, owing to the sharp bend of the valley to the north-east, is here very intricate. Unluckily I stepped into a crevasse, thereby so severely sprain- ing my left foot that, although still able to get on that day with some trouble, I was afterwards fain to reserve my strength and prevented for some time from taking part in trying excursions. After a rapid march of some eight miles over hard-frozen snow we reached the foot of the last stage of the glacier, from which access is gained to the highest nevefield, which is still some three miles distant from the " SemenofF pass." the extreme point. From this stage of neve we ascended across steep, much-fissured snowy slopes in an approximately easterly direction 54 THE SARY-JASS VALLEY and made good progress over the snow, which, thanks, to the early hour (5 a.m.)» was in a favourable con- dition. We soon reached a considerable altitude, so that we felt a confident hope of scaling the top of the giant mountain, and from it at last acquir- ing some certainty, regarding the position of Khan- Tengi'i and the ramification of the loftiest crests. This hope urged us rapidly forward ; but as we mounted higher and higher, up to an elevation of some 16,400 ft. (5,000 m.), the hard-frozen surface gradually gave way under our feet and, hence- forth, was formed of snow, which assumed more and more a powdery consistency. I have already indicated one of the causes of this phenomenon {vide pp. 30, 49). The moisture precipitated as snow on the extreme heights of the Tian-Shan possesses a peculiar crystalline form, and is dry as powder. The atmospheric strata of these altitudes are un- usually dry, but on snow of this nature they cause no appreciable amount of evaporation. Moreover, even under the influence of insolation the surface layers cannot, owing to the constant circulation of the upper atmospheric strata and their low tempera- ture,thaw during the day, and consequently cannot form a frozen crust at niglit. If anywhere, such processes take place on the slopes, facing south and west, though even there only to a small extent, but as a rule liardly at all on those with a northerly or easterly exposure. There, on the contrary, the severe night frosts only make the snow all the drier. This prevents any congelation, and one sinks a yard deep in tiie powdery snow. But when this powdery A TOILSOME ASCENT 55 snow lies on a layer of old snow, which by the above-mentioned processes has acquired a glacial surface in any places, favourable for such a forma- tion, or else has been gradually hardened by the pressure of the overlying layers, then there is great danger of the loose surface layer, when trodden upon, breaking away from the steep slope and glid- ing with the persons trespassing on it down to the bottom. In a few days this was, in fact, verified. For us, however, there Avas no imminent analogous danger during the ascent ; but we sank at every step to the middle, and could no longer find any firm footing. All oiu' efforts failed to discover a zone of snow in better condition by changing the direction of our ascent. In order to divide fairly the toilsome labour of treading down the snow, we changed the leader every ten minutes. Still the strength of the three Alpine climbers gradually flagged, and despite the most heroic efforts, we no longer made any appreciable progress. There were still to be sur- mounted over 3,300 ft. (1,000 m.) of absolute elcA'ation, which, taking into account the angle of the slope and the winding of the way, was equivalent to a distanceof over 5,000 ft. (1,500 m.). Even if our strength was equal to the task, which was not to be thought of in such snow and in such extremely rarefied atmosphere higher up, night would have overtaken us on the summit. And how easily might the weather have clianged, so that even then we should have no longer been able to make observations. 56 THE SARY-JASS VALLEY The undertaking had to be given up as hopeless. Still, it was not quite useless, since the prospect ti-oni the elevation attained gave us much supple- mentary information. The state of my foot obliging me to return to headquarters, my place was taken by Herr Keidel, and one of the Narynkol Kossacks was summoned to carry the large photographic apparatus. The party now ascended a summit 15,100 ft. (4,600 m.) high on the southern margin of the SemenofF glacier, the primary object being to take photo- graphs and to obtain bearings for our further investigations. From this point Herr Pfann's next goal was a mountain, rising on the southern margin of the neighbouring JNIushketofF glacier, which runs parallel with the SemenofF. The assumption — verified a year later — was that from the top a decisive view must be had of the valley, towering above which, the pyramidal crest of Khan-Tengri is always seen. As the mountain at least its flank up which the ascent was to be made, is not very steep, and moreover faces west, all the conditions were present for a successful issue. At midnight the party of four started from a camp 18,450 ft. (4,100 m.) high, situated on the left margin of the SemenofF glacier at the junction of a broad level secondary glacier valley. They traversed its course of about fi\'e miles, and thus before daybreak reached tlie foot of a broad low ridge, which is crowned with stunted snowy domes and separates the SemenofF basin from the upper A NARROW ESCAPE 57 part of the INIushketofF glacier, of which more anon. A deep depression about 14,450 ft. (4,400 m.) in this valley — which I name the " JNIushketofF pass " in honour of the never-to-be-forgotten naturalist, MushketofF — was now surmounted. As the surface of the MushketofF glacier is here about 500 ft. (150 m.) higher than that of the Semenoff glacier, but little height was sacrificed in descending to reach it. It was crossed from side to side at a point, where it is about two miles wide, so that by day- break the party reached the foot of the completely snow-clad mountain on the opposite margin, which was to be ascended, and had an altitude roughly estimated at 17,700 ft. (5,300 m.). Scaling over a snowy ridge, trending away to the west, they reached the shoulder of the mountain and began the ascent on the west flank of the actual summit itself All went well, and the snow remained firm under the feet of the climbers, who w^ere connected together by a stout Alpine rope. Towards 11 o'clock in the forenoon they found themselves within 300 — 600 ft. (100 — 200 m.) of the very top of the mountain. Then was heard a sudden crash : a surface layer of snow, loosely overlying a substratum of hard snow had cracked ; it gave way and slid towards the bottom with all four climbers. They all seemed lost, when their downward course was fortunately arrested by a small snowy ledge projecting some 650 ft. (200 m.) deeper out of the slope. All four were able to work their way uninjured out of the sno\\y masses, and nothing had to be regretted, save the loss of some hats and ice-axes, which could not 58 THE SARY-JASS VALLEY be recovered. The Kossaek, paralysed by fright, completely lost his senses. The three others were inconsolable at the failure of the attempt, which in Herr Pfann's opinion must have led to the discovery of the position of Khan-Tengri ; yet another year elapsed, before he was found to be right. So near to the wished-for goal had their hopes been wrecked. And now for me the outcome of all past ex- periences was that in the highest regions of the Tian-Shan it was perhaps only under quite ex- ceptionally favourable conditions that the snow can acquire that consistency which permits the ascent of peaks, rising above 16,404 ft. (5,000 m.), unless indeed it can be made on rocky ground. Only the lofty rocky crests are for the most part ex- tremely precipitous, and, as appeared from further experiences, owing to the influence of excessive thermal contrasts, so profoundly shattered that the attempt to scale them often encounters un- surmountable obstacles. Ascents through rocky gorges and couloirs have to be avoided on account of the great risks, incurred from falling stones. Hence only a very few of the loftiest Tian-Shan summits hold out favourable prospects to the Alpine climber. liearing this in mind during the subsequent course of the expedition, I avoided difficult Alpine undertakings, and henceforth ascended only such mountains as might from their position offer the promise of connnanding prospects, affording an insiglit into the structure of the mountain range, SNOW-STORMS 59 and at the same time seemed accessible to ex- perienced Alpinists without exposure to great danger. INIeanwhile a period of unfavourable weather had set in, and all operations were prevented by daily snow-storms. This obliged me to leave the Semenoff glacier, the accurate measurement of which by triangulation was not carried out till the following year. As we had made sure that Khan-Tengri does not lie too within the basin of the IMushketoff glacier, I decided at once to pene- trate into the next great parallel valley, that of the Inylchek, and there look for it. CHAPTER IV TO THE INYLCHEK GLACIER AND FARTHER SOUTH We moved some twenty-three miles down the Sary- jass valley, which soon loses its picturesque aspect. Omng to the already mentioned causes, the chain, skirting the right bank, shows rounded crests, pierced only by a few upland gorges, but no glaciers. The left bank still maintains for a short distance its high Alpine character. It is dis- posed in separate sections by deep transverse valleys, harbouring glaciers. These glaciers breaking out from the ravines, combined with the glittering snows of the peaks enclosing them, form a lovely contrast to the deep green of the main valley and its slopes, carpeted with Alpine meadows. The most important of these tributaries is the Adyr-tyr valley, which above its mouth, turning in a swift course to the east, flows approximately parallel to the SemenofF valley, wliich it nearly rivals in length, breadth, and wealth of glaciers, and even surpasses in the height and grandeur of its mountains. Its upper course is filled by a glacier, which IgnatiefF has named the MushketofF glacier, of which more i'arthcr on. 60 GREEN PRAIRIES Gl The wide green prairies of the Sary-jass, averag- ing one mile, but in places broadening out to two miles, showing the character of the treeless and scrubless upland steppe, display soft rounded forms, which are due to the old morainic deposits, fringing the scarps of the valley. Such lateral moraines on the left side, well preserved in two stages, accompany at intervals the upper course of the valley. On the right side, even on the plateau- like ridges are morainic deposits and debris now to be seen, frequently also glacier scorings liigh up on the rocky walls. The valley-bed is filled with old ground-moraines, covered by marshy meadows with small tarns, relics of the large lakes, which, being dammed up by terminal moraines, formerly filled the basin-shaped expanses. The origin of some of these expanses is obviously due to the lateral erosion of the river. Another, above the Adyr-tyr valley, has been caused by a kind of cleavage {Scharung), the ridges receding from each other in consequence of some sudden change in the strike of the strata. The phenomenon must in some way be connected with the already-men- tioned {vide p. 42) faults and fractures in the lateral valleys. At the mouth of the Adyr-tyr valley, over one verst broad, the granite and the phyllitic rocks associated with it, disappear below the surface ; the limestones of the chain on the left bank of the Adyr-tyr valley strike outwards, and form farther on in the Sary-jass the southern ramparts, whose ridges hence rapidly diminish in height. Beyond them the superb glaciated highlands of 62 THE INYLCHEK GLACIER the Kulu-Tau come into view, with an extremely bold eminence, towering up as a solitary peak. The slates and marbles, present on the right bank in flakes are absent on the left bank. From a broad gap in the limestone range on the left bank some seven miles below the mouth of the Adyr-tyr, the copious Tys-ashu river flows to the Sary-jass, and drains a much-ramified valley region, not shown on the maps. This group of valleys lies in a tangle of mountains with a north-westerly slope, and is enclosed between the high range, forming the left bank of the Adyr- tyr valley, trending north-westwards, and the chain, stretching south-westwards along the right bank of the Inylchek valley. In the obtuse angle, formed by the two widely diverging chains, lies an extensive plateau-like neve, which is gently inclined, and in both chains develops blunt, tent-shaped, snowy summits. From the breaks be- tween these eminences, sweeping round in a wide curve descend flat, trough- shaped ravines, filled with nev^, and disposing in radiating sections the broad stretches of land, which slope quite gently down to the Sary-jass valley. By a lofty plateau ridge, " Tur," which has escaped all erosive action, the whole system of valleys is disposed in two groups — that of the Kusgun-ya valleys, which will be dis- cussed farther on, and that of the Tys-ashu valleys. Kongul-jol, Achik-tash, Mai-bulak, Tys-ashu I. and II. are the names of the more important radiating head-channels, which unite in a main stream also called Tys-ashu (Tys-ashu means the A SUPERB MORAINIC LANDSCAPE 63 ramifying of a level tract). The tracts of nev^ which lie in the broad, shallow upland troughs of these valleys are at present separated one from the other by considerable masses of morainic refuse, dis- posed in ridges. Two only still show conspicuous stretches of glaciers, which, however, soon come to an end amongst the debris of their ground-moraines. From the whole aspect of the land it is seen at once that all that now remains of isolated nev^ is but the remnant of a once continuous and very extensive ice-cap. From these glacial masses was developed a huge glacier, which formerly spread over the lower parts of the district, and joined the at-one-time-mighty Sary-jass glacier. The whole of the wide Tys-ashu domain, which is amongst the favourite grazing-grounds of the Kirghiz, presents a superb morainic landscape of a typical character, such as is elsewhere rarely to be seen. The rocky Avails, too, are polished to a great height by glacial action. Standing later on an elevated position, I was able to ascertain that the great glacier, to which it owed its existence, was formed by the combined glacial masses of the southern border-range of the MushketofF glacier and the chain, skirting the north side of the Inylchek glacier. In the trough-shaped depression of the Tys-ashu district the hills are literally buried beneath morainic drift, now covered with swampy Alpine meadows, so that only in a few places the rocks are seen cropping out — limestone disposed in narrow folds, stretching northwards, granite, and phyllitic schists. 