Digitale Ausgabe

by baron a. von humboldt. In the journey I made, during the summer of 1829, in Northern Asia, be-yond the Ob, I passed nearly seven weeks on the frontiers of Chinese Zunga-ria, between the forts of Oust-Kamenogorsk and Bouktaminsk, and the Chi-nese advanced station named Khoni-maïlakhou, to the north of lake Zaïsang;on the line of the Cossacks of the steppe of the Kirghiz, and on the coastsof the Caspian Sea. At the important emporia of Semipolatinsk, Petropaulov-ski, Troïtzkaïa, Orenburg, and Astrakan, I endeavoured to obtain from theTatars, who travel about so much, — and by Tatars I understand, as the Rus-sians do, not Mongols, but men of Turk family, Bokhars, and Tashkandis, —information respecting the tracts of Inner Asia adjoining their country.Travels to Toorfan, Aksu, Khoten, Yarkand, and Cashmer are very rare; butCashgar, the country between the Altaï and the northern slope of the CelestialMountains (Tëen-shan), where Choogulchak,* Korgos, and Gulja, or Kura,are situated, the Khanat of Kokand, Bokhara, Tashkand, and Shersaves, tothe south of Samarkand, are frequently visited. At Orenburg, where cara-vans of some thousand camels arrive annually, and where the exchange iscrowded with different nations, M. de Gens, a well-informed man, has beencollecting, with care, for the last twenty years, a mass of very important materialsrelating to the geography of Inner Asia. Amongst the numerous itineraries,which M. de Gens communicated to me, I found the following remark: “inproceeding from Semipolatinsk to Yarkand, when we reached lake Ala-kul, orAla-dinghis, a little to the north-east of the great lake Balkashi (the Palcati-nor of D’Anville), which receives the waters of the Ele, we saw a very loftymountain, which formerly emitted fire. Even at the present day, this moun-tain, which raises itself in the lake like a little island, occasions violent tem-pests, which annoy the caravans; wherefore a few sheep are sacrificed, inpassing, to this ancient volcano.” This fact, taken from the mouth of a Tatar traveller, at the beginning of thepresent century, excited in me the more interest, since it recalled to my mindthe burning mountains of Central Asia, which we have become acquaintedwith from the learned researches of M. M. Rémusat and Klaproth in Chineseworks, and the position of which, at so great a distance from the sea, hascaused so much astonishment. Shortly before my departure from St. Peters-burgh, I received, by the great politeness of M. de Klosterman, the followinginformation, which he had procured from Bokhars and Tashkandis:—“ thedistance from Semipolatinsk to Gulja is twenty-five days: the route is bymounts Alashan and Kondegatay (in the steppe of the Kirghiz of the middlehorde), the borders of lake Savande-kul, the Tarbagatay mountains in Zunga-ria and the river Emyl: when it is traversed, the road unites to that whichleads from Choogulchak to the province of Ele. From the banks of the Emylto lake Ala-kul, the distance is sixty versts. The Tatars estimate the distanceof this lake from Semipolatinsk at 455 versts. It is on the right, and its ex-tent is 100 versts from east to west. In the midst of this lake rises a very

---

* A frontier station established by the Chinese in 1767. This town has ramparts of earth; the magis-trates and inspectors of the frontier reside here. The garrison consists of 1,000 Chinese soldiers and1,800 Manchoos or Mongols. The Chinese are permanently placed there; they constitute a militarycolony, and are obliged to cultivate the earth for their subsistence. The Manchoos and Mongols aresent from Ele, and are relleved every year.—Klaproth. |150| high mountain, called Aral-toobeh. From thence to the Chinese station be-tween the small lake Yanalashe-kul and the river Baratara,* on the banks ofwhich Kalmuks dwell, it is reckoned fifty-five versts.” Comparing the itinerary of Orenburg with that of Semipolatinsk, there re-mained no doubt that the mountain which, according to the tradition of thenatives, consequently in historical times, emitted fire, was the conical isle ofAraltoobeh. As the most important point in these statements concerned thegeographical position of the isle, and its relative situation in respect to thevolcanos discovered by Messrs. Rémusat and Klaproth, in very ancientChinese books, as then existing in the interior of Asia, to the north and southof mount Tëen-shan, it will not be out of place to insert here some detailsrespecting the geography of this region. The central and interior portion of Asia, which forms neither an immensecluster of mountains nor a continued table-land, is crossed from east to westby four grand systems of mountains, which have manifestly influenced themovements of the population; these are, the Altaï, which is terminated tothe west by the mountains of the Kirghiz; the Tëen-shan, the Kwan-lun, andthe Himalaya chain. Between the Altaï and the Tëen-shan, are placed Zunga-ria and the basin of the Ele; between the Tëen-shan and the Kwan-lun, Littleor rather Upper Bucharia, or Cashgar, Yarkand, Khoten, the great desert ofGobi (or Cha-mo), Toorfan, Khamil (Hami), and Tangout, that is, thenorthern Tangout of the Chinese, which must not be confounded with Tibet,or Se-fan; lastly, between the Kwan-lun and the Himalaya, Eastern andWestern Tibet, where H’lassa and Ladak are situated. 1. The system of the Altaï encompasses the sources of the Irtish, and ofthe Yenisseï or Kem; to the east, it takes the name of Tangnu; that of theSayanian mountains between lakes Kossogol and Baikal; farther on, that ofthe lofty Kentaï and the mountains of Dauria; lastly, to the north-east, itjoins the Yablonnoy-khrebet, the Khingkhan and the Aldan mountains, whichstretch along the sea of Okhotsk. The mean latitude of its course from east to westis between 50° and 51° 30′. We shall soon have satisfactory notions respect-ing the geography of the north-eastern part of this system, between the Baikal,Yakutsk, and Okotsk, for which the world will be indebted to Dr. Erdman,who has recently traversed those parts. The Altaï, properly so called, scarcelyoccupies seven degrees of longitude; but we give to the northernmost portionof the mountains encompassing the vast mass of high land of Inner Asia, andoccupying the space comprised between the 48th and 51st parallels, thename of the System of the Altaï, because simple names are more easily impressedupon the memory, and because that of Altaï is best known to Europeans fromthe great metallic wealth of these mountains, which now annually yield 70,000marks of silver and 1,900 marks of gold. The Altaï, in Turkish, in Mongol,Altaï-in-oola, “gold mountain,” is not a chain of mountains, forming thelimit of a country, like the Himalaya, which bounds the table-land of Tibet,and which consequently lowers itself abruptly only on the side of India, whichis lower than the other country. The plains adjoining lake Zaïsang, and espe-cially the steppes near lake Balkashi, are certainly not more than 300 toises(1968 English feet) above the level of the sea. I avoid intentionally, in this paper, conformably to the statements I collectedon the spot, employing the term Lesser Altaï, if this term is applied to thevast mass of mountains situated between the course of the Narym, lake Telet-