64 THE INYLCHEK GLACIER Just here the cham on the north side of the Inylchek valley is greatly depressed, while the southern range under the same meridian is up- heaved to one of its highest elevations, one of the most imposing mountains in the whole Tian-Shan system. Hence to a person, standing in the lower part of the Tys-ashu valley, and looking upwards along the broad, gently rising, trough-like valley, the deceptive impression is conveyed that the ex- tensive neve at the head of the valley runs straight up to the wild and precipitous glacial walls of the huge Inylchek peak, which seems to close the Tys-ashu valley. What lies between remains hidden from the eye of the observer. Evidently this impression misled Professor KrassnofF himself when, in bad weather too, he penetrated a little way into the Tys-ashu valley, named by him the Tesnyk-basy. He A\Tites {SapisJd, Imp. Russ. Q. G. vol. xix. 1888, p. 89): "The third glacier, which is not even mentioned by Ignatieff, and is omitted even on his map, is the glacier which lies at the foot of Tesnyk-basy, perhaps one of the highest peaks after Khan-Tengri, and bears, like the peak, the name Tesnyk-basy. This glacier with its snow-fields is obviously connected with those of the Inylchek glacier region. The valley of the Tesnyk- basy, the second affluent of the Sary-jass on the left, I followed as far as the frontal moraines of this glacier, which was evidently but little inferior to the Mushketoff glacier. To my regret I was prevented by the bad weather," etc. THE TYS-ASHU PASS 65 In the dividing wall between Tys-ashu and Inylchek there occurs an ice-clad pass about 13,300 ft. (4,050 m.), which, as affording the nearest access to the Inylcliek valley, I crossed with the caravan, though not without some difficulty. This I call the Tys-ashu pass. During the ascent of the pass the track lies between limestones and limestone schists, which have an east-north-east trend, but in the vicinity of the pass, develop folds dipping north and with granite cropping out along their edge. Owing to its close contact with the granite, very little of its great wealth of fossils has been preserved by this carboniferous limestone forma- tion. Nevertheless, by repeatedly crossing the pass we managed later, to collect some that could be identified. On the south side of the pass the limestones are tinged red, calcined, and greatly disintegrated. There also occur conglomerates and friction-breccias, indicating the discharge of eruptive matter from a spot, which I was later able to locate on the north-east side of the pass in the neighbouring Kusgun-ya valley. Flanking the gate-like entrance of the pass there shoot up hundreds of obelisk-like limestone crags, into which the masses have been decomposed by the marvellous action of erosion. If from these strange surroundings we turn to the south and east, we see, some 3,300 ft. (1,000 m.) lower down, the boulder-strewn floor of the broad trough of the Inylchek valley, walled round by many- crested snowy ranges, whose crest-line rises with an average elevation of over 8,000 ft. (2,500 m.) 5 66 THE INYLCHEK GLACIER above the valley bottom. The eye also lights a little higher up on an ice-field of extraordinary extent, similarly bordered and stretching away to the east. Even though the observer be accustomed to the sight of the loftiest eminences on the globe, the Himalayas, Karakorum, etc., a feeling of wonder and amazement will still be produced by the first view of the extraordinarily abrupt southern border- range of the Inylchek valley. Here are unfolded the mightiest elevations of the Tian-Shan. A gigantic range, surmounted by the wildest and most rugged snowy peaks of the most diverse forms ever sculptured by the creative forces of nature, is seen stretching away to the east for a distance of some fifty miles, altogether one of the grandest Alpine pictures on the globe. Amid this proud phalanx the most magnificent is one mountain which rises opposite the pass, the same that, as already stated, is partly visible from the Tys-ashu valley. It is difficult to conjure up an adequate picture of the mighty, far-reaching spurs of this giant, of the wildness of its many fractured crests, the splendour of its precipitous glaciers, carved in a thousand varied forms and broken into endless fragments. I do not hesitate to pronounce this marvellous moun- tain mass, some 21,300 ft. (6,500 m.) high, to be the grandest in the Tian-Shan. For it an appropriate name sliould certainly be found. The Central Tian-Shan attains its highest mean altitude not, as hitherto supposed, in the southern chain THE INYLCHEK GLACIER 67 of the SemenofF glacier, but in this range, whose crest trends away to the east-north-east at an average elevation of 18,000 ft. (5,500 m.). From this point there is a gradual fall of the mountain mass towards the south. To our surprise, however, Khan-Tengri, the absolutely highest eminence of the Tian-Shan, did not appear in this range, and the question of the position of its actual basis still remained unsolved. Seen from the pass the Inylchek glacier already produces a profound impression, although its lower section, being for many miles completely covered with debris, has not at all the air of an ice-stream, and, owing to the windings of its bed, its whole course cannot be taken in at a glance. Neverthe- less, it struck us all at once that IgnatiefF's estimate (eight miles long) fell far short of the reality, although the enormous extent of the glacial stream was not made fully evident till the next year's exploration. The bed of the valley beyond the glacier has an extremely slight incline, and throughout its upper course, mth an average breadth of a mile, it forms a shingly desert, completely levelled in by its covering of detritus, through which the mighty stream ramifies ir- regularly. Despite this distribution of its volume the crossing is difficult, as each branch still has a deep bed of considerable breadth — in fact, a copious and rapid stream. Where these waters unite in a single arm, at certain reaches of the middle course, the crossing is possible only in the early morning. As during the next year I ascended (58 THE INYLCHEK GLACIER the valley through its whole length from its con- fluence with the Sary-jass to the Tys-ashu pass, and as in the later part of this report I shall have to deal with the observations then made, I will confine myself for the present to a few details of the physical features of the upper course. Here also basin-shaped expanses up to a breadth of two miles are met with. Such a tract, some thirteen miles below the lower end of the glacier, is blocked by a low ridge of limestone schist cliffs, forming a barrier, which with a length of nearly two miles, stretches obliquely across the bottom of the valley, here some two miles wide, so that an opening of not more than about 500 ft. (150 m.) is left for the outflow of the waters. On the extremely disturbed and dislocated rocks of this old barrier there still lie the remains of old ground moraines. In this valley, too, the old glacial deposits acquire quite an extraordinary development. On the descent from the Tys-ashu pass we already meet with them here and there, 1,000 ft. (300 m.) below the level of the pass— that is, 2,000—2,300 ft. (GOO — 700 m.) above the bottom of the valley— and in hke proportions they are seen along the downward course of the valley. Hence nearly all the mouths of the transverse \'alleys stand very high above the present bed of tlie main valley. Of these transverse valleys, however, only a very few occur throughout the whole middle and lower course ol this long ri\-cr-bcd. Owing to the rapid change of chmate after the retreat of the lateral glaciers in the post- FLORA 69 glacial epoch, erosion had no longer produced any great effects in this district, as I have already shown by other examples {vide pp. 35, 45). In the Inyeklch valley also, as in the other large longitudinal valleys and for like reasons that have already been dis- cussed, the imposing Alpine character is confined to the southern flanking range, at least in the ice-free part of the valley. In the upper course all vegetation, except a rubble flora, is banished from the bed of the valley, and confined to the slopes on both sides, where, however, it is displayed in the sharpest contrasts. The slope on the right bank, facing southwards, is treeless and scrubless, and only the lower part covered with thin stunted, grassy growths, which assume the aspect of meadows only in a few tracts, sheltered from intense insolation by the disposition of the slope. On the other hand, the slope on the left bank, facing northwards, is decked with bright Alpine meads, and strangely contrasting with the woodless Sary-jass valley, even shows some- what dense patches of pines. This is all the more remarkable since the Inylchek valley, although it has the same trend as the Sary-jass valley, lies considerably farther south, and according to my meteorological records is distinguished by greater dryness of the atmosphere ; while on the other hand, in the Sary-jass valley even the slope facing southwards is carpeted with lovely Alpine meadows, which are missing in the parts of the Inylchek valley, enjoying the same aspect. On the contrary, on the southern slopes of the Inylchek valley pine- 70 THE INYLCHEK GLACIER groves are met with, wherever the least mountain debris has been brought down from the steep rocky walls of the valley and deposited in cones at the foot of the cliffs, or where morainic drift lies on banks and terraces. The discrepancy of these re- lations cannot be explained by the nature of the soil, since the constituent elements of the mountains are of much the same character in both valleys. For a distance of about twelve miles on the same side of the valley a green zone stretches from the end of the glacier tongue along the foot of the rocky walls far up in the frigid zone. Short Alpine grasses, a rich Alpine flora, and, besides other bushy plants, the Caragana shrub (Siberian Pea-tree) of dense forest-like growth, form the chief components of this pleasant floral zone, which extends right up to the region of perennial frosts and is associated witli old lateral morainic drift. Strange to say, for about the same distance (some twelve miles) the glacier is covered across its whole average width of about two miles by a mound of morainic debris and large boulders at least 450 ft. (100 m.) high. By atmospheric influences, by the erosion of the waters and by the movement of the glacier, this mound has been disposed in ridges and peaks of the most diverse forms, valleys, troughs, cauldrons — in a word, every form developed by a real mountain range. The material for this work has for the most part been supplied from the slopes of the chains skirting the main valley and from those of its ravine-like lateral valleys, which along the lower course of the glacial stream are AN EARTHQUAKE 71 free from ice up to a considerable height. Owing to the extraordinary fluctuations of the tempera- ture in this valley, and its southerly position, the disintegration of the rocks has been carried to an unusual extent, while the material entering into the structure of the mountains, here mainly schists, offers but slight resistance to such influences. Still, the climatic conditions alone could not have caused such great effects, had they not been supplemented by the incredible disturb- ance of the strata. Here we are in a region of the most profound and manifold dislocations, which are exposed in many places on both sides in the cliffs, skirting the lower course of the glacier. That seismic movements, however, have not yet ceased in this region was shown by an earthquake, which occurred on the morning of August 22nd, 1902, lasting about half a minute, and making itself felt in three very severe shocks, proceeding from below upwards. A fearful, never-to-be-forgotten spectacle was presented as an immediate result of this disturbance on the precipitous glaciers of the gigantic mountain above described, at the foot of which we had established our headquarters. Huge masses of ice were set free, and came tumbling down with an indescribable crash into the gorges of the huge rocky buttresses, from which great coluinns of powdery snow and ice then rose up to the level of the sno^vy crests of the great mountain. The mound of detritus, piled up on the glacier is so compact that ice crops out only at the edges, 72 THE INYLCHEK GLACIER so that the ice-stream, although it descends lower than the Semenoff, is thus prevented from thawing, despite the fact of its penetrating far into a southern cHmate. As the melting waters are forced to seek an underground outlet beneath the overlying drift, they excavate rudimentary crevasses at the end of the glacier, scooping them into hollows, where the water is then collected. On the advent of the warm season the pent-up waters seem to occasionally burst their fetters, and discharge themselves with irresistible force over the plains, carrying with them huge masses of ice. Even so late as the end of August in the year 1902, and at a distance of two miles from the glacier, I came upon several blocks of ice as big as a house in the boulder-strewn Inylchek valley, exposed though it is to such extreme insolation. The only explanation I can offer of such a phenomenon is that above suggested. A visit to the glacier in two successive years enabled me to determine the altitude of its lower end, at about 10,500 ft. (3,200 m.). There were no indications at all of any recent retreat of the frozen stream. Its stability is sufficiently ex- plained by its enormous development, its slight incline — only about eighty feet per verst — and compact morainic covering, which itself stands in close relation with the slight incline. Tliis mound of detritus necessarily makes the exploration of the lower section of the glacier extremely toilsome and fatiguing. In a day's march one can cover only a few miles. Being CRIPPLED RESOURCES 73 unmindful of this circumstance, and also un- prepared for the vast dimensions of the glacier from the hitherto published reports of its magni- tude, and moreover unaware that at this season the valley is not even visited by the nomad Kirghiz, I had not brought sufficient supplies to meet the wants of the party for eight or ten days, the minimum of the time, required for profitable work on the glacier. The number of porters was also insufficient for such undertakings, while these fellows themselves struck work at critical moments, and broke out into open revolt against me. Under such circumstances I was fain to confine myself to a short excursion in the region of ice. The expedition separated into two parties. Herr Keidel descended the valley with a small party in order to make a survey of its geological structure, and with a view to acquiring some knowledge of the local conditions, he pushed on to the next large parallel longitudinal valley, the KajTidy valley, which lay to the south, but was still entirely unknown, and not e\en figured on the maps. As I next year explored this valley and another, stretching still farther south, informa- tion regarding them Avill be found in the later parts of this report. Herr Pfann and I plodded across the morainic mound of the glacier, making very slow progress. After covering about two miles, we saw, rising behind the heap of drift, a broad, massive rocky wall, dark, but capped with neve. IMuch farther on, where 74 THE INYLCHEK GLACIER the ice begins to be free from ddbris, this rampart diWdes the A\dde glacial stream into two branches. When we got a Uttle higher up, a bright, slender pyramid was seen towering into the air, but much fiirther back, at the side of the dark mass and close to its northern flank. We at once recognised it as the summit of Khan-Tengri. Owing to a peculiar bend in the axis of the valley and in the trend of the range, of which the dark rampart evidently forms part, the interesting picture seems to the eye shifted in such a way that the observer remains uncertain as to the grouping of the mountain ranges and the position of the breach, from which rises the pyramidal peak. A few hundred steps farther, and this peak is no longer seen at all. Still, there was great probability that it must stand somewhere in the Inylchek valley, or in one in some way connected with it. In order, tlierefore, to get a better insiglit into these relations, we decided to make our way over to the left side, bivouac tliere on the edge of the glacier, and ascend a lofty summit, rising above the border-range. From such an elevation w^e hoped to get a clear notion, regarding the trend of the ranges along the valley and the position of Khfin-Tengri, and to be able also to take telephoto- graphic views, since the unfavourable circumstances, ah-cady mentioned, prevented us for the present from penetrating farther into the mysterious glacial region. Leaving the execution of this project to Ilcrr Pfaim, I undertook to investigate the com- plicated disturbances in the structure of the range ; A CLIMATIC PHENOMENON 75 these can be best observed in the fine exposures of the steep walls on the right side of the valley. However, the extremely crumbling state of the schists forming the rocky crest of the mountain to be scaled, together with the treacherous con- stitution of the upland snow, prevented Herr Pfann from reaching the summit. During the ascent atmospheric disturbances also set in, so that very little remained visible of the mountain ranges, A strong atmospheric pressure now prevailed, heralding snow-storms. To my deep regret I had in all haste to quit the valley, which I had so hastily explored, and