---

* The name of this river is Boro-tala-gol; it flows, not from east to west, but from west to east; andit empties itself not into the Alak-tugul-noor, but into the Khaltar-usike-noor.—Klaproth. |151| sky, the Bia, Serpent Mountain, and the Irtish above Oustkamenogorsk,consequently the territory of Russian Siberia, between the 79th and 86th me-ridians east of Paris, and between the parallels of 49° 30′ and 52° 30′. ThisLittle Altaï is probably, owing to its extent and elevation, much more consi-derable than the Great Altaï, whose position and existence, as a chain ofsnowy mountains, are, perhaps, equally problematical. Arrowsmith, and se-veral modern geographers, who have followed the model he has arbitrarilyadopted, give the name of Great Altaï to an imaginary continuation of theTëen-shan which is carried to the eastward of Khamil (Hami) and Barkoul(Chin-se-foo), a Manchoo town, and runs to the north-east, towards theeastern sources of the Yenisseï and Mount Tangnu. The direction of the lineof separation of the waters, between the affluents of the Orkhon and those ofthe Aral-noor, lake of the steppe, and the unfortunate practice of marking byhigh chains of mountains where systems of streams separate, have occasionedthis error. If it be desired to retain on our maps the name of Great Altaï, itshould be given to the succession of lofty mountains ranged in a course directlyopposite (parallel to the chain of the Khangai),* or from the north-west tothe south-east, between the right bank of the Upper Irtish, and the Yeke-Aral-noor, or Lake of the Great Isle, near Gobdo-Khoto. There, consequently, to the south of the Narym and of the Bukhtorma,which bounds what is called the Little-Russian Altaï, was the primitive abodeof the Turk tribes; the place where Dizabul, their grand khan, towards theclose of the sixth century, received an ambassador from the emperor of Con-stantinople. This gold-mountain of the Turks, the Kin-shan of the Chinese,a name with the same signification, bore heretofore also those of Ek-tag, and Ektel, both of which probably have an analogous meaning. It is said thatmore to the south, under the 46th parallel, and almost in the meridian of Pi-jan and Toorfan, a lofty peak is still called in Mongol Altaïnniro, “summit ofthe Altaï.” If some degrees farther to the south, this Great Altaï unites itselfto the Naiman-ula mountains, we there find a transverse ridge which, runningfrom the north-west to the south-east, joins the Russian Altaï to the Tëen-shan, northward of Barkoul and Hami.† This is not the place to developehow the system of north-western direction so general in our hemisphere istraced in the beds of the rocks, in the line of the Alps of Alghin, of thelofty steppe of the Chuya, of the chain of the Jyiktu, which is the culminatingpoint of the Russian Altaï, and in the hollows of the narrow valleys, whereflow the Chulyshman, the Chuya, the Katunia, and the Upper Charysh;lastly, in the whole course of the Irtish from Krasnoyarskoi to Tobolsk. Between the meridians of Oust-Kamenogorsk and of Semipolatinsk, the sys-tem of the Altaï mountains extends from east to west, beneath the parallelsof 59 and 50, by a chain of hills and low mountains, for 160 geographicalleagues, as far as the steppe of the Kirghiz. This range, though of very smallimportance in respect to size and elevation, is highly interesting to geognosy.