of which I had seen so little ; but there was no option if my retreat over the pass was not to be cut off by the snow. Not till the next year, when I returned better prepared, was I successful in unravelling the mystery of the conformation of this valley, on which more details will be found in the later parts of this report. Here I should like just to draw attention to a peculiar phenomenon in the climatic relations of the valley. During the five days of my sojourn in it, there regularly sprang up in the later hours of the afternoon whirlwinds, which carried aloft considerable quantities of dust from the ground and again deposited them as loess on high-lying ledges and little terraces in the walls along the margin of the glacier. Extensive banks of this aeolian precipitation may be observed, especially along the left edge of the glacier. CHAPTER V FROM THE KAPKAK VALLEY TO THE GREAT MUSART VALLEY On our return from the Tys-ashu to the Sary- jass valley Herr Pfann and I left the caravan, and ascended the above-mentioned dividing ridge between the Kusgun-ya and Tys-ashu groups of valleys, the Tur plateau, about 12,300 ft. (3,750 m.). Here we saw the pyramidal peak of Khan-Tengri, rising above the surrounding ranges far more boldly than from any of the other, even higher points hitlierto visited. The ranges, however, as seen thence, seemed to be shifted in quite a peculiar way, so as to give the impression that Khan-Tengri rose at the head of a valley with a north-easterly trend towards the JNIusart pass, or a little to the south of this point, but at its origin apparently connected with the head of the Inylchck valley. The view was sketched and photographed, which took up so much time that we had to pay for our exploring zeal by an exposed bivouac without shelter or provisions, and did not overtake the caravan till the following day, after crossing the Kapkak pass, about 12,150 ft. (3,700 m.), in the valley of the same name. 76 LUXURIANT GRASSES 77 This valley, running some forty-three miles in a south to north direction, is amongst the most important transverse valleys along the upper course of the Tekes. The Kapkak pass lies at the converging point of four valleys, as owing to a lateral thrust ( Vorwe7''fung) the ranges here diverge widely from each other. For this reason the Kapkak river, with its large and widely ramifying affluents, effects the drainage of a very extensive territory. The trip across this charming valley ranks amongst the most enjoyable excursions in the Tian-Shan. All the elements that combine to form a romantic Alpine dale are here repre- sented in the greatest profusion. The pine forests are magnificent, and contain trees of gigantic size. The r" -^velopment of the Alpine flora is, next to that of the INIukur-mutu valley, the richest and most luxuriant that I have seen in the Tian-Shan, while the gro\\i:h of Alpine grasses is astonishing. Phyllites, granite, syenite, fossiliferous limestones, and calcareous slates form the geological structure, which resembles that of the Bayumkol valley, but owing to the disturbances, that have here taken place, presents in many respects a special interest. For the study of the later vicissitudes of many Tian-Shan valleys the Kapkak basin offers some special features, particularly in its lower course. Although at its head neve and ice are at present quite insignificant, every indication of complete former glaciation may here be observed. In its upper course ancient moraines acquired an inmiense 78 TO THE GREAT MUSART VALLEY development, and in its lower course the masses of flmioglacial rubble have in some places been deeply eroded by the river. Its course, which formerly lay more to the east, having been shifted by the resistance of such boulder deposits or by the ice, the stream has been compelled, in order to reach the Tekes, to eat its way through a mighty barrier of hard limestone in a steep im- penetrable canyon. The waters, formerly dammed up by glacial drift, have flooded basin-shaped expanses, thus forming lakes. The side- valleys which debouch here lie very high. They have been eroded in trough-like form, and although formerly enclosing small lakes, are now empty, while their mouth lies high above the beds of the former lakes of the main valley. Reasons for this disposition have ah'eady been several times dwelt upon {vide pp. 35, 45, 69). Later irruption of considerable bodies of running water may be assumed from the circum- stance that loose younger conglomerates are found deposited, high above tertiary formations, like that lying on the borders of the old basin-shaped expanses of the Tekes valley. These deposits extend in places even beyond the tertiary to the limestones. Besides the tertiary beds we also see, exactly as in the Tekes valley and at other places, great quantities of sand and debris, w^hich are derived from disintegrated and eroded granites. During the su])sequent course of the expedition I visited tlie Kara-kul-say, one of the largest lateral valleys of the Kapkak, in which there is still a lake, dannned up by old moraines, and in which the REPLACING PORTERS 79 indications of those, already vanished, are well pre- served (see particulars farther on). Towards tlie end of August, after my return to Narynkol, I lost some valuable days there in exchanging the worn-out horses, and especially in procuring fresh " Jigits " and porters to replace the former, to whose refractory conduct is partly attri- butable the fact, that so far the journey had yielded such slight results. At last I was able, at the beginning of September, to return to the Bayum- kol valley to resume the previous operations, inter- rupted by bad weather. I hoped to be favoured by more settled weather in the advanced season, when the contrasts of temperature between the plains ^xid uplands are less pronounced. Unfortu- nately general atmospheric disturbances took place, again seriously impeding and delaying our work. For the same reason the intended ascent of one of the high snowy peaks at the head of the valley had to be put off. The only ascent was that of a gi'anite eminence, about 14,450 ft. (4,400 m.), at the northern edge of the western glacier, from the top of which a panoramic view of the surrounding- mountain ranges was obtained. Herr Pfann also, despite the unfavourable weather, was able to complete the survey of the western glacier, and to determine the height of the peaks of the border- ranges from an elevated basis. In the course of our wanderings in connection with these operations I came upon a breach about 14,000 ft. (4,250 m.), free from ice, in the ridge, which separates the Karakol river, flowing to the Sary-jass (see above) 80 TO THE GREAT MUSART VALLEY from the basin of the western Bajoimkol glacier ; and here also 1 obtained a magnificent view of Khan-Tengri, appearing through a gap in the southern border-range. In this breach 1 found five decayed posts, jammed in between rocky boulders. At first I supposed they might have belonged to the Ignatieff expedition, and that the breach in question was identical with the gap, which this traveller named the "Narynkol pass," and which he states is 13,580 ft. high. After, however, again reading the passage in IgnatiefF's report {Isvestiya Russ. Geograph. Soc. vol. xxiii.), I hesitated as to this assumption, because Ignatieff made the descent from the pass down to a glacier and traversed it lengthways on horse- back, which for the western Bayumkol glacier must be pronounced absolutely impossible. Nor could Narynkol be reached in one day from this glacier, as is asserted by Ignatieff. Lastly, the difference between our two determinations of height is so great, that these cannot have reference to the same position. Hence Ignatieff must presumably have crossed at some other point. Tlie western Bayumkol glacier is formed by the confluence of fiv^e glaciers, issuing from recesses in tlie walls enclosing the valley, and is much broken up, especially in its middle course, and its upper nevd is likewise much crevassed. Here it communicates with the Semenoff glacier by a snowy saddle a})out 14,4.50 ft. (4,400 m.), which I readied the following year from the Semenoff (see below), and is also connected with the upper nev^ HEAVY SNOWFALLS 81 basin of the same SemenofF glacier by the SemenofF pass (see above). A connection also undoubtedly existed formerly with the Karakol glacier, and in the ice age all these glaciers evidently formed a continuous ice-field. At pre- sent the ridge between Karakol and Bayumkol is free from ice on the side (south-east), facing the latter valley, and here the sedimentary rocks (lime- stone, marble, clay schists) are seen lying finely exposed in several strata between the granites. Unusually heavy snowfalls at last drove us (September 20th) from the uplands, where no fodder was any longer procurable for the horses. The snow already reached down to the Tekes valley, and I was fain to postpone till the next year all the explorations in my programme on the north side of the highlands, and cross over to the south side, wliere more favourable conditions might perhaps permit of more protracted operations. 6 CHAPTER VI NORTHERN MUSART VALLEY, MUSART PASS, AND SOUTHERN MUSART VALLEY After a few days' preparatory work, the expedi- tion left Narynkol on September 23rd, in order to surmount the Great INIusart pass, which had ah-eady been traversed by a few Russian ex- peditions. Von Kaulbars has pubUshed some notes on the topography of the district, and Ignatieff on its geology. I shall, therefore, in this report limit myself to some hitherto only partially known or quite unkno^vTi particulars, reserving for the more detailed account of the journey a series of physico-geographical observations, for which the crossing of this pass afforded ample opportunity. The downward route from Narynkol through the Tekes valley leads through one of the best-defined basins of the old frontal lakes which formerly lay at the base of the mountain range. On the southern border the outlines of the old terraced beaches have been excellently preserved. At the wide entrance to the INIusart valley beds of fiu\'ioglacial deposit form five ancient terraces, and for several miles, follow the course of the valley as longitudinal banks, nearly up to the foot of the mountain mass. 82 A DISASTER 83 Here, in the vicinity of the first Chinese miHtary post, where the copious Musart river emerges from the highlands, it is joined by its equally copious affluent, the Dondukol (on which more below). The united stream is not easily crossed, and through the carelessness of a " Jigit " during the passage, the expedition met with an accident of far-reaching consequence. One of the packhorses stumbled, and his load — two tin boxes bought as " air-tight " — fell into the water. When fished out, the contents were found thoroughly saturated. Amongst them were a great number of large, exposed " Edward films," shut up in tin boxes, which were supposed to be " absolutely air-tight." Relying on this, we omitted to open them im- mediately after the mishap, and when we did so later on, it was found that water had penetrated through, destroying all the films. Sixty views, 6^ by 8 English inches in size, mostly panoramas and telepanoramas, taken from lofty positions, the fruits of indescribable toil and care, the main result of the summer's photographic operations, geographical documents of priceless value, were irreparably lost. By this disaster the course to be followed by the expedition in the next year was in a way already marked out. Records so important for the topography of the Central Tian-Shan, could not be dispensed with. It was necessary above all to revisit the more conspicuous points, from which the lost photographs had been taken. However keenly felt tlie damage was at the time, still it proved beneficial in the end. 84 MUSART VALLEY AND PASS Being compelled again to retraverse the uplands already visited, and being then also more famihar with all the local conditions, and moreover favoured by good weatlier, I was able in the follov\ing year to work better and more successfully than in the first summer, and in most cases to find the solution of what had hitherto puzzled me in the structure of the Central Tian-Shan. At the entrance of the Great ^lusart valley is seen a vast series of chloritic schists, often inter- stratified with phyllitoid schists. Just before emerg- ing from the highlands, the river breaks through masses of red granite, which are followed higher up by a narrow zone of gneiss. But aphanites are soon developed over a wide area, and farther up the valley, where they again approach a granite stratum, they assume more and more the char- acter of schists. These schists, with an almost northerly strike (N. by 10^ E., which for this region is abnormal), have been throM^n into irregular narrow folds. Pressure-phenomena occur also in the granite, which not infrequently assumes the form of granitic gneiss. Limestones and clay- slates, cropping out between the granites, have, owing to dynamo-metamorphic processes, been pressed, the former into schistose, the latter into crystalline forms. More regular conditions do not appear till far back in the valley, where a normal easterly trend (N. by 7° E.) is resumed. Here the granite occurs under very diversified forms, even as porphyritic granite, and in some parts is replaced by syenite. Over ^ somewhat wide THE VALLEY 85 zone it is followed by gneiss and other crystalline schists. Yet the nearer we approach the head of the valley the more prevalent become dark, more or less crystalline limestones, clay-slates, and marbles, from which, as in the other large valleys, are exclusively built up the sections of the crests, be- longing to the main water-dividing range. Here, however, also occur great masses of dolomitised limestones, which present the same bold and fantastic summits as we are familiar with in the dolomitic limestone highlands of South Tyrol, and under these forms they flank the defile of the Musart pass southwards throughout nearly its whole length. The Great Musart valley, as far as it lies within the highlands, has a length of from thirty-six to forty miles, and is distinguished from the other large Central Tian-Shan valleys by the somewhat steeper incline of its bed (average about ninety to ninety-five feet per mile). At the outlet of the river from its narrow upland course, about 6,200 ft. (1,900 m.) great quantities of fluvio-glacial drift are deposited on both sides of the valley, where they form terraces {vide p. 28). At the confluence of the Dondukol (not Maralta, as it is wrongly called by Ignatieff'), which here joins the main stream at an obtuse angle, these terraces either intersect or are piled up against similar formations deposited by this afiluent. Like other Tian-Shan valleys, that of the Great Musart is likewise disposed in basin- shaped expanses, which are connected by gorge- like narrows. These narrow^s are for the most part 8G IMUSART VALLEY AND PASS choked by old moraine deposits, through which the river has everywhere cut itself deep channels, which, liowever, rarely reach down to the bed-rock. In the basin-shaped expanses the moraine debris is found deposited for the most part on the left bank in step-like terraces, rising one above the other. During the ascent the traveller wanders along tne picturesque valley on the slopes of the left bank, decked with magnificent pine groves, especially in the middle course, exclusively over Alpine meads and forest-bearing old morainic soil. In several places the old terminal moraines are of enormous dimensions. At the confluence of the Khamer-davan, about 7,900 ft. (2,400 m.), of which more farther on, lies the largest of these beds, which has a width of nearly two miles and forms a huge mound in the valley. Another nearly as large lies about seven miles higher up, at an altitude of some 8,400 ft. (2,600 m.), and still maintains a height of 250 ft. above the level of the \alley. The morainic drift is conspicuous to a considerable height on the walls of the valley, while ice-worn rocks and i^oches moutonnes may be observed on the face of the cliffs. Here also, besides several deeply eroded lateral valleys of ancient origin— Dondukol, Khamer-davan, Atun- bulak, etc. — one can distinguish a series of high- lying trougli-shaped, younger valleys with cirques at their licads, and tlie mouths of which are raised higli above the present bed of the main stream, thus indicating the former level of the chief glacier, \\ liicii once filled up the whole valley. They still A WILD GT.ACIAL STREAM 87 retain small glaciers. A peculiar phenomenon in this wind-sheltered valley is the deposit of loess-like beds of considerable thickness (fifty to sixty feet) on old morainic terraces. They appear to be formations of fluvial origin, although showing a resemblance to aerial loess. About half-way up the valley there occur hot springs (48° C), near which the Kalmuks have built some primitive huts, while utilising them as medical baths. They well up in the valley-bed at the level of about 8,400 ft (2,550 m.), in the zone, where crystalUne schists and granites come in contact with greatly disturbed limestones. At the point where the valley-bed describes a semicircular curve of short radius towards the east, the range on the right bank, apparently shutting in the valley, rises to a series of bold lofty peaks, about 18,000 ft. (5,500 m.), which, owing to their northerly exposure, tower with their fronts completely enveloped in snow and ice, superbly above a darkly-wooded old moraine. At their foot the wildest of the big glacial streams, that I have seen in the Tian-Shan, bursts out from a lateral valley, coming from the east, and is dis- solved into a series of wonderful cataracts, with thousand-fold scracs. This glacier on reaching the bottom of the main Aalley turns northwards, and ends at a height of 9,000 ft. (2,750 m.), a httle above the third Chinese post, where it is separated from the main stream by a huge lateral moraine, which it has here deposited. Judging from the height of this morainic ridge (up to 200 ft), from the 88 MUSART VALLEY AND PASS immense size of the boulders, consisting exclusively of light dolomitic limestone and marble, and from the great thickness of the ice at its tongue, this still unexplored glacial stream must be of gi'eat length. Beyond doubt it has its source at the dividing ridge, by which the head of one of the upper affluents of the Agiass, which flows north to the Tekes, is separated from the Musart valley. From there also — that is to say, from the main crest of the Khalyk-Tau in the east — stretch the dolomitic limestones and marbles, composing the above-mentioned lofty snowy peaks, and abutting here on the granites and gneisses. The protection, afforded by this rampart with its northerly face, has secured an exceptionally mild climate to the part of the valley, lying behind, despite its great elevation (9,200 ft. ; 2,800 m.) The result is seen in a wonderfully beautiful bush and forest vegetation, ranging right up to the glacier ice. The INIusart pass is a "wall-pass," whose ir- regular flat top has an extent of over ten miles. From the north side, the ascent, which starts from tlie elevated terraces at the head of the northern Musart valley (about 9,500 ft. ; 2,900 m.), is short and steep up to the plateau ; the descent to the south, down to the 'J'amga-tash post (about 9,050 ft. ; 2,700 ni.) is long and gradual, excepting a few steep steps, so tliat the two sections are unlike. An anomaly is seen in the fact that the glacier on the nortli is short, while that on the south side is very extensive. The Yalin-Klianzyn glacier, THE JIPARLIK GLACIER 89 descending northwards, is now merely an insignifi- cant remnant of a formerly extensive ice-stream. It terminates at a level of about 10,200 ft. (3,100 m.), and is almost entirely covered with detritus, so that a little ice is visible only at the confluences of small lateral glaciers. The watershed between it and the Jiparlik glacier, descending southwards, is almost obliterated. Owing mainly to the very shifting accumulations of morainic drift, it is difficult to determine the culminating point of the pass. We considered this to coincide with a small plateau, whose altitude was calculated by a preliminary survey at about 11,480 ft. (3,500 m.). Ignatieffs figure is 12,240 ft. (3,730 m.). Near the top of the pass on its southern slope the mighty Jiparhk glacier descends from the east- north-east. The glacial stream where it covers the highest plateau of the pass is nearly free from debris, and over a slightly inclined stretch of several versts the surface ice is divided into millions of tiny, tent-shaped knolls, the origin of which is to be attributed to peculiar melting processes. As far as the eye can penetrate up the course of the glacier, from 300 tO 400 m. or about 1,200 ft. broad, high snowy mountains (limestone and marble) are visible along its margins. But owing to a bend in the valley the source itself cannot be seen. It seems to He in the same dividing ridge as the already mentioned large glacier, which joins at the bend of the main stream. Near its outlet on the plateau of the pass, an arm of the chief glacier, branching off to the south-west, stretches obhquely 90 INIUSART VALLEY AND PASS across the plateau, and disappears in an opening facing the south-west in the wall of the west margm. The main glacier itself trends, with an average breadth of one and a half miles, first south-eastwards, then southwards, and terminates at a height of about 9,500 ft. (2,900 m.) in a tongue (now rapidly retreating) above the Tamga- tash post. Here a waterfall bursts out through a gate-like aperture in the ice-wall. At the time of my visit, above the lowest cavity were still to be seen two other, quite similar but empty cave- like outlets, standing one above the other in the terminal wall of the glacier. Hence the stream had evidently cut its bed deeper and deeper in the ice. Its waters had once been dammed up in a morainic lake about two miles long and one mile wide in front of the glacier. As far as the glacier covers the plateau of the pass on its gently inclined southern slope, the ice is almost hidden by a coating of debris ; where it shows itself it is beset with a very large number of funnels, in each of which lie one or more large boulders, whose great absorption of heat gave rise to these hollows. On the rocky enclosing walls, over 3,300 ft. (1,000 m.) liigh, tlie traces may everywhere be noticed of the grinding force of the glacier ice, indicating how it formerly completely filled the upland valley. On the left bank, at the foot of an ice-polished marble wall 1,300 ft. (400 m.) high, the ruins of a mazar and of the Mazar-bashi post stand on a rocky, gently inclined terrace. At this point, where a lateral glacier debouches, the main glacier breaks with a A TERRIBLE PASSAGE 91 fall of about 350 ft. to a lower terrace, where its glacial masses are dissolved in wild seracs, horns, and pinnacles, separated by yawning chasms. This is the famous passage which has been a terror for hundreds of years, and cannot be surmounted by the caravans without the aid of the guards at the Tamga-tash post, who have excavated regular stair- cases in the icy pinnacles. But the skeletons of pack-animals, strewn about in large numbers, show how great are the perils of the passage, despite all aid. Nevertheless, this pass is still relatively the easiest for communication between the north and south sides. A caravan, floundering amid this maze of icy turrets presents a strange spectacle. At the foot of the succeeding glacial terrace an extensive lake occupies a hollow in the ice near the left bank. The whole length of the Jiparlik glacier cannot be estimated at less than sixteen miles. It has already been pointed out that dolomite Hmestone, carved into exceptionally bold peaks, together with white marble, forms the prevailing constituent of the ramparts, flanking the JNIusart pass. These light-coloured masses stand in sharp contrast with the dark jagged walls of highly meta- morphosed eruptive rocks, which uninterruptedly accompany the metamorphic sedimentary beds from the head of the defile in the north down to its southern end and far beyond it, both sets of strata sharing in the later contortion, of which extraordinary instances are here and there grandly exposed. Owing to the prevalence of a north- easterly trend with a marked incline to the east, 92 MUSART VALLEY AND PASS gneiss and syenite are noticed only on the north side of the pass. The route through the southern JNIusart valley, which has a length of about sixty miles, with a breadth of from one to two miles, presents great interest in two respects. In the first place, there are the tremendous dislocations, to which not only the igneous rocks (gneiss, granite, syenite), but also the sedimentary formations, have all been sub- jected, and then the great masses of eruptive rock (diorite, porphyrite) which have burst through both series. A more careful study of the con- ditions observed will be needed, before it can be decided whether the disturbances were in the first instance caused by the intrusive igneous rocks, and hence were to a certain extent local, or whether the whole massfif was affected by wide-ranging convulsions, followed or accompanied by the in- trusion of the magma in the chasms thus formed. Here, too, as is so often the case, the zone of contact awakens the deepest interest. Extensive metamorphic plienomena are seen, not only in the contact zone of the erupted matter with the sedimentary and old crystalline rocks, but also wlicrc tlicse two are found associated together. During our second visit to the valley, Herr Keidel made a complete collection of specimens from the contact zones. In the soutlicrn JNIusart valley granite, syenite, gneiss, etc., occur, only at greater distances from the central ranges than in any of the northern and soiillicni transverse valleys, visited by me — that GEOLOGICAL STRUCTURE 93 is, only in the outer section of the valley, up to where sedimentary rocks alone enter into the structure of the highlands. Gneisses are far more extensively developed than had hitlierto been sup- posed. Between the Khailik-IMabuse and Tograk posts they form an uninterrupted zone three miles long, sharply limited at both ends. Chloritic and highly metamorphosed schists are interstratified with granites. The limestones too, occuring here, became more or less crystalline. The walls of the ranges, running obliquely to the trend of the valley, often perpendicularly cut to a height of 4,900 ft. (1,500 m.) and more, show in their strata the most remarkable and diversified bendings, crumplings, and foldings of the vertically disposed sedimentary beds even dowii to the minutest wrinkles, with exposures on the grandest scale, and always most pronounced in the neighbourhood of intruding dioritic rocks. In some places the intrusion of the magma is in dykes, accompanied by extensive apophysic formations. Despite the dynamic effects, accompanying the powerful dis- location of the sedimentary beds, Herr Keidel succeeded in finding a limestone bed that had been spared, and in it collected a fauna belonging to the upper carboniferous age. This justifies the conclusion that these limestones of the middle and lower valley, and the crystalline masses between which they rest, are from a tectonic point of view to be separated from the older pala30zoic lime- stones of the head of the valley and from the metamorphic eruptive rocks, folded together with 94 MUSART VALLEY AND PASS them. Old crystalline conglomerates begin to occur in the second section of the valley, but do not appear in larger masses till near its outlet, Avhere they are associated with sandstones and metamorphic schists between the lateral Ak-topa and INloro^khotan valleys. Exposures in walls 1^300 — 1,600 ft. (400 — 500 m.) high also reveal in these conglomerates extraordinary strains and twists in the strata. The great pressure is attested by blocks of conglomerate, which are streA\Ti about, and whose constituents have been crushed out lengthwise. These conglomerates also form the slope of the range, facing the valley of the Musart- daria, flowmg to the east, of which more farther on. A few miles beyond, where we struck south from the JNIusart, sandstones again occur on the slope of the range facing the steppe at the mouth of the Kash-bulak valley. These are compressed in close folds together with coarse, schistose- calcareous and fine conglomerates like grauwacke, and in places contain fractured, laminated, shiny carboniferous clay slates {Lettenkohlenschiefer), elsewhere even real anthracite. No less interesting than the peculiar geological features of the Musart valley are the indications of its former extensi\'e glaciation. If in this valley, trending southwards, the old morainic deposits are seen in much larger quantities and less destroyed than in the great glacial valleys on the north side, the ex})Ianation, as already rightly suggested by Ignaticff, is tliat in the north, owing to the very extensive glaciation, wliich to a considerable extent OLD MORAINES 95 even now continues, the old masses of boulder drift were for a long period, and are to the present day, exposed to the vigorous erosive action of the melt- ing waters. Here in the south, on the contrary, where the present glaciation is relatively shght, the cHmate much dryer, and in any case, even in the post-glacial period, was subjected to more rapid changes than in the north, the destructive and obliterating forces were less effective in the interior of the valleys. Here we see first of all that in some places the valley was blocked by old frontal moraines, and elsewhere by the accumulation of diluvial deposits at natural constrictions, forming six basin-like expanses, which represent so many former lakes. In the second basin, morainic drift lies on high terraces, from 1,000 to 1,300 ft. (300—400 m.) above the level of the valley, while scorings on the surface of the rocky walls here, as farther out in the valley, extend considerably higher up. In some places, as in the fourth basin, the foot of the mountain barrier is Uterally buried in morainic drift up to a considerable height, and this debris forms, for a distance of one and a half miles, a compact covering of the broad channel of the valley still over 200 ft. thick, although much of it has afready been swept away. There, dry weathering has reduced the boulders (marbles, limestones) to sand and dust, above which the still remaining blocks partly project. By these products of weathering a long stretch of the valley has been transformed to a 96 MUSART VALLEY AND PASS real sandy desert, whose dune-like eminences are bound together by plants of genuine desert type. The finer particles have been borne aloft and deposited as loess on elevated terraces, where they often attain a thickness of from forty to fifty feet. At the Khaihk-Mabuse camping-ground, about 8,130 ft. (2,480 m.), old morainic drift rises some 1,300 ft. (400 m.) above the level of the valley. But the greatest accumulations are found in the neighbourhood of the Tograk post, about 7,700 ft. (2,350 m.), where exceptionally large masses of drift have been deposited by the Tograk- Yailak, which joins on the right bank. These were heaped up on those of the chief glacier, whereby the detritus was raised to the enormous height of from 1,C00 to 2,000 ft. (500—600 m.) against the opposite mountain wall. Here the valley is blocked by a barrier of morainic debris some 650 ft. (200 m.) high, through which the river cuts its way in a romantic gorge several miles long. While on the moraines, so far described, the boulders consist of marble and limestone, here scarcely any but gneiss blocks are seen, which aeolic excavation (corrugation) has fashioned into thousands of fantastic forms. Below Tograk tlie lateral Jin-Jilga valley joins on the left side, and from the confluence the gigantic ground-moraine of the old glacier projects in ex- cellently preserved form far into the main a alley. The immense masses of shifted debris, however, cannot be derived from this lateral glacier alone, since they extend as a rampart a distance of six ENORMOUS DEPOSITS OF DRFFT 97 to eight miles downward from 130 to 160 ft. above the level of the river, which has cut its bed deeply into them. The conformation of the mountain range here indicates rather that the gigantic glacier, which has deposited all this detritus, once overflowing