---

* Mount Khanggay-ula is to the north of the source of the Orkhon. Its summits are lofty and con-siderable. This chain is a branching off of the Altai, which comes from the north-west: it extends to theeastward to the rivers Orkhon and Tula with their affluents, and becomes the Kenteh of the Khinggan.A branch of this chain separates to the west and runs to the north under the name of the Kuku-daban;it encompasses the Upper Selengga and all its affluents, which take their origin in it, and then runs adistance of 1,000 le into the Russian territory. The Orkhon, the Tamir, and their affluents have like-wise their sources in this chain, which is probably the same which the Chinese distinguish by the nameof Yang-jin-shan.—Klaproth. † The Chinese (in their imperial geography of China), in tracing the direction of the Great Altai fromthe north-west to the south-east, makes it almost re-unite itself to the Tëen-shan, which correspondsexactly with what M. de Humboldt states.—Klaproth. |152| There does not exist a continued chain of Kirghiz mountains, which, as themaps represent, under the names of Alghidin-tsano or Alghidin-chamo, unitesthe Ural and the Altaï. Some isolated hills of 500 or 600 feet high, groupsof small mountains which, like the Semi-tau near Semipolatinsk, rise abruptlyto the height of 1,000 or 1,200 feet above the plains, deceive the traveller whois not accustomed to measure the inequality of the soil; but it is not less re-markable that these clusters of hills and small mountains have been raisedacross a furrow which forms this line of division of the waters between theaffluents of the Saras, or to the south in the steppe, and those of the Irtishto the north: a fissure which follows uniformly, as far as the meridian of Sve-rinagolovskoy, the same direction for sixteen degrees of longitude. In the line of division of the waters between the Altaï and the Ural, betweenthe 49th and 50th parallels, is observable an effort of nature, a kind of attemptof subterranean energy, to force up a chain of mountains; and this fact recallspowerfully the similar appearances I remarked in the new continent. But the non-continued range of low mountains and hills of crystallized rocks,by which the system of the Altaï is prolonged to the west, does not reach thesouthern extremity of the Ural, a chain which, like that of the Andes, presentsa long wall running from north to south, with metallic mines on its eastern side:it terminates abruptly under the meridian of Sverinogovloskoy, where geo-graphers are accustomed to place the Alghinic mountains, the name of which isentirely unknown by the Kirghiz of Troitsk and of Orenburg. II. System of the Tëen-shan.—Their mean latitude is 42°. Their culminatingpoint is perhaps the mass of mountain remarkable by its three peaks, coveredwith eternal snows, and celebrated under the name of Bokhda-ula, or “HolyMountain,” in the Mongol-Calmuc tongue; which has caused Pallas to give tothe whole chain the denomination of Bogdo. From the Bokhda-ula, theTëen-shan runs easterly towards Barkoul, where, to the north of Hami, itsinks abruptly, and spreads itself to the level of the high desert called theGreat Gobi, or Shamo, which extends south-west and north-east, fromKwa-chow, a town of China, to the sources of the Argun. Mount Nomkhun,to the north-west of the Sogok and the Sobo, little lakes of the steppe, de-notes, perhaps, by its position, a slight swell, an angle in the desert; forafter an interruption of at least ten degrees of longitude, there appears, a littlemore to the south than the Tëen-shan, in my opinion, as a continuation ofthis system, at the great bend of the Hwang-ho, or Yellow River, the snowychain of the Gajar, or Yn-shan, which runs likewise from west to east, underthe parallels of 41 and 42, consequently to the north of the country of Ordos. Let us now return to the neighbourhood of Toorfan and the Bokhda-ula, andfollow the western prolongation of the second system of mountains; we shallperceive that it extends between Gulja (Ele), the place whither the Chinesegovernment exiles criminals, and Kucha; then between Temoortu, a largelake, the name of which signifies “ferruginous water,” and Aksu, to thenorth of Cashgar, and runs towards Samarkand. The country comprised be-tween the first and second systems of mountains, or between the Altaï and theTëen-shan, is closed on the east, beyond the meridian of Peking, by theKhingkhan-ula, a mountainous crest which runs S.S.W. and N.N.E.; but tothe west, it is entirely open on the side of the Chwei, the Sarasu and theLower Sihoon. In this part there is no transverse ridge, provided, at least,we do not regard as such the series of elevations which extend, north andsouth, to the west of lake Zaisang, across the Targabatay, as far as the north- |153| eastern extremity of the Ala-tau,* between lakes Balkash and Alak-tugul-noor, and then beyond the course of the Ele, to the eastward of the Temoor-tu-nor (between lat. 44° and 49°), and which present the appearance of awall occasionally interrupted on the side of the Kirghiz steppe. It is quite otherwise with the portion of Central Asia, which is bounded bythe second and third systems of mountains, the Himalaya and Kwan-lun. Infact, it is closed to the west in a very evident manner by a transverse ridge,which is prolonged from south to north, under the name of Bolor or Beloor-tagh.† This chain separates Little from Great Bucharia, and from Cashgar,Badakshan, and the Upper Jihoon or Amoodaria. Its southern portion, whichconnects with the system of the Kwan-lun mountains, forms, according to thedenomination used by the Chinese, a part of the Tsung-ling. To the north, itjoins the chain which passes to the north-west of Cashgar, and bears the nameof the defile of Cashgar (Cashgar-divan, or davan), according to the narrativeof Nasaroff, who, in 1813, travelled as far as Kokand. Between Kokand,Dervazeh, and Hissa, consequently between the still unknown sources of theSihoon and Amoo-daria, the Tëen-shan rises previous to sinking again in theKhanat of Bokhara, and presents a group of lofty mountains, several summitsof which, such as the Takt-i-Suleyman, the crest called Terek and others, arecovered with snow even in summer. Farther to the east, on the road whichruns from the western bank of lake Temoortu to Cashgar, the Tëen-shan doesnot appear to me to attain so great an elevation; at least no mention is madeof snow in the itinerary from Semipolatinsk to Cashgar. The road passes tothe eastward of lake Balkashi, and to the westward of lake Yssi-kul or Temoor-tu, and traverses the Narun or Narym, an affluent of the Sihoon. At 105versts to the south of the Narun, it goes over Mount Rovatt, which is prettyhigh, and about fifteen versts wide; it has a large cavern, and is situated be-tween the At-bash, a small river, and the little lake of Chater-kul. This is theculminating point previous to arriving at the Chinese post placed to the south ofthe Aksu, a small river of the steppe, the village of Artush and Cashgar.This city, built on the banks of the Aratumen, contains 15,000 houses and80,000 inhabitants, but is yet smaller than Samarkand. The Cashgar-davan ‡ does not appear to form a continuous wall, but to offer an open pas-sage at several points. M. Gens expressed to me his surprise that none of thenumerous itineraries of Bokharians which he has collected, make mention ofa lofty chain of mountains between Kokand and Cashgar. The great snowy

---

* This is a name which has occasioned much confusion. The Kirghis, particularly those of the grandhorde, give the title of Ala-tagh (Alatau, “speckled mountains”) to a series of elevations extendingfrom west to east, under the parallels of 43° 30′ to 45°, from the Upper Sihoon (Syr-daria or Jaxartes),near Tonkat, towards lakes Balkashi and Temoortu. The eastern portion of the Ala-tau rises consider-ably at the great sinuosity made by the Sihoon to the north-west, and connects with the Kara-tau(“Black Mountain”) at Taras or Turkestan. The natives likewise give the name of Ala-tau to themountains to the south of the Tarbagatay between lakes Ala-kul, Balkashi, and Temoortu. Is it fromthese denominations that geographers have been in the habit of calling the whole second system ofmountains that of Tëen-shan, Alak or Ala-tau? The Oolug-tagh, or “Great Mountain,” named onsome maps Oulug-tag Oolu-tau, and Ooluk-tagh, must not be confounded with the Ala-tau or Ala-taghi.† According to M. Klaproth, this transversal ridge is named in Ouigoor Boolyt-tagh, “CloudyMountain,” on account of the extraordinary rains which fall uninterruptedly in this latitude, duringthree months. West of this transverse ridge of Beloor, is the station of Pamir, nearly under the pa-rallel of Cashgar. Marco Polo has named, after this station, a table-land of which modern geographershave made sometimes a chain of mountains, sometimes a province situated farther to the south. Thisdistrict is still interesting to the naturalist, on account of the celebrated Venetian traveller having firstobserved there a fact, which has so often occurred in my experience, at considerable elevations, in theNew World, namely, that it is extremely difficult to light and to keep fire in there.‡ The terms davan, in Oriental-Turki, dabahn, in Mongol, and dabagan, in Manchoo, denote not amountain, but a pass in a mountain; Cashgar-davan, therefore, signifies only the pass across the moun-tains to Cashgar.—Klaproth. |154| mountains seem not to re-appear till east of the meridian of Aksu, forthese same itineraries mention Jeparleh,* a glacier covered with perpetual snow,on the Kura road, on the banks of the Ele at Aksu, nearly half-way, betweenthe warm springs of Arashan to the north of Kanjeilao, a Chinese station, andthe advanced post of Tamga-tash. The western prolongation of the Tëen-shan or Mooz-tag, as the editors ofthe Memoirs of Sultan Baber call it by pre-eminence, deserves a particularnotice. At the point where the Beloor-tag joins the right angle of the Mooz-tag,or traverses as a lode or vein this great system, the latter continues its coursewithout interruption from east to west, under the denomination of Asferah-tag,to the south of the Sihon, towards Khojand and Urateppeh, in Ferghana. Thischain of Asferah, which is covered with perpetual snow, and is improperlycalled the chain of Pamer, separates the sources of the Sihon (Jaxartes) fromthose of the Amoo (Oxus); it turns to the south-west, nearly in the meridianof Khojand, and in this direction is called, as far as near Samarkand, Ak-tag(“White or Snowy Mountain”), or Al-botom. Farther to the west, on thesmiling and fertile banks of the Kohik, commences the great dip or depressionof land, comprehending Great Bucharia, the country of Maveralnahar, which isso low, and where the highly-cultivated soil and the wealth of the towns attractperiodically the invasions of the people of Iran, Candahar, and Upper Mon-golia; but beyond the Caspian Sea, nearly in the same latitude, and in thesame direction as the Tëen-shan, appears the Caucasus, with its porphyritic andtrachytic rocks. One is inclined, therefore, to regard it as a continuation ofthe furrow, in the form of a lode, on which the Tëen-shan rises in the east,just as, to the west of the great cluster of the mountains of Azerbaijan andArmenia, is observable, in Taurus, a continuation of the action of the fissureof the Himalaya and the Hindu Coosh. It is thus that, in a geognostic sense,