the left, here greatly depressed, scarp of the valley, descended from more elevated parts of the Khalyk-Tau in the east. At the last post also (Koneshar, not Kunya-Shar, as it is called in the forty-verst map) the main valley, about 6,900 ft. (2,100 m.), was blocked by morainic drift, which on the left side envelops the mountain walls some distance up. That the old glaciers also extended out into the plain is shown, not only by the morainic mounds, which lie at the foot of the range, where it bends towards the east, and which were crossed the next year by the expedition on the route along the Khalyk-Tau (on this see below), but also by the enormous deposits of shifted glacial drift, including boulders, which to a thickness of several hundred feet still extend for over twenty miles out into the plain, here partly forming closed plateaux, partly disposed by erosion in little ridges of manifold shape. I must here lay stress on the fact that these last-mentioned deposits differ in some essential features from those formations for which INI. Bogdanovich has introduced the term " Kuren " ( Trudi Tibetskoi Eccpedizii, p. 88 et seq.). These masses have been preserved in a region, where erosion, dispersion, and denudation have operated more vigorously than in most other 7 98 MUSART VALLEY AND PASS lands. Granite boulders I found strewn over the desert more than twenty-six miles distant from the foot of the mountains. The lateral valleys of the southern Musart valley, whose parched soil is traversed by a potent stream from which it no longer derives any appreciable fertility, still contain a con- siderable store of glacier ice, where rise lofty and magnificently glaciated ranges, the most superb and richest in glaciers being in the Turpal-che valley, in the cirque-like Chiran-toka valley, in the Serakh-su valley, Tograk-Yailak, etc. Into these valleys the pinewoods also have retreated from the almost dried-up main valley, and where they appear, present the finest contrast to the desert character of the main valley. In this we see one of the most remarkable upland valleys, remodelled by tectonic movements, and the action of ice, water, and wind, a juxtaposition of steppe and desert amid grandiose Alpine sur- roundings. Many other physical features would have still to be dwelt upon to complete the picture. But this would exceed the limits of this preliminary report. CHAPTER VII FROM THE MUSART VALLEY TO KASHGAR Our intention to continue our work a little longer in the liighlands of the great lateral valleys of the southern Musart river could not be carried out, as the valley offered no supplies, either for our men or the pack-animals. Hence the expedition could only be regularly provisioned from a station, lying far beyond the district, but for the organisa- tion of such supplies there was no longer time at this advanced season of the year. The plan was therefore postponed to the spring of the next year, and we took the route, which leads out of the valley to the town of Ak-su, and then for a stretch of about twelve miles, intersects the ranges of the Topa-davan tertiary uplands between the Lyangar and Abad posts. As I am not aware of anything, having yet been published on these uplands or altogether on the tertiary formations at the southern foot of this sec- tion of the Tian-Shan, I may here give some details on the subject. In the structure of the Topa- davan range the same red sandstones are ex- hibited, that we meet with in the tertiary of the Tekes plain and elsewhere, besides red, salt- bearing clays and marls, with gypsum-bearing a9 100 MUSART VALLEY TO KASHGAR marls in some places, and lastly conglomerates of liffht and dark limestones. The whole massif has a general west-north-west trend, and in some places is distinguislied by narrow intricate foldings. Althouoh the mountains are in summer and autumn waterless, they have been carved into several ranges by the powerful erosion of running water, setting in with the melting of the snows and acting all the more vigorously on the mountain mass since it is built up of easily soluble materials, aided also by atmospheric influences, especially wind. The action of all these agencies is helped by the narrow foldings and the vertical disposition of the strata. By such forces these ranges have again been sculptured into a number of the most diversified crests, often affecting the most fantastic forms. In these clay and marly uplands we again find, crowded together in a narrow, readily OAcrlooked space, the same varied features in valley and hollow, the same manifold conformation of mountain and surface as are presented by the higli ranges in wide inaccessible areas, of which we can obtain no comprehensive view. Many of the processes that there took place in a large way, have here been repeated on a small scale. In a word, the mountain-shaping and mountain- destroying forces have combined to produce a relief, which offers an instructive object-lesson on or()graj)iiic structure, so far as regards the diversified character of tlie surface modelling. I later explored tlie tertiary highlands west, north, THE TOPA-DAVAN 101 and north-east of Kashgar, which are built up of similar materials, and also surveyed the Chul-Tau, a southern prolongation of the Topa-davan (on which subjects more will be found farther on) ; but however varied may be the articulation of the surface in some parts of those districts, it nowhere displayed such varied features as the Topa-davan. Its average altitude, rising gi'adually from east to west, is 5,250 ft. (1,600 m.). While the first foothills only reach a height of from 100 to 130 ft. above the almost table-like level surface of accumulated rubble, those approaching the south- west border rise to over 650 ft. (200 m.) above it. Here we are often surprised at the sight of perpendicular mountain walls, about 500 ft. (150 m.) high, which are formed of a single layer of clay, honey-combed like a sieve by the decom- position of easily soluble inclusions. At the A bad post, about 5,100 ft. (1,550 m.), near the south-western border, there occurs a tortion of the axis, combined with a change in the trend — the ridges of the Chadan-Tau, which run from south- west to north-east, here converging with those of the Topa-davan, which run west-north-west. AA^ith this change are connected serious dis- turbances in the lie of the strata. Salt occurs especially on the south-west border in troughs and cavities in the form of efflorescences, which acquire a thickness of up to twenty inches, and are ex- ploited by the Chinese. The mountains appear to terminate abruptly towards the desert, because the low ridges of the outer folds are completely buried 102 MUSAKT VALLEY TO KASHGAR beneath a mass of detritus, several hundred feet thick. The road from Abad through Jam to Ak-su may be passed over as well known. I also omit any remarks on the long stretch from Ak-su through Maral-bashi to Kashgar, although offering occasion for many interesting observations, since it has already been to some extent described by other travellers, as, for instance, most recently by Sven Hedin. On October 18th, 1902, the expedition took up its winter quarters in Kashgar, from which place Herr Pfann and the preparator, Herr Russel, set out on the homeward journey. As the southern border-ranges of the Tian-Shan often remain free from snow even in winter, as was particularly the case in the winter of 1902-03, we ultilised this season, despite the intense cold, to make some excursions to that region, chiefly for the purpose of collecting palaeontological specimens. This object was also effected, thanks to the collecting zeal of Herr Keidel, and we returned to Kashgar loaded with rich spoils. CHAPTER VIII EXCURSIONS ON THE SOUTHERN EDGE OF TIAN- SHAN TO COLLECT PALAEONTOLOGICAL MATERIALS The first excursion led us to the Toyun valley, at first through the narrow gorges of the " Artysh strata" — made known by the publications of Stoliczka and Bogdanovich, — which attain so great a development at the southern foot of the Tian- Shan. Amid these profoundly disturbed strata lies a group of large villages which bear the collective name of Artysh. In August 1902, not long before our arrival, earthquakes had almost utterly ruined both this and the other group, collectively known as Altyn- Artysh, which lies farther east on the southern border of the tertiary range, and was likewise visited by us. These places, now lying in ruins, presented a sad spectacle. For a wide space the ground was seen torn by rents and fissures, and in some places little mud volcanoes were noticed. In connection with these events the study of the " Artysh strata," as they are called, was of special interest to us. Later conglomerates, discordantly overlying these marly-clay and sandstone strata likewise exhibit indications of considerable dis- location. Even in very recent conglomerates, 103 104 SOUTHERN EDGE OF TIAN-SHAN dislocations Avere observed by us in several districts, especially in the Kummduk valley, lying to the east of Altyn-Artysh. There can be no doubt that the seismic movements appearing in the upfolding of the Artysh strata, which are referred by BogdanoA'ich to the pliocene epoch, were continued in later formations, and persist to the present day (more on this in the detailed report). In the district, already mentioned, such movements led to the almost complete destruction of from ten to twelve populous villages which, standing on well-watered loess terraces, occupy the richest and most productive tracts in the neighbourhood of Kashgar. The epicentrum of the seismic forces nearly coincides with Artysh-bazar, and the destructive effects of the earthquake waves, radiating from this point, made themselves felt even in the city of Kashgar and its environs. We were able to follow these movements, somewhat weakened but still very destructive, over a wide area, up the Toyun valley, in the JVIaydan-Gess valley, fartlier east in the Kurumduk valley, and later even still farther east. During our stay in Kashgar, more or less violent and destructive underground shocks were of such daily occurrence that one grew accustomed to them. In the Toyun valley Devonian fossils were found, partly in the places already visited by StoHczka and Bogdanovich, north of the Chou Tcrck grazing ground (not village), jmrtly in other districts. On the whole, however, the finds were FOSSILS 105 not great, although we pushed northwards far be- yond Yakub Beg's old fortified post of Chakmak. On the other hand, we ascertained the presence of erupted basaltic rocks in the zone of the most violent dislocations, in the schists and also in the sandstones, which are embedded in them, and are, by Bogdanovich, referred to the tertiary epoch, all at a considerable distance to the south of the localities, where they had been found by Bogdanovich (Suyok valley) and by Stoliczka (Chakmak) ; for further details see below. During the following excursion our collecting work was more fruitful. The route led by Altyn-Artysh northwards, up the vast tertiary basin of Argu, which was formerly flooded by a lake and still shows well-preserved terraces. It was approached through a narrow portal cut into a rampart of conglomerate 650 ft. (200 m.) high, and was quitted by a similar outlet, leading into the Tangitar ravine, by which are reached the basin-shaped expanses of the former large lakes Tegermen and Arkogak, which follow from west to east, one a stage higher than the other. StoUczka found some fossils north of Yakub Beg's barrier fort of Tangitar, consequently north of the gorge through which the river forces its way. The places where we made our great finds lie, some a little to the west of the old fort, some south of it. The fauna is probably partly Devonian, partly carboniferous. Most surprising is the thickness of the con- glomerates immediately before and beyond the lOG SOUTHERN EDGE OF TIAN-SHAN place, where the Tangitar river breaks through the carboniferous limestones in a romantic gorge from fifty to sixty-five feet ^^dde, which winds for some two and a half miles between almost vertical walls, and escapes to the rocky Tangitar cirque. Although partly worn away, the conglomerates, in which are often embedded huge boulders, here rise in places up to 1,150 ft. (350 m.) above the level of the valley on the limestone walls, and project, as mighty buttresses far into the plain. Beyond the gorge, ancient valley terraces {Ldngssttifen) are seen in these conglomerates, which are over-lain by loess to a considerable thickness. In the vast Tegermen basin, where no water now flows, except a narrow rivulet, deposits of shingle are of such enormous thickness that they partly hide the foot-ranges of the mountains, to such an extent that only a few of their cones and domes rise like islands above the overlying drift. In the left scarp of the basin Herr Keidel found upper carboniferous brachiopods, and in a narrow gorge Devonian corals. The bed of the extensive Arkogak basin, formed by level accumulations of rubble, is reached by a breach in the low range and over a broad sill of the soil. We followed the second basin a long way in a north-easterly direction. By a lateral valley, branching off to the east and draining indirectly to the Kurumduk river, access is gained to the extensive Bash-Sugun pastures of the Kirghiz. A bed of coarse, white limestone, containing a mass of excellently preserved fossils, was found BRACHIOPODS 107 in the limestones of the ramparts, enclosing the Sugim valley, which are of very diversified character and exhibit complex stratified con- ditions. Here we were able to obtain a rich upper- carboniferous fauna of brachiopods, representing about fifty species in several hundred specimens. Bash-Sugun was already known from the fossils found by Stoliczka (E. Suess, Contyibutions to the Stratigraphy of Cenf?ril Asia). But whether the locality, exploited by us, is identical with Stoliczka's seems doubtful when we consider that this explorer found only a few, apparently lower-carboniferous fossils in this place, whereas such an accumulation of organic remains as occurs at our " storehouse " could scarcely have escaped the trained eye of the distinguished naturalist. On the further journey to the south-east through the Sugun valley, which here contracts and forms a series of small, cauldron-shaped expanses con- nected only by narrow passages, we noticed ex- tensive intrusions of basaltic rocks in the form of domes, but also in dykes. Shattering breccias and conglomerates also occur, while the surrounding Hmestones have been greatly metamorphosed. The outbreak of basaltic rocks, ascertained by us as occurring in this region, as well as in various localities on the extreme southern border of the Tian-Shan, show that their intrusion is not confined to the line of fault, assumed by Bogda- novich to exist on the northern slope of the Kok-tan range {Trudu, etc. p. 72). Such basalts were found by us not only at the already- 108 SOUTHERN EDGE OF TIAN-SHAN mentioned places in the Toyun valley (see p. 105), but also in the farthest off-shoots of the mountains at Tafifh-Tumshuk, not far from IMaral-bashi. Through a breach, a hundred feet wide, the Sugun river continues its easterly course and debouches to a spacious valley, about two miles broad, which in its turn again trending to the south- east, falls into the Kurumduk river. It should be noticed that the delineation of this region on all the maps known to me, especially the hydrographic system between the plateau of Tegermen and the Sugun district, and the continuation of this river system through the Kurumduk and right out into the Kaldy-Yailak plain, does not even remotely correspond with the actual facts. From Ayak- Sugun, which lies at the confluence of the already- mentioned lateral valley with the Kurumduk, we made our way to Sugun-Karaul. The route from the Kurumduk valley (which was itself traversed only for a short distance) to the plateau at the southern foot of the mountains, leads for o\'er sixteen miles through narrow winding defiles across that section of the tertiary highlands, con- sisting of soft clays and marls, which have been subjected to the most profound shiftings of level. Owing to tliis fact it has been shattered, and, for the most part, buried beneath its own debris to an extent tliat has elsewhere been rarely observed. He fore the marls lies a thick zone of very fine, hard conglomerates, whicli extend for two miles into the desolate, high Kaldy-Yailak plain. 