---

* This is the Moosar-tag, or glacier between Ele and Kucha. The ice with which it is sheeted givesit the appearance of a mass of silver. A road, called Mussar-dabahn, cut through these glaciers, leadsfrom the S. W. to the N. or, to, speak more accurately, from Little Bucharia to Ele. The following isa description of this mountain by a modern Chinese geographer: “to the north is the post-station ofGakhtsa-karkai, and to the south that of Tamga-tash, or Terma Khada; they are distant from eachother 120 le. On proceeding to the south, after quitting the former, the view extends over a vast spacecovered with snow, which, in winter, is very deep. In summer, on the top of the ice, snow andmarshy places are found. Men and cattle follow the winding paths at the side of the mountain. Who-ever is so imprudent as to venture upon this sea of snow is irrecoverably lost. After traversing upwards oftwenty le, you reach the glacier, where neither sand, trees, nor grass can be seen: the most terrifyingobjects are the gigantic rocks formed into one by masses heaped upon one another. When the eyedwells upon the intervals which separate these masses of ice, a gloomy chasm appears, into which thelight never penetrates. The sound of the water rushing beneath the ice resembles the report of thunder.Carcasses of camels and horses are scattered here and there. In order to facilitate the passage, stepshave been cut in the ice, to ascend and descend, but they are so slippery that they are extremely danger-ous. Too frequently travellers find their grave in these precipices. Men and cattle walk in file, trem-bling with alarm, in these inhospitable tracts. If night surprises the traveller, he must seek shelterunder a large stone; if the night happen to be calm, very pleasing sounds are heard, like those of seve-ral instruments combined: it is the echo which repeats the cracking noise produced by the breaking ice.The road, which is pursued the day before, is not always that which it is convenient to follow the nextday. At a distance, to the west, a mountain, which has been hitherto inaccessible, displays its scarpedand icy summits. The halting-place of Tamga-tash is eighty le from this place. A river, called Moos-sur Gol, rushes with frightful impetuosity from the edges of the ice, flows to the south-east, and joinsthe Erghew, which falls into lake Lob. Four days’ journey to the south of Tamga-tash, is an aridplain, which does not produce the smallest plant. At eighty or ninety le further off, gigantic rocksstill recur. The commandant of Ushi sends every year one of his officers with oblations to this glacier.The formula of the prayer recited on this occasion is transmitted from Peking by the Tribunal of Rites.Ice is found along the whole crest of the Tëen-shan, if it is traversed lengthwise; but, on the contrary,if it is crossed from north to south, that is in its width, ice is found only in a space of a few le. Everymorning, ten men are employed, in the pass of Mussar-tag, in cutting steps for ascending and descend-ing; in the afternoon, the sun has either melted them or rendered them extremely slippery. Sometimesthe ice gives way under the feet of the travellers, and they are ingulphed, without a hope of ever seeingday-light again. The Mohamedans of Little Bucharia sacrifice a ram previous to traversing these moun-tains. Snow falls there throughout the year: it never rains.—Klaproth. |155| the disjointed members of the mountains of Western Asia, as Mr. Ritter callsthem, connect themselves with the forms of the land in the east. III. The System of the Kwan-lun, or Koolkun, or Tartash-davan, entersKhoten (Elechi),* where Hindu civilization and the worship of Buddha pene-trated 500 years before it reached Tibet and Ladak, between the cluster ofmountains of Kookoo-noor and Eastern Tibet, and the country called Kachi. This system of mountains commences westward of the Tsung-ling (“Onion orBlue Mountains”), upon which M. Abel Rémusat has diffused so much light inhis learned History of Khoten. This system connects itself, as already ob-served, with the transverse chain of Bolor; and, according to the Chinesebooks, forms the southern portion of it. This quarter of the globe, betweenLittle Tibet and Badakshan, abounding in rubies, lazulite, and turquoise, isvery little known; and, according to recent accounts, the table-land of Kho-rasan, which runs towards Herat, and bounds the Hindu-Kho or Hindu-Coosh, to the north, appears to be a continuation of the system of the Kwan-lun to the west, rather than a prolongation of the Himalaya, as commonlysupposed. From the Tsung-ling, the Kwan-lun or Koolkun runs from westto east, towards the sources of the Hwang-ho (Yellow River), and penetrates,with its snowy peaks, into the Chinese province of Shen-se. Nearly in themeridian of these sources, rises the great cluster of mountains of lake Koo-koo-noor, a cluster which supports itself, on the north, against the snowychain of the Nan-shan, or Ki-lian-shan, extending also from west to east. Be-tween the Nan-shan and the Tëen-shan, on the side of Hami, the mountains ofTangout bound the edge of the high desert of Gobi or Shamo, which stretchesfrom south-west to north-east. The latitude of the middle portion of theKwan-lun is about 35° 30′. IV. System of the Himalaya.—This separates the valleys of Cashmer (Seri-nagur) and Nepal from Butan and Tibet; to the west, it stretches, by Jevahir,to 4,026 toises (26,420 feet); to the east, by Dhavalaghiri, to 4,390 (28,809feet) of actual height above the level of the sea; it runs generally in a directionfrom N.W. to S.E., and consequently is not parallel with the Kwan-lun; itapproaches it so nearly, in the meridian of Attock and Jellalabad, that betweenCabul, Cashmer, Ladak, and Badakshan, the Himalaya seems to form only asingle mass of mountains with the Hindu-Kho and the Tsung-ling. In likemanner, the space between the Himalaya and the Kwan-lun is more shut upwith secondary chains and isolated masses of mountains, than the table-landsbetween the first, second, and third systems of mountains. Consequently, Tibetand Kachi cannot properly be compared, in respect to their geognostic con-struction, with the elevated longitudinal valleys,† situated between the chainof the eastern and western Andes, for example, with the table-land which en-closes the lake of Titicaca, a correct observer of which (Mr. Pentland) foundthat its elevation above the sea was 1,986 toises (13,033 feet). Nevertheless,it must not be represented that the height of the table-land between the Kwan-

---

* The position of Khoten is very incorrectly laid down in all the maps. Its latitude, according to theastronomical observations of the Missionaries Felix d’Arocha, Espinha, and Hallerstein, is 37° 0′; thelongitude 35° 52′ W. of Peking. This longitude determines the mean direction of the Kwan-lun.† In the Andes, I found that the mean height of the longitudinal valley between the Eastern andWestern Cordilleras, from the cluster of mountains of Los Robles, near Popayan, to that of Pasco,as well as those in 2° 20′ N. lat. to 10° 30′ S. lat., was about 1,500 toises (9,843 feet). The table-land, orrather longitudinal valley, of Tiahuanaco, along the Lake of Titicaca, the primitive seat of Peruviancivilization, is more elevated than the Peak of Teneriffe. However, according to my experience, itcannot be asserted generally that the absolute height to which the soil of the longitudinal valleys appearsto have been raised by subterranean force, augments with the absolute height of the neighbouringchains. In like manner, the elevation of isolated chains above the vallies is very various, showing thatat the foot of the chain the raised plain is elevated at the same time, or has preserved its ancient level.|156| lun and the Himalaya, as well as in all the rest of Central Asia, is equalthroughout. The mildness of the winters, and the cultivation of the vine,* inthe gardens of H’lassa, in the parallel of 29° 40′,—facts ascertained by the ac-counts published by M. Klaproth and the Archimandrite Hyacinth,—proclaimthe existence of deep valleys and circular hollows. Two considerable rivers,the Indus and the Zzambo (Sampoo),† denote a depression in the table-landof Tibet, to the north-west and south-east, the axis of which is found nearlyin the meridian of the gigantic Javahir, the two sacred lakes of Manassora-vara and Ravana Hrada, and Mount Caïlasa, or Caïlas, in Chinese O-new-ta,and in Tibetan Gang-dis-ri. From this nucleus springs the chain of Kara-ko-rum-padisha, which runs to the north-west, consequently to the north of La-dak, towards the Tsung-ling; and the snowy chains of Hor (Khor) and Zzang,which run to the east. That of Hor, at its north-western extremity, connectsitself with the Kwan-lun; its course, from the eastern side, is towards theTangri-noor (“Lake of Heaven”). The Zzang, farther to the south than thechain of Hor, bounds the long valley of the Zzangbo, and runs from westto east, towards the Nëen-tsin-tangla-gangri, a very lofty summit which,between H’lassa and lake Tangri-noor (improperly called Terkiri), terminatesat Mount Nom-shun-ubashi. Between the meridians of Ghorka, Katmandu,and H’lassa, the Himalaya sends out to the north, towards the right bank, orthe southern border of the valley, of the Zzang-bo, several branches coveredwith perpetual snow. The highest is Yarla-shamboy-gangri, the name of which,in Tibetan, signifies “the snowy mountain in the country of the self-existingdeity.” This peak is to the westward of lake Yamruk-yumdzo, which ourmaps call Palteh,‡ and which resembles a ring, being almost filled by an island. If, availing ourselves of the Chinese writings which M. Klaproth has collect-ed, we follow the system of the Himalaya towards the east, beyond the Eng-lish territories in Hindustan, we perceive that it bounds Assam to the north,contains the sources of the Brahmaputra, passes through the northern part ofAva, and penetrates into the Chinese province of Yun-nan, where, to thewestward of Yung-chang, it exhibits sharp and snowy peaks; it turns abruptlyto the north-east on the confines of Ho-kwang, of Keang-si, and of Fuh-kien,and extends, with its snowy summits near to the ocean, where we find, as if itwas a prolongation of this chain, an island (Formosa), the mountains of whichare covered with snow during the greatest part of the summer, which showsan elevation of at least 1,900 toises (12,469 feet). Thus we may follow thesystem of the Himalaya, as a continuous chain, from the Eastern Ocean, andtrack it by the Hindu-Coosh, across Candahar and Khorasan; and lastly as faras the Caspian Sea in Azerbaijan, through an extent of seventy-three degreesof longitude, half that of the Andes. The western extremity, which is vol-canic, but covered likewise with snow to Demavend, loses the peculiar charac-ter of a chain in the cluster of the mountains of Armenia, connected with theSangalu, the Bingheul, and Cashmer-dag, lofty summits in the pashalic of Er-zeroum. The mean direction of the system of the Himalaya is N. 55° W. (To be concluded next Month.)