1 was now obhgcd, during the prevalence of the THE GESS VALLEY 109 severest winter weather, to undertake the long and difficult journey to Tashkent. Here I had to procure the instruments and photographic materials, long before ordered in Europe, to complete our equipment, and also to direct a second Alpine guide by telegraph from his home to Kashgar, where there is no telegraphic communication with Europe or even Turkestan and only a very defec- tive and tedious postal service. I took the route over the Terek-davan (Irkishtam — 23° C, Kok-su — 28° C). As this route has already been several times described, and most recently by Futterer {Through Asia), I may here pass over the observa- tions that I made on my journey, though they present many points of interest. During my absence Herr Keidel occupied him- self with the investigation of the loess deposits in the Kashgar-daria valley, and also made an excursion to the southern border of the Kashgar basin. The way led through Boruk-tai to Tash- malik ; a rich fossil fauna was discovered south- west of this place. From Tash-malik Herr Keidel went on to the Gess valley, which he followed up to Ak-chiu, where he made a collection of fossil plants of the Angara series in the coal-beds, worked in a primitive fashion by the Kirghiz. The return journey was made through Eski and Yangi-Hissar. A second excursion to Bash- Sugun, undertaken towards the end of February, had for its object the completion of the geological collection by a study of other levels in the local limestones. In the finds here made, various stages 110 SOUTHERN EDGE OF TIAN-SHAN of the carboniferous age are represented. Some specimens of the species of the Bash-Sugun fauna found tlieir way to Calcutta, where they were recognised in the Geological Survey Office of India as corresponding with the Productus Lime- stones of the Punjab Salt-range. CHAPTER IX THE SOUTHERN MARGIN OF TIAN-SHAN BETWEEN KASHGAR AND UCH-TURFAN At the beginning of March I was back in Kashgar from Tashkent, where, through the kindness of His Excellency the Governor-General, two valiant young Kossacks were placed at my disposal as es- corts. At last, after much trouble, diverse bad incidents and unpleasant delays, the new Alpine guide, Sigmund Stockmayer of Neukirchen in Pinzgau (Salzburg), arrived with a portion of the instruments and materials that had been ordered. After completing all the other troublesome pre- parations, the hitherto very cold weather ha^dng also become a little milder, a start was at last made on April 14th, 1903, for a fresh expedition to the highlands. Beside myself and Herr Keidel, the party now consisted of the two guides, Kostner and Stockmayer, the preparator Herr JNIaurer, the two Kossacks BesporodofF and Simin, with the corresponding accompaniment of Sart attendants and horse-keepers. Later we were joined by ChernofF, another Kossack, who had been one of Sven Hedin's assistants. All the military posts along our route had previously received due in- formation from the Chinese authorities in a way 111 112 SOUTHERN MARGIN OF TIAN-SHAN that deserved my thanks. They also gave me \\Titten instructions and a policeman ("Beg") for a part of the way. Through the kindness of His Excellency N. F. Petrovsky, Imperial Russian Consul-General at Kashgar, to whom I am greatly indebted for much help, the Russian "Aksakals" in Uch-Turfim and Ak-su received notice of my approaching arrival. Though my residence in Kashgar had not been too pleasant, I still parted reluctantly from persons, whose kindly advances and disinterested support had stood me in good stead on many trying occasions. As the raw weather and the masses of snow, lying on the uplands, did not yet allow us to push into the high valleys, I decided to journey at first for several weeks as closely as possible along the southern escarpments, in order to study their geological structure, as on this particular section of the Tian-Shan next to nothing was known. The route once more necessarily lead through Altyn-Artysh and Tangitar to Bash-Sugun. Still, the repeated visit to this locality was not thrown away, since it lead to the discovery of carboniferous Permian deposits. My intention was to make my way over the Kara-bcl passes into the Aiktyk valley, whose southern bank is formed by the " Kok-kya range," as it is called by Severzoff, why I do not know. Thence the descent could be made into tlic narrow ravines of the Kok-shaar river, cut between the aforesaid range and the section of the southern border-range also named by THE KHIRCxHIZ KARA-JIL 113 SeverzofF the " Bos-aidyr Chain." The project, however, was thwarted by the stupidity or the ill-will of the " Beg," appointed by the Chinese authorities to accompany me. I should here point out that tlie terms Kok-kya and Bos-aidyr, applied to mountain ranges, are unknown to the natives along the southern border. From Bash-Sugun the way led east and north- east in narrow gorges through light-coloured, coral- bearing limestones, then along the southern border of the great mountain range across the surface of the plateau, where the outer range rises only in isolated crags above enormous, heaped up masses of mountain drift, like cliffs out of the sea. At the Kirghiz settlement of Kara-jil these crags of the front range reach a height of only fifty to sixty feet, and consist of interstratified light and dark limestones, the latter of which yielded a rich upper carboniferous fauna. The place must not be confused with the Chinese military post of like name, which lies farther north in the Aiktyk valley. Of this district, which was subsequently traversed by the ex- pedition, the maps give an altogether inadequate representation, wliich will in many respects be coiTCcted and completed by our route surveys. From Kara-jil we travelled in an east-north- east direction along the foot of a limestone ridge, 1,600—2,000 ft. (500—600 m.) high, through the loess steppe, where the outer range, buried in the drift, may still be followed in island-like fragments for a long distance. Then we turned 8 114 SOUTHERN MARGIN OF TIAN-SHAN a little south to the Kh-ghiz settlement of Jai- teve (tube?), on the shore of the saline lake Shor-kyl. At this point the expedition touched Sven Hedin's route of the year 1895, but again immediately diverged in a north-easterly direc- tion, and made its way into a valley, which intersects the mountain range at an acute angle. This valley is of typical form, narrowing in its upper course to the shape of a ravine and becoming a perfectly developed transverse valley, sunk in hard strata of limestone, grauwacke and phyllite-like schists. We found this typical erosion- \'alley (Apatalkan) and its secondary valleys water- less, and only at the valley head came on a feeble stream, issuing from the snow-fields lying there. The origin of such a valley, and of other erosion- valleys, traversed by the expedition shortly before and afterwards in the journey to Uch-Turfan, cannot be explained satisfactorily by the periodical streams which flow through them only for a small part of the year, but much rather points to great climatic change. The route led between mountains, whose steep sides, in consequence of the slope of the strata towards the north, were turned towards the \ alley, up to the Apatalkan pass, nearly 10,000 ft. (3,000 m.) in altitude; then down through the snow-clad northern Apatalkan valley (Uyuk- Apatalkan), of trough-like profile, where, notwith- standing the advanced time of the year (April 22nd), we re-entered in the region of winter and encountered ^^olent snowstorms. The ramparts THE KOK-SHAAL VALLEY 115 of the valley consist of a regular folded mass, already much worn down, of phyllite-like schists and grey-blue grauwacke, both of very varying appearance. This horizon of great thickness can be followed a distance of twenty-five or thirty miles in the Kok-shaal valley. Kok-shaal is the name, given in general to the upper course of the Taushkan-daria by the population living on its bank. At the mouth of the Apatalkan valley, the Kok-shaal valley is already a mile or a mile and a half wide, and looking back one sees, only a little farther to the west, the river leaving the ravine, whence it issues through a door-shaped opening, and then rushing in a majestic curve into the distance. To my regret time did not suffice for in- specting the ravine, from which this river issues, especially as it had never before been visited by any expedition. It is significant that in the Kok-shaal valley, as in all the less recent Tian- Shan valleys, we at once met with unusually large masses of conglomerates, which constantly accompany the course of the river, irregularly overlying the old schists, and in their turn over- laid by more recent conglomerates, etc. At the locality of Abdul-kia, ahas Alep-turga, about 8,200 ft. (2,500 m.) in altitude (these, like most of the following names, are not to be found on any of the existing maps) — the Kok-shaal river should be crossed, but, owing to the strength of the current, the passage proved impracticable. We had to travel along the limestone range on 116 SOUTHERN MARGIN OF TIAN-SHAN the right bank, against the bluff walls of which the river beat for a long stretch, and hence induced us to take our way through defiles of the surprisingly eroded limestone ridges. Thus, travelling for some time near the border of the mountain mass, we again reached the main valley, where the river, now divided into several branches, could be crossed. Already at Ak-kia the view up the river had opened out to the fine chain of snow-clad, rocky mountains belong- ing, at least in a purely orographical sense, to the so-called Bos-aidyr chain {vide p. 113), for the separation of which from the continuous wall of the Kok-shaal-Tau, however, I can find no satisfactory boundary-line either from a geo- logical or an orographical point of ^^ew. The route over the wide, slightly inclined steppe- terraces of the northern bank was now open to us. Passing the great Kirghiz settlement, Kara- bulak, with one of Yakub Beg's dilapidated forts, we crossed a plateau of consolidated pudding-stone {Deckcnschottei^), rising gently towards the north- east, and approached the foot of the rugged precipitous mountain rampart at the Aul of Chagash-gumbes, about 8,000 ft. (2,450 m.) in altitude. The secondary range of the Kok-shaal- Tau, which here attains a height of ciixa 11,500 ft., though it must, from a geotectonic point of view, be considered as separate from the higher ranges behind it, should, according to the maps, belong to Severzoff's " Bos-aidyr range." The native Kirghiz call it Markesh-tagh, The drift mounds THE BOS-AIDYR CHAIN 117 at its base contain no crystalline material, this front wall being composed of limestones, calcareous slates, and very dense, highly metamorphic, bright- coloured clay-slates and sandstones, which dip now to north-north-west, now in the reverse direction. Crystalline fragments (granite and syenite) which are brought down in the beds of several streams that break through the front range, appear to be derived from the higher ones behind ; but lower down the valley I found in more recent drift, which there covers the base of the hills to a great depth, crystalline material (large granite blocks) in places, where no passage leads back through the lower range. They were doubtless carried hither by ice from the inner recesses of the mountain range. These are not the only traces of former glacial action which we found in the Kok-shaal valley ; on the right as well as on the left bank such traces were proved to exist, though not very abundantly. The section of the Kok-shaal-Tau, to which the name of the Bos-aidyr chain is given consists of several nearly parallel chains, of which the hinder, more northerly, is much higher and possesses more variety of mountain shapes than the range in front. Its summits, clad with nevd, are very steep. Here there is displayed a characteristic of the configuration of the Tian-Shan, which I had already observed and afterwards often con- firmed — that of its parallel structure. SemenofF, the most acute explorer who has ever visited any part of the Tian-Shan, long ago directed attention 118 SOUTHERN MARGIN OF TIAN-SHAN to this law, which finds such abundant expression in the configuration of this giant range. The Kok- shaal-Tau shows, on the whole, a gradual ascent from west to east, as far as the neighbourhood of the Bedel pass, where a sinking takes place. At the Kirghiz settlement of Kysyl-gumbes, about 7,500 ft. (2,300 m.) above the sea, which owes its name to the red colour of the loess surface, a result of the decomposition of the bright red ("kysyl" = red) limestones, conglomer- ates, and sandstones which form the steep, finely peaked ramparts of the valley, and to the many Kirghiz burial chambers ("gumbes") which dis- tinguish the region. An excursion to the so-called Bos-aidyr chain was arranged, and, for a better insight into its formation, one of its lofty peaks was to be climbed. But, to my regret, this purposed excursion came to nothing, owing to a phenomenon, which, regularly, during the long time the expedition was engaged on the south side of the range, made obser\-ations exceedingly difficult and partly impossible — namely, continuous and unusually dense fog. The fog was, in this early part of the year, almost denser — at all events it was much more continuous — than we find it in the Alps in Novcm})er (a surprising phenomenon in this southern region, distinguished by the dryness of its climate) ; for weeks at a stretch there was IK) clear weather. The explanation lies in the intense licating of the loess soil during spring- time. Tiiis, at some hours of the day, whirls DUST AND MIST 119 the fine dust aloft, and, even when the wind is cahn (to say nothing of the strong winds which often prevail), drives it in commotion up to the higher strata of the air, where it remains floating. In spring, in consequence of the melting of the snow, the mountain slopes give off much moisture by evaporation, and these vapours are condensed on the fine floating dust-particles to mists which neither fall nor yield. In April and May we had often a cloudless sky, but seldom a clear atmosphere. Our photographic work had to be discontinued often for many days — a serious loss. Over much that was worth observing on our route, lay an impenetrable veil. In the limestones which are mainly concerned in the formation of this front range, we found beds containing corals, whose identification will perhaps help to determine the age of these strata, which, also on the right bank of the Kok-shaal, form ranges of imposing height. At the place called Ak-tala we crossed again to the right bank. Here, and even earlier, the mountain range along the bank, the Sogdan-Tau, showed remarkable development in the impressive perspective of a lofty rampart with a crest-line of about fourteen miles (twenty versts) length, almost without a gap, deeply covered with snow, and on the average about 4,000 ft. (1,200 m.) above the floor of the valley. Behind it a far higher range (again a parallel structure), with a some- what greater variety of form, and bearing small glaciers, became \'isible. 