by baron a. von humboldt. The aforegoing are the principal features of a geognostic description ofCentral Asia, which I have drawn up with the aid of numerous materials ac-cumulated by me during a long series of years. Of these materials, the por-tion for which we are indebted to modern European travellers is of small impor-tance, in comparison with the prodigious space which is occupied by the chainof the Altaï, the Himalaya mountains, the transverse ridges of the Bolor andthe Kingkan. Those who, at the present day, have published the most im-portant and complete details on these subjects are the learned persons whoare conversant with Chinese, Manchoo, and Mongol literature. The moregeneral the cultivation of the Asiatic dialects shall become, the better shall weappreciate the utility of these so-long-neglected sources, for the study of thegeognostic constitution of Middle Asia. Until M. Klaproth diffuses a newlight upon this study by a special work of his own, the picture which I havehere exhibited of the four systems of mountains which run from east to west,the materials for which were, in a great part, furnished by the learned personwhom I have just named, will not be without its use. In order to ascertainthe characteristic properties which are to be found in the inequalities of theglobe’s surface, and to discover the laws which regulate the local dispositionof the masses of mountains, and the dips or depressions, we may have recourseto the analogy which other continents may offer. If once the grand forms andpredominating courses of the chains are well determined, we shall see con-nected with this fundamental principle, as with a common type, whatever ap-peared at first isolated in these phenomena, and at variance with rules, pro-claiming another date of formation. This method, which I followed in mygeognostic description of South America, I have endeavoured to apply hereto the limits of the grand masses of Middle Asia. In bestowing a parting glance upon the four systems of mountains whichdivide the continent of Asia from east to west, we observe that the southernhas the greatest extent, and the fullest developement in respect of length.The Altaï hardly attains, with its elevated summits, the 78th degree; theTëen-shan, the chain at whose foot are situated Hami, Aksu, and Cashgar,reaches at least to the meridian of 69° 45′; provided we place Cashgar, ac-cording to the authority of the missionaries, in 71° 37′ east of Paris.† Thethird and fourth systems are, as it were, blended in the grand clusters of Ba-dakshan, Little Tibet, and Cashgar. Beyond the 69th and 70th meridiansthere is but one chain, that of the Hindu-Kho, which is depressed towardsHerat, but which afterwards, to the southward of Asterabad, rises to a con-siderable height towards the volcanic and snowy mountain of Demavend.The table-land of Iran, which, in its greatest extension, from Tehran to Shi-raz, appears to attain the average height of 650 toises (4,265 feet), throws off,towards India and Tibet, two branches, the Himalaya and the Kwan-lunchain, and forms a bifurcation of the furrow from which the mountainousmasses rise. Thus the Kwan-lun may be considered as a saliant fracture ofthe Himalaya. The intermediate space, comprising Tibet and Kachi, is inter-sected by numerous rents in all directions. This analogy with the most com-

---

* Concluded from p. 156.† The astronomical geography of Inner Asia is still very confused, because the elements of the obser-vations are not known, merely the results.|233| mon phenomena of the formation of lodes or veins is manifested in a verystriking manner, as I have elsewhere shewn, in the long and narrow line ofthe Cordilleras of the New World. We may trace beyond the Caspian Sea, to the 45th meridian (of Paris), thesystems of the Himalaya and the Kwan-lun, which are prolonged till they joinin the cluster situated between Cashmer and Fyzabad. Thus the chain of theHimalaya remains to the south of the Bolor, the Ak-tag, the Mingboolak, andthe Ala-tau, between Badakshan, Samerkand, and Turkestan; to the east ofthe Caucasus it joins the table-land of Azerbaijan, and bounds to the south thegreat dip or valley, of which the Caspian Sea and lake Aral* occupies the lowestbasin, and in which a considerable portion of land whose surface is probably18,000 square leagues, and which lies between the Kooma, the Don, the Vol-ga, the Yak, the Obsheysyrt, lake Aksakal, the Lower Sihon, and the Khanatof Khiva, upon the shores of the Amoo-daria, is situated below the level ofthe ocean. The existence of this singular depression has been the object oflaborious barometrical observations of levels between the Caspian Sea and theBlack Sea, by MM. Parrot and Engelhardt; between Orenburg and Gourievat the mouth of the Yayk, by MM. Helmersen and Hoffmann. A country solow is abundant in tertiary formations, whence proceed garnets, and debris ofscorified rocks, and it offers to the geognostic inquirer, from the constitutionof its soil, a phenomenon hitherto unique in our planet. To the south ofBaku, and in the gulf of Balkan, this aspect is materially modified by volcanicinfluence. The Academy of Sciences of St. Petersburgh has recently com-plied with my solicitations to get determined by a series of stations of baro-metric levels upon the north-eastern edge of this basin, upon the Volga betweenKamyshin and Saratov, upon the Yayk between the Obsheysyrt, Orenburg,and the Uralsk, upon the Yemba and beyond the hills of Mougojar, by whichthe Ural extends itself towards the south, on the side of lake Aksakal and to-wards Sarasu, the position of a geodœsic line, uniting all the points at thelevel of the surface of the ocean. I have referred already to the hypothesis, according to which this great dipof the land of Western Asia was formerly continued as far as the mouth ofthe Ob and the Frozen Sea, by a valley traversing the desert of Kara-koumand the numerous groupes of oases in the steppes of the Kirghiz and Baraba.Its origin appears to me to be more ancient than that of the Ural mountains,the southern prolongation of which may be traced in an uninterrupted coursefrom the table-land of Gaberlinsk to Oostoort, between lake Aral and theCaspian Sea. Would not a chain, whose height is so inconsiderable, have en-tirely disappeared if the great furrow of the Ural had not been formed subse-quently to this depression? Consequently, the period of the sinking ofWestern Asia coincides rather with that of the swell of the table-land of Iran,that of Central Asia, the Himalaya, the Kwan-lun, the Tëen-shan, and all theold systems of mountains running from east to west; perhaps also with theperiod of the exaltation of the Caucasus and the cluster of mountains of Ar-menia and Erzeroum. No part of the earth, not even excepting South Africa,presents a mass of land so extensive, and elevated to so great a height, as inInner Asia. The principal axis of this exaltation, which probably precededthe eruption of the chains from the clefts running from east to west, as in thedirection of S. W. and N.E., from the cluster of mountains between Cashmer,

---

* A series of barometrical levels continued throughout a very severe winter, during the expedition ofColonel Berg, from the Caspian Sea to the western shore of Lake Aral, at the Bay of Mertvoy Kultuk,by Captains Duhamel and Anjou, has demonstrated that the level of Lake Aral is 117 English feetabove that of the Caspian Sea.|234| Badakshan, and the Tsung-ling in Tibet, where are situated the Caïlasa, andthe sacred lakes,* as far as the snowy summits of the Inshan and Khingkan.† The elevation of so enormous a mass would suffice to produce a hollow which,even at the present day, is perhaps not half filled with water, and which, sinceit was formed, has been so modified by the action of subterranean forces, that,according to the traditions of Tatars, collected by Professor Eichwald, thepromontory of Absheron was formerly united by an isthmus with the oppositecoast of the Caspian Sea in Turcomania. The great lakes, which have beenformed in Europe at the foot of the Alps, are a phenomenon analogous to thecavity in which the Caspian Sea is situated, and owe in the same manner theirorigin to a sinking of the soil. We shall soon see that it is principally in thecompass of this hollow, consequently in the space where the resistance wasleast, that recent traces of volcanic action are apparent. The position of mount Aral-toobeh, which formerly emitted fire, of theexistence of which I became aware from the itineraries of Colonel Gens, be-comes more interesting when we compare it with that of the volcanos of Pïh-shan and Ho-chow, on the northern and southern sides of the Tëen-shan,with that of the solfatara of Urumtsi, and with that of the adjoining chasm oflake Darlay, which exhales ammoniacal vapours. The researches of MM.Klaproth and Rémusat acquainted us with this last fact upwards of six yearsago. The volcano situated in about the latitude of 42° 25′ or 42° 35′, betweenKorgos, on the banks of the Ele, and Kucha, in Little Bucharia, belongs tothe chain of the Tëen-shan: perhaps it may be on the northern face, threedegrees to the eastward of lake Yssi-kul or Tremoortu. Chinese authors callit Pïh-shan (“White Mountain”), Ho-shan, and Aghi (“Fiery Mountain.”)‡ It is not known with certainty whether the name of Pïh-shan implies that itssummit reaches the line of perpetual snow, which the height of this mountain