120 SOUTHERN MARGIN OF TIAN-SHAN The existence of glaciers is indicated also by the name of a transverse valley, Uch-INIusduk, which means " Five Glaciers." To this Sven Hedin has already directed attention. This great mountain range, occupying a wide area, is still altogether terra incognita. Our route along its border led us into a longitudinal valley of considerable breadtli, where there are strata of laminated, green phyllitic schists, interchanging with strata of grey sandstone in a regular, rather sharply folded formation, whose partially eroded arches may be followed for a great distance. These strata, as was afterwards proved, at different points, overlie discordant limestones, which strike obliquely across from the left bank. Also, in this now waterless region there are wonderfully perfect erosion- valleys. Near the ^Vul of Sum-tash, in the neighbourhood of which are the still unknown ruins of an ancient town, complicated foldings in the same rock-series are disclosed, and the limestones, seen below, crop up to the surface at the Kok-belys pass, which we crossed, where they contain a bank bearing bracliiopods, and lie discordant among argillaceous schists. The structure of the mountain chain onwards steadily engages the attention in con- scfjncncc of tlie magnificent disclosures of its interesting stratification, but this subject cannot be discussed in this summary report. Herr Keidel will make good this and other deficiencies in liis detailed geological description of the regions traversed. UCH 121 As we descended an affluent to the main valley we reached the Kirghiz settlement of Uch (on Hassenstein's map wrongly placed on the left bank) about 6,400 ft. (1,950 m.) in altitude, and subsequently we again struck Sven Hedin's route of 1895. Among the magnificent, gorge-cleft mountain ramparts at Uch, where, from a height I climbed, the three parallel chains of the Sogdan- Tau were visible, a collection could be made of the fine rich fauna of the upper carboni- ferous formation, here existing in two distinct horizons, lying in slight discordance. This rock sequence can be followed far to the east. Here for the first time we discovered foraminiferous schists (bearing Schivagerinas), which from this point steadily accompanied our route along the southern slopes as far as the Khalyk-Tau. The extraordinarily wide area, on which these fora- minifera, which characterise the uppermost car- boniferous formation, are distributed, is a new fact in the stratification of Central Asia. In the continuation of our journey eastwards we constantly found magnificent disclosures of the same compactly folded system, running north-east and south-west, especially fine in the Aul of Shinne. Immediately thereafter, toward the Kara- turuk gorge (this on Hassenstein's map is marked to the east instead of to the west of the pass), the river dashes impetuously against a cape-like pro- jecting spur of the mountain range, and makes it necessary to cross the rocky pass of Shinne- davan, in the neighbourhood of which, owing to 122 SOUTHERN MARGIN OF TIAN-SHAN the oblique cutting through the folds, highly interesting geological disclosures can be seen. Again there appears the rock-series of the horizon of Uch, discordant under schists, and, farther on, old stratified conglomerates overlaid with black limestones and reddish clay-schists, a series which accompanies our route over the next pass and onwards through a valley into the plain, where, in the neighbourhood of the Aul of Sary-turuk, it is replaced by hard, dull-coloured, crystalline limestones, which now, in a series of banks of immense thickness, form the mountain wall over Ak-kia to the richly cultivated Aul of Safar-bai, about 6,000 ft. (1,850 m.) above the level of the sea. The much higher, snow-clad mountain range, which accompanies the river on the left, remained, during this long journey through the Kok-shaal ^'alley, which is often as much as two and a half miles (four versts) in breadth, in consequence of the thick fog {vide p. 118 scq.), almost continually invisible. The river-bed is several times constricted to a breadth of 000 to 900 ft. (200—300 m.) by cape-like projections, where the ridges, set obliquely to the axis of the main valley, have been much eroded by the action of water and wind, but the average l)rcadth of the valley is not diminished. At tlie Kirghiz settlements of Kara-bulung on the riglit bank and Bulung-turuk on the left bank the river makes a great bend, and bears henceforth the name of Taushkan-daria, or is even called simply Daria. There, from the wall on the right shore, which at the bend swings far to the south- STRANGE EFFECTS OF CORROSION 123 west, low ridges of fossil-bearing limestones run forward to the river. Beyond the Aul of Koshe- bashe, where the loess-plain on the right bank showed rich cultivation, the river presses suddenly against this bank and reduces it to a mere strip. Even this A^anished at last, and then our route, since it was found impracticable to cross to the low left bank, led over a projecting cliff of marble- like limestone to the pass of Denge-davan. In the ascent I found the cliffs, up to approximate heights of sixty-five feet (twenty metres) worn by water — one sure sign of many I observed in the main valley, either that the river has deepened its bed, or that it formerly had a larger flow of water, or that both of these suppositions are true. On the eastern side of these cliffs the rock-walls are, owing to aeolian corrosion, pierced high up with thousands of little holes, a phenomenon which may be observed at many places in the Kok-shaal valley on the windward side of the cliffs, but nowhere else so finely as here. In the neighbourhood of the Aul of Konganishuk-Yangyll, again, a low ridge juts forward from the chain bounding the main valley on the right to the river-bed, or, indeed, into it ; this is through erosion, partly by water and partly by wind, di\'ided into separate small rocky islands, two of which rise in the middle of the bed of the river. This row of cliffs, which the Kirghiz call Mai-tewe (tube ?), consists of coarse, dark- coloured limestone conglomerate, interstratilied with sandstones ; the limestone fragments contain a rich fauna, belonging to the upper carboniferous 124 SOUTHERN MARGIN OF TIAN-SHAN formation, of which we collected specimens. To judge from the flat depression of the strata and from the arrangement of the folds, this horizon mioflit be followed far to the east and south-east ; it was. in fact, again met Avith farther to the east. At Bash-chakma, about 5,600 ft. (1,700 m.) in altitude and at Tagh-tumshuk, the commanding mountain range on the right bank is developed on a great scale (here also three parallel chains could be observed), and by its height and arrangement it forms a remarkable shelter for the region, which now at length (at the end of April) showed tlie first green of spring and the charming hues of blossoming peach- and apricot-trees. There, on a sharply projecting mountain spur, could be observed complicated disturbances, several flexures, faidts, and ruptures, which could be followed away to the east and north-east in a complex stratum of slabby limestones, quite void of fossils, loose sandstones, and red-brown quartzites. Farther on, at tlic Aul of Kum-bulung, however, only these sandstones appear ; they form great arches in tliick layers, and the products of their decom- position have transformed the region far and wide into a desolate sand-desert, from which a !)it of soil for cultivation can only with difiiculty be wrung. The protruding dark-coloured lime- stone ridge of Ot-bashi-tagh sets the first limit to the driving of the sand, at a bend of the river. Under its shelter the diligence and skill of the population (from here onwards exclusively Sarti.'ui) ]ia\c turned the region into an incom- A PICTURESQUE CITADEL 125 parably luxuriant garden-land, which extends to the town of Uch-Turfan, about 5,000 ft. (1,500 m.) in altitude, and beyond it. These dark-coloured limestones accompany the route in crowded folds, frequently with remarkable bendings of the strata ; in them also there is an upper carboniferous fauna, of which Herr Keidel collected fine specimens. A projecting spur, formed of these limestones, bears on its summit the picturesque citadel, command- ing not only the town and its handsome walls, built on the Vauban system, but also the garden- like region far and wide. This cliff consists partly of great banks composed exclusively of Productus and Spirifh% from ^ to 4f in. (2 — 12 cm.) in diameter. CHAPTER X TO THE KHALYK-TAU AND BACK TO UCH-TURFAN In accordance with information obtained, I had to put off my purpose of penetrating from Uch- Turtan at this time into the transverse valleys of the left mountain range, since in those valleys there was at the time snow, of course, but no fodder for the horses, and the helpful Kirghiz had not yet come up. I therefore resolved to go farther east into the Khalyk-Tau, which had never before been visited by any explorer, and whose transverse valleys, opening directly to the south, might be expected to offer more favourable conditions. Our way would lie across Ak-jar, Shah-shambe, and Tjaggcrak to the town of Ak-su. Along this route we could at length (in the first week of May) begin gathering the first spring flora of the steppe. At Ak-su we had to stay for several days to complete our number of attendants and of horses as well as to meet the Chinese authorities. We left the interesting town on May 7th by the old caravan road to Bai, and crossed, between Kara- Yulgun and Tugarakdan (according to the incorrect rcprcscntalion of llie Ibrty-verst map, it would lie between Jurga and Yakka-aryk), the tertiary 126 BAI 127 mountain range of the Chul-Tau which trends west-north-west, the route lying in an obhque cutting through its fine system of arched folds. Bright-coloured banks of sandstone and clay marls, bearing gypsum, overlaid by slabs of conglomerate, compose the mountain range, whose structure is far less complicated and whose appearance therefore has less variety of form than the Topa-davan range in the north-west {vide p. 99 seq). The crest elevation of the central portion is of course higher than there, but in its eastern chains between Jurga and Yakka-aryk and farther east it is much weathered and already reduced to insignificant dune-shaped swellings ; by its decomposition it has furnished the material for a considerable rise in the level of the plain eastwards. This plain reaches its highest point at Chakh-chi, about 4,760 ft. (1,450 m.) in altitude, and from this spot it falls away towards the Musart-daria. The visit to the town of Bai was of doubtful value. The information there obtained with great difficulty from the Chinese authorities con- cerning routes and conditions in the Khalyk-Tau turned out to be incorrect. It appears that nobody there is well acquainted with that almost inaccessible mountain region. The forty-verst map here leaves us completely in the lurch ; between Bai and the mountain range it presents nothing but a blank, and what of Khalyk-Tau is elsewhere shown proved wTong. As the topographical sketches, taken during the journey are not yet worked out, our course, in the absence 128 TO THE KHALYK-TAU of a topographical basis, could only be made clear by very detailed explanation ; I must there- fore reserve a minute description of this portion of our journey and, in this provisional report, only state the most essential points. But it must be mentioned that the direction and course of the rivers, given in tlie forty-verst map do not agree with fads : The Kapsalyan river, the most impor- tant of the rivers of this range, on issuing from its narrow, ravine-Uke ^ alley, takes a direction towards the soutli-west and west along the southern slope of tlie range, and the river which issues from the valley, wrongly named Kasnak-su on the forty- verst map, but in fact bearing the name of Terek, does not flow into the Musart-daria but into the Kapsalyan, whicli for its part only reaches that stream in the neighbourhood of Chakh-chi. Finally, Bai is at a much greater distance from the foot of the mountain range, than it appears to be on the forty-verst map. Our route passed from Bai in a north-westerly direction over Terte and Uskim through the desert to the small kishlak of Masar-Yakub, which is still at a considerable distance from the edge of the mountain range. Here it turned out that our next destination, the Tilbichek side- valley, cannot be reached directly, since its lower course forms a ravine, inaccessible for beasts of })iirdcn. We had to bear westwards and traverse the desert- valley of Kah-agach, which is cut into recent crystalline conglomerate. We crossed a small pass and, going along a hollow, running MARL TERRACES 129 soutli-west, reached an elevated plain, and then went to the village of Jam-kuluk, about 5,500 ft. (1,600 m.) in altitude, situated on the plain at the foot of the first chain of older conglomerates, near the mouth of the Kapsalyan river. The route upwards in this valley had an easterly and north- easterly direction ; the valley is sunk between rugged, high, red conglomerate ramparts (of which more hereafter), and is divided into three small basins (old lake-bottoms), connected with one another by door-like openings, cut in the enclosing rock-walls. Thus we reached the region of tertiary, bright-coloured clay marls, which are steeply folded together with the hard, violet-red conglomerates, but, already very much destroyed, are for the most part now only found at the foot of the steep conglomerate walls, 650 to 1,000 ft. (200—300 m.) high. Over these marl terraces, at one side of which the river flows in a deep gorge, we travelled up the valley to the Taranchi settlement, called Musulyk, about 6,000 ft. (1,820 m.) in altitude, lying on a raised sill, deeply pierced by the river. Thence we went to the junction of the Terek with the Kapsalyan, and approached over a boulder- plain the spot, where the latter issues from between the steep walls of the lofty limestone range. We then left the basin of the river, and crossed the broad water-shed between it and the Tilbichek river-basin through a defile about seven miles (ten versts) long, which, folloAving the strike of much decomposed, variegated banks of marl, displays wonderfully \'aried and bright-coloured 9 130 TO THE KHALYK-TAU strata, especially where the red conglomerate walls, with their boldly peaked crest-Hne, overtop these steep, jagged marls. Over a steep slope Ave descended into the wide plain of the Tilbichek \-alley, whose door-hke entrance into the gorge of its lower course was at once visible behind us to the right. In the middle portion of the Tilbichek valley, the soft marls are almost cleared away, and the conglomerates alone form, by their strike, tlie ramparts of the valley. Since they dip steeply towards the south-east, the orographic right wall is sloped steeply enough, but the left wall presents towards the \'alley perfectly perpendicular precipices, forming a wall twenty versts long, sheer by the plumb-line, red, crowned with odd peaks and pinnacles, a sight such as is seldom seen elsewhere. A small Taranchi settlement in tlie valley is called Sukhun, about 6,400 ft. (I,{)o0 m.) above the sea. Thence we penetrated deeper into the Aalley, first going north-eastwards, then north, where tlie still-preserved parallel folds of the steep, \'ariegated marls, rising in serrated crests behind one another, together with the conglomerate walls, group themselves into tlie most peculiar shapes. In this geological horizon lie three basin-shaped widenings, which are coi nice ted witli one another only by door-shaped openings, thirty to forty feet (ten to twelve metres) wide. Tlirougli tlie last opening access is gained to a region of light grey, fine, sandy con- glomerates, which pass into actual sandstone and enclose clay-coal schists {Lcttcnkohknfichkfcr) IRON-SMELTING 131 with impressions of plants, and higher up there are dark-brown, poor clay-ironstones and grey, dense limestones. Far behind in the valley a Taranchi, living in a cave, is occupied in smelting iron. The main valley here branches and leads towards the north-west, over lofty terraces, clothed with Alpine meadows, to a pass ; the main branch, however, leads northwards as a narrow gorge, with almost perpendicular walls of dense lime- stone, between which flows a raging torrent. To Herr Keidel's attempts to penetrate deeper into the ravine, and so from the limestone belt to reach the crystalline, insuperable obstacles at once presented themselves. The second excursion into the mountain range was farther west, through a narrow, door-like and difficult breach in the red conglomerate walls to the Kepek-chai valley, where the region of the bright, grey, sandy conglomerate, sandstone, clay- coal schists, and clay-ironstones, already mentioned, is reached much sooner than in the Tilbichek valley, since this system of strata runs about north-east and south-west. In the background of the valley the most complicated forms of stratification, inclined folds, contortions, over- shdings {iyberschicbungen), etc., accompanied by chaotic destruction of the rock-series, can be observed at magnificent exposures. These disturbances, after more minute examination of the observed conditions, may probably prove to be associated with the already-mentioned disturbances in the southern Musart valley {vide 132 TO THE KHALYK-TAU p. 93scq.) since tlie crystalline rocks strike across from that region and go somewhat deeper into the mountain range in contact with the sedimentary rocks. The red conglomerates and tertiary marls, being much more recent, were not involved in this movement. We ascended the Busai-tash pass, about 9,200 ft. (2,000 m.) in altitude, leading into the Tilbichek valley, and thence over extensive Alpine plateaus, about 800—1,000 ft. (250—300 m.) higher, which spread out between the two valleys, named and tlie Kapsalyan valley. These plateaux pro- vide a fine view of the snowy chain of the Central Kluilyk-Tau. The highest peaks lie to the north and west ; towards the south and east there is a gradual falUng away. Turning back towards Musulyk, Herr Keidel attempted to penetrate into the Kapsalyan valley, but was baffled, the narrow gorge being completely filled with water. Only in winter, if the river is low or is frozen, the Taranchi penetrate into the valley and carry away fir-wood. Herr Keidel, in order to obtain an insiglit into the structure of the range, now resolved on the ascent of a high peak, about 12,000 ft. (3,()00 m.) in altitude, standing between the Terek and the Kapsalyan valleys, while I penetrated into the Terek valley, which likewise has the cliaracter of a mucli-winding ravine, but, ncxcrtheless, proved passable. From a bivouac aljout 8,000 ft. (2,450 m.) in altitude, midway up tlie gorge, I was fortunate enough to reach its head, about 0,700 ft. (2,950 m.), where it divided LIMESTONES AND SCHISTS 133 into two clefts, running up to the main water-parting ridge. I could thus traverse the whole series of the sedimentary rocks lying on the outer border, the crystalline belt, forming the middle course of the valley, and the limestones and old argillaceous schists forming the valley-head. Thus I was able to collect a complete sequence of the rocks. Just as in all other transverse valleys of the Central Tian-Shan, so also in the Khalyk-Tau, it is not the crystalline rocks, but the limestones and argil- laceous schists, which form the highest and most central portion of the range. Here these lime- stones and schists on the whole strike east and west, with slight deflections towards the south and north. These conditions, however, according to observations already made on the JNIusart pass, could not be expected otherwise. In the crystalline rock of the Terek valley, remarkable disturbances, inclined folds, violent compressions and overslidings, etc., were noted. Even far up the valley, but especially at the entrance into the Terek gorge, at the little settlement of Bom-khotan, there exist Sc/iwagcri?ia-hearmg limestones, which interchange with plant-bearing schists ; a little farther down the valley, after red sandstone there follows a belt of porphyry between the former and the frequently mentioned grey sandstones and conglomerates. It surprised me to find in this southern valley, opening to the south, the features of a narrow transverse valley of the northern limestone Alps : terraces with Alpine meadows, and on steep, rocky slopes forests of pine, which extend into the 134 TO THE KHALYK-TAU recesess of the ra\dne and, on the valley-terraces ( Thahtufcn), form dense permanent forests ; a main stream, well supplied with water by many brooks, flowing from genuine Alpine side-valleys ; splendid snow-clad, rocky peaks. At the head of the valley, where, as before mentioned, it diA^des into two narrow clefts, no glaciers can be formed, but there are small glaciers at the heads of the side-valleys, which widen out into the form of cirques. At the mouths of some of these valleys, though much has been washed away by floods of the stream, there are still considerable quantities of moraine debris, piled as evidence of former extensive glacia- tion. The whole length of the Terek valley amounts to about thirty-five miles (fifty versts) ; at a sliort distance from its head it di\ddes into two brandies : the one, running north-westwards, is called Ya-konash ; the other, rimning mainly nortliwards, is called Jan-kasnak. From this latter name the appellation of Kasnak-su, which is given in tlie forty- verst map to the wliole valley, is perliaps deri\ed. I repeat that the inhabitants of tlie region denote the whole valley simply by the name Terek. Our return-route from the Klialyk-Tau lay close to the skirt of the mountain laiigc : first upwards, in the lower course of the Tcrck \ alley ; then across the lofty terrace of Yar- jilga, wliich apparently closes the valley, down to the wide plain of Karabag, which extends be- tween the longitudinal course of the IMusart-daria aiid the foot of the range. The transverse valleys at this part ol" the Khalyk-Tau are not inserted, ^ CKOSSING THE KISHLAKS 135 much less named, in any map. In the order from east to west their names are : Yagus-tal, Kysyl-tal, Tutuk-tery, Cholok-su, Alagyr, Tyukiir-myt. I found all of tliem, notwithstanding their southerly exposure, very snowy, and in some there were considerable glaciers. They are terminated by a mountain ridge, running north-west and south-east, which crosses obliquely fi'om the Musart valley, and for this reason the most easterly are short, while, in general, their length increases the farther they lie to the west. The most important of them is the Tutuk-tery valley, from which a great mountain-river flows. Most of these valleys contain pine forests, in which the inhabitants of the widely scattered kishlaks of the high plain burn charcoal. Our route lay across the kishlaks : Kish-talga, Karabag, Kok-kya, Little Karabag, Kyssalik, and Chapta-khanne — continually along the edge of the mountain range, which falls away towards the high plain in walls about 4,000 ft. (1,200 m.) high. Along the foot, however, there is a belt of tertiary deposits, more or less destroyed and carried away. After crossing the INIusart-daria, which here, at Cliapta-khanne, presses quite up to the mountain wall, the road leads without intermission over old moraine-soil, o\'er-grown with verdure, across a number of moraine-ridges, running north and south, cut by little cross-valleys ; on these ridges lie massive erratic blocks {vide p. 97). From this enormous accumulation of moraine debris the route descended steeply to the first Chinese picket. 136 TO THE KHALYK-TAU Koncshar, at the entrance of the southern Musart \'alley, where we made our appearance on May 2.'3rd. On tlie assurance that, in accordance with the orders of the Chinese authorities at Ak-su, wliich had been forwarded to the Sartian " Begs," I sliould find fodder for the horses and provisions for the company ready at all stations, I resolved to pay another visit to the southern Musart valley. My principal aim was to penetrate from the last picket, Tamga-tash, to the unexplored Karakol \'alley, which extends thence to the north-east, and become acquainted with the very important glacier of tliis valley, probably one of the largest in the Tian-Slian, and also to explore its sur- roundings, which consist of completely ice-covered mountain chains of gigantic height, whose con- nection M'ith the great main ranges still remains o])scurc. The extensive glacier background of the Turpal-che valley was also to be investigated. To my deep regret these plans could not be carried out, since the " Begs," notwithstanding the orders sent them, left me in the lurch. I first made, from Tamga-tasli, a tour of in- spc(ti(jn to tlie great Karakol glacier, w^hereby it was establislied that this glacier, hke the Inylchek glacier, is o\'erspread with a great coating of moraine dc'-bris, the passage of wiiicli, to a length of only lor two and a half miles (four versts), took much time and proved very troublesome. So far as could be seen from an elevated crest on the left bank of the valley, this debris mound extends still farther over a stretch of about seven miles (ten versts) up UCH-TURFAN 137 the glacier before free ice can be reached. This has certainly three times the length of the moraine- covered part. At the end of the glacier tongue there is a small moraine-lake. The traversing of the glacier and the investigation of its environs would have required at least a week. When I had returned from this trip to the picket, it turned out that only an insignificant quantity of fodder had been brought, and there was no more in prospect. I had therefore to hasten my retreat from the inhospitable valley, and, to my regret, to give up the investigation of this region, the most unexplored of the Central Tian-Shan. Though this trip had cost me a week's time, it had not been taken in vain, inasmuch as the geological, glacial-geological, and orographical conditions of the Musart valley, already briefly described, had this time been more fully investigated, than was possible in the flying visit of the year before. Unusually violent winds, sand-storms, and mist to some extent interfered with the work. By the route, already indicated, we returned to Ak-su, where now the Kossack Chernoff*, one of Sven Hedin's attendants, joined the expedition, and at last, after incredible difficulties, the long- expected supplies and outfit, absolutely indispens- able for the continuance of our work on the high mountain range, were completed. Uch-Turfan is more favourably situated as a point of departure for the investigation of the southern transverse valleys, since it is nearer the mountain range, and we therefore returned to that 138 TO THE KHALYK-TAU place. On the way a rich collection was made of the steppe and desert flora, now in the freshest full bloom. Being properly supported by the Chinese " Ambal " at Uch-Turfan, a well-informed and agree- able man, as well as by the Sartian " Aksakal " of the Russian Consulate in Kashgar, I could satis- factorily carry out my investigations in the hitherto quite unexplored side-valleys of the south Central Tian-Shan. The atmosphere had in the meantime become transparent, and from Uch-Turfan we had daily a clear view of the southern mountain range. The great abundance of snow, and especially the rich glaciation of these southern chains, far exceeded my conceptions. The background of the Kaiche valley, with the wonderfully bold mountain peak marked by Kaulbars with the name of PetrofF peak (not Peter peak), the magnificent Bos-tagh group, and, more than all, the mighty, completely glaciated Sabavchy chain, formed a series of surprises, considering the exposure of the slopes facing to the south, or partly to the west. CHAPTER XI THE SOUTHERN TRANSVERSE VAELEYS — THE AL- LEGED AND THE ACTUAL BREACH, FORCED BY THE NORTHERN WATERS. We left Uch-Turfan on June 11th, crossed without difficulty the Taushkan-daria, which, however, had become much swollen, and, gradually ascending the deeply intersected debris-coating of the desert, approached the foot of the mountain range. AVhat we had already learned in our journey along the southern base of the Tian-Shan was here for the first time shown in the most con- vincing manner : of the so-called wall-like descent of the Tian-Shan towards the Tarim basin, of which so much has been written and which one would expect from the representations of maps, there was, except at a few places, nothing to be seen. The veil of haze surrounding the mountain range and the sharp light of the steppe produced this felse impression on travellers, who passed along at a greater distance from the mountains. Nearly everywhere the Tian-Shan slopes away gradually towards the high plain at its southern base, in places (according to peculiarities of structure of its different parts and the corresponding course of erosion), sub- 139 UO SOUTHERN TRANSVERSE VALLEYS siding gradually in ranges of transverse spurs, whose cape-like ends project far into the desert, or in other places in the step-like tailing off of longitudinal ranges. Besides, if it is considered liow much of the outermost skirting range lies buried in the enormous rubbish-heaps of the high plain, frequently mentioned in this report, the hitherto prevailing conception of the wall- like descent of the range must be given up. In some places limestones appear as projections from the range ; at others, conglomerates and tertiary clay marls form the outermost folds. Our first stopping place was the oasis of Kukurtuk, on the little stream called Ui-Bulak, distant aljout sixteen miles (twenty-five versts) from the outlet of the Kaiche valley, and about 5,300 ft. (1,620 m.) in altitude. With the help of the Kirghiz of that place we penetrated into the Janart valley, to determine what connection it has with the alleged Janart-breach through the mountain range, and how far the representa- tions, hitherto given on maps would be confirmed. On the liigli plain, on approaching the Janart river, I found indeed a river-bed about 130 ft. (40 m.) deep, cut into the boulder deposits, and wide enough even for great floods, but not such as would indicate a powerful river. The quantity of water, flowing through it, was clearly only moderate. These circumstances alone sufficed to make me d()ul)t the existence of the so-called Janart-breach. At the entrance into the mountain valley, 7,100 ft. (2,250 m.) in altitude, where FLOOD-MARKS 141 the inevitable Sc/iwago'ina-hear'mg limestones — much compressed, however — again appeared, I was surprised to find a shallow, trough-shaped river section and a rather strong mountain stream, but no mighty river, such as must be formed by the united waters from the largest glaciers on the north side : the Sary-jass, Inylchek, Kayndy, etc. The flood-marks on the rock walls showed a height of ten to thirteen feet (three to four metres) over the then level of the river. A^^ith the determina- tion of this point my conviction was sealed, that not a drop of water flows through the Janart valley from the northern glaciers. However, I wished to exhaust the evidence of this, and re- solved to traverse to its head the valley, which has a length of about thirty miles (forty-five versts). We only rendered this undertaking feasible by moving oiu' camp forward three times. In the first third of the valley, light-coloured, dense limestones form the boundaries, and the character of the southern steppe is displayed in the midst of a magnificent rocky circumvallation. In the second third, where the valley assumes a north- Alpine character, with good meadow-spaces and fine pine-woods, it is bounded by crystalline schists and granite rocks, followed by a second series, consisting of light-coloured limestones — interstratified with dark limestone schists, and a ponderous series of dark schists and light marbles succeeds this. A thin belt of green grauwacke- schists and phyllites appears to be the outcrop of the similar rock-series, observed in the upper 142 SOUTHERN TRANSVERSE VALLEYS Kok-shaal valley, but there in a much thicker horizon {vide p. 115 scq.). Then again follow lime- stone schists and marbles, reaching almost to the head of the valley. The last third shows a gorge- like form, but quite at the end is a fan-shaped widening, where the glaciers spread out. In the highest region we found in the circumvallation of the pass a belt of granite which, at least in its southern slope, has but little breadth. The whole of the complex stratification is very steeply set; the mean direction is E. 10° N. In interstratified limestones and limestone schists Herr Keidel found a carboniferous fauna, which seems to belong to two distinct horizons. The glacier in the main valley has no great extent ; in the side-valleys, especially in those on the west, the glaciers are somewhat more extended, but are rapidly retreating. So much the more surprising are the great accumidations of moraine debris, which even at the outlet of the valley, rise liigli up against the rocky walls. In the middle of the valley, where the steep form of the rock-walls did not admit of their preservation, the ri\'er has cut its bed deep, and we saw the remains of ancient ground-moraine under the