---

* The lakes Manasa and Ravan Hrad. Manasa, in Sanscrit, signifies “spirit.” Manasa-vara isthe easternmost of these two lakes: its name means literally “the most perfect of honourable lakes.”The westernmost lake is named Ravanah Hrad, or “Lake of Ravana,” after the celebrated hero of the Ramdyana. Bopp. † This direction of the axis of exaltation from the S.W. to the N.E. is again found beyond the 55thdegree of latitude, in the space comprized between Western Siberia, a low country, and Eastern Si-beria, a country full of chains of mountains: this space is bounded by the meridian of Irkutsk, theFrozen Sea, and the Sea of Okotsk. Dr. Erdman has discovered among the Aldan mountains, atAllakh-yuma, a peak 5,000 feet high. To the north of the Kwan-lun, the chain of Northern Tibet,and to the west of the meridian of Peking, the portions of elevated land most important in respect tothe extent and height, are the following:—1. To the east of the cluster of the Khokhonoor, the spacebetween Toorfan, Tangout, the great sinuosity of the Hwang-ho, Garjan, and the chain of the Khing-khan, a space which comprehends the great desert of Gobi. 2. The table-land between the snowymountains of Khanghay and Tangnu, and between the sources of the Yeniseï, the Selengga and theAmoor. 3. To the west of the district watered by the upper course of the Oxus (Amu), and of theJaxartes (Sihoon); between Fyzabad, Balkh, Samarkend and the Ala-tau near Turkestan, to thewestward of the Bolor (Beloot-tag). The elevation of this transverse ridge has produced in the soil ofthe great longitudinal valley of the Tëen-shan-nar-lu, between the second and third systems of moun-tains from east to west, or between the Tëen-shan and the Kwanlun, a counter-slope from west toeast, whilst in the longitudinal valley of the Tëenshan-pe-lu in Zungaria, between the Tëen-shan andthe Altaï, a general declivity is observable from east to west.‡ The details given M. Klaproth (Tabl. Hist. de l’ Asie, p. 110; Mém. relatifs à l’Asie, t. ii. p. 358)are the most complete, and derived principally from the history of the Ming dynasty. M. Abel-Ré-musat (Journ. Asiat. t. v. p. 45; Descrip. de Khotan, t. ii. p. 9), has added more in the Japanesetranslation of the grand Chinese Encyclopædia. The root ag, which is found in the word Aghi, accord-ing to M. Klaproth, signifies “fire” in Hindustani. To the south of Pïh-shan, in the neighbourhoodof Khoten, belonging to the Teen-shan-nar-lu, there can be no doubt that, prior to our era, Sanscritwas spoken, or a language possessing a strong analogy with it: but in Sanscrit a flaming mountain iscalled Agni-ghiri. According to M. Bopp, Aghi is not a Sanscrit word. Humboldt. The root ag, which is found in the word Aghi, signifies “fire” in all the dialects of Hindustan;this element is ag in Hindustani, agh in Mahratta, and the form of agi is still preserved in the dialectof the Punjab. The word agni, by which “fire” is commonly designated in Sanscrit, belongs to thesame root, as well as agun in Bengalee, ogun in Russian, and the ignis of the Latins. Klaproth. |235| would determine, at least the minimum; or whether it merely denotes theglittering hue of a peak covered with saline substances, pumice stone, and vol-canic ashes in decomposition. A Chinese author of the seventh century says:at 200 le, or fifteen leagues, to the north of the city of Kwei-chow (nowKucha), in about the latitude of 41° 37′ and longitude 80° 35′ E., accord-ing to the astronomical determination of the missionaries made in thecountry of the Eleuths, rises the Pïh-shan, which emits fire and smoke with-out interruption. It is from thence sal ammoniac is brought; upon one of thedeclivities of the Fiery Mountain (Ho-shan), all the stones burn, melt, andflow to a distance of some tens of le. The fused mass * hardens as it becomescold. The natives use it in disorders † as a medicine: sulphur is also foundthere. M. Klaproth observes that this mountain is now called Khalar,‡ and that,conformably to the account given by the Bokhars who bring to Siberia salammoniac (called nao-ska in Chinese, and nōshāder in Persian), the mountainto the south of Korgos is so abundant in this species of salt that the nativesfrequently employ it as a means of paying their tribute to the emperor ofChina. In a recent Description of Central Asia, published at Peking in 1777,we find the following statement:—“the province of Ku-cha produces copper,saltpetre, sulphur, and sal ammoniac. The latter article comes from an ammo-niac mountain to the north of the city of Ku-cha, which is full of chasms andcaverns. These apertures, in spring, summer, and autumn, are filled withfire, to such a degree that, during the night the mountain appears illuminatedby thousands of lamps. No one is then able to approach it. In winter alone,when the vast quantity of snow has extinguished the fire, the natives are ableto labour in collecting the sal ammoniac, for which purpose they strip them-selves quite naked. The salt is found in caverns, in the form of stalactites,which renders it difficult to be detached.” The name of Tartarian salt, for-merly given in commerce to this salt, ought to have long ago directed attentionto the volcanic phenomena of Central Asia. M. Cordier, in his letter to M. Abel Rémusat, “on the existence of twoburning volcanos in Central Asia,” calls Pïh-shan a solfatara like that of Puz-zuoli. In the state in which it is described in the work cited further back, thePïh-shan might well deserve only the name of an extinct volcano, althoughthe igneous phenomena are wanting in the solfataras I have seen: such as thoseof Puzzuoli, the crater of the peak of Teneriffe, the Rucu-pishinsha, and thevolcano of Jorullo; but passages in more ancient Chinese historians, who re-late the march of the army of the Heung-nus, in the first century of our era,

---

* The history of the Chinese dynasty of Tang, speaking of the lava from the Pïh-shan, states thatit ran like liquid fat. Klaproth. † This is not lava, but the saline particles which appear in the form of an efflorescence on itssurface.‡ The Pïh-shan of the ancient Chinese, at present has the Turk name of Eshik-bash: Eshik is aspecies of goat, and bash signifies “head.” Sulphur is produced there in abundance. The Eshik-bashbelongs to the elevated mountains which in the time of the Wei dynasty (the third century) bounded,to the north-west, the kingdom of Kwei-tsu (Ku-cha); it is the Aghi-shan under the Suy dynasty (inthe early moiety of the seventh century). The history of this dynasty relates that this mountainalways shewed fire and smoke, and that sal-ammoniac was obtained there. In the description of theWestern country, which forms a part of the history of the Tang dynasty, we find that the mountain inquestion was then called Aghi-teen-shan (which may be translated “mountain of fields of fire”), orPïh-shan (“white mountain”), that it was to the north of the city of Ilolo, and that it emittedperpetual fire. Ilolo (or perhaps Irolo, Ilor, or Irol) was then the residence of the King of Kwei-tsu.The Eshik-bash is to the north of Ku-cha, and 200 le to the west of the Khan-tengri, which formspart of the chain of the Teen-shan. The Eshik-bash is very large, and much sulphur and sal-ammo-niac is even now collected there. It gives birth to the river Eshik-bash-gol, which flows to the southof the city of Kucha, and falls, after a course of 200 le, into the Erghew.|236| speak of masses of rocks in fusion flowing to the distance of some miles: sothat it is impossible, in these expressions, not to understand eruptions of lava.The ammoniac mountain between Kucha and Korgos has also been a volcano,in activity, in the strictest sense of the word: a volcano which emitted torrentsof lava in the centre of Asia, 400 geographical leagues from the Caspian Seato the west, 433 from the Frozen Sea to the north, 504 from the Great Oceanto the east, and 440 from the Indian-Ocean to the south. This is not theplace to discuss the question relative to the influence of the proximity of thesea on the action of volcanos; I merely solicit attention to the geographicalposition of the volcanos of Inner Asia, and their reciprocal relations. ThePïh-shan is distant from 300 to 400 leagues from all the seas. When I returnedfrom Mexico, some celebrated mineralogists expressed their astonishmentwhen they heard me speak of the volcanic eruption of the plain of Jorullo,and of the volcano of Popocatepetl, as still in activity; although the formeris only thirty leagues from the sea, and the latter forty-three leagues. GebelKoldaghi, a conical and smoking mountain of Kordofan, of which Mr. Rüp-pell was told at Dongola, is 150 leagues from the Red Sea, and this distanceis but a third of that at which the Pïh-shan, which for 1,700 years has emittedtorrents of lava, is situated from the Indian Ocean. The hypothesis, con-formably to which the Andes present no volcano in activity in those partswhere the chain recedes from the sea, is without foundation. The system ofmountains of the Caraccas, which run from east to west, or the chain of thecoast of Venezuela, is shaken by violent earthquakes, but has no more aper-tures which are in permanent communication with the interior of the earth,and which discharge lava, than the chain of the Himalaya, which is little morethan 100 leagues from the gulf of Bengal, or the Ghauts, which may almost betermed a coast-chain. Where trachyte has been unable to penetrate acrossthe chains when they have been elevated, they discover no chasms; no chan-nels are opened whereby the subterranean forces can act in a permanent man-ner at the surface. The remarkable fact of the proximity of the sea wherevervolcanos are still in activity,—a fact which, in general, is not to be denied,—seems to be accounted for less by the chemical agency of the water, than bythe configuration of the crust of the globe, and the deficiency of resistancewhich, in the vicinity of maritime basins, the elevated masses of the continentoppose to elastic fluids, and to the efflux of bodies in fusion in the interior ofour planet. Real volcanic phenomena may occur, as in the old country of theEleuths, and at Toorfan, to the south of the Tëen-shan, wherever, owing toancient resolutions, a fissure is opened in the crust of the globe at a distancefrom the sea. The reason why volcanos in activity are not more rarely remotefrom the sea, is merely because, wherever an eruption has been unable to forceitself through the declivity of continental masses towards a maritime basin, avery unusual concurrence of circumstances is requisite to permit a permanentcommunication between the interior of the globe and the atmosphere, and toform apertures which, like intermittent warm springs, effuse, instead of water,gases and oxidised earths in fusion, in other words, lava. To the eastward of the Pïh-shan, the “White Mountain” of the Eleuths,the whole northern slope of the Tëen-shan presents volcanic phenomena:“lava and pumice-stone are seen there, and even considerable solfataras,which are called ‘fiery places.’ The solfatara of Uroomtsi is five leagues incircumference; in winter it is not covered with snow; it would be supposedto be full of ashes. If a stone be thrown into this basin, flames issue forth, as |237| well as a black smoke, which continues some time. Birds dare not fly overthese fiery places.” Eastward, sixty leagues from Pïh-shan, is a lake of veryconsiderable extent, the different names of which in the Chinese, Kirghis, andCalmuc languages, signify “warm salt and ferruginous water.” If we cross the volcanic chain of the Tëen-shan, we find E.S.E. of lake Yssi-kul (so often mentioned in the itineraries which I have collected), and of thevolcano of the Pïh-shan, the volcano of Toorfan, which may also be called thevolcano of Ho-chow (“City of Fire”), for it is very near that city.* M.Abel Rémusat has made particular mention of this volcano in his Histoire deKhoten, and in his letter to M. Cordier.† No reference is made to stonymasses in fusion (torrents of lava), there, as at Pïh-shan; but “a column ofsmoke is seen continually to issue; this smoke gives place at night to a flamelike that of torch. Birds and other animals, upon which the light falls, appearof a red colour. The natives, when they go thither to collect the nao-sha, orsal ammoniac, put on wooden shoes, for leather soles would be very soonburned.” Sal ammoniac is procured at the volcano of Ho-chow not only in theform of a crust or sediment, according as it is deposited by the vapours whichexhale it; but Chinese books likewise make mention of “a greenish liquorcollected in cavities, which is boiled and evaporated, and from it sal ammoniacis obtained in the form of small lumps like sugar, of extreme whiteness andperfect purity.” Pïh-shan and the volcano of Ho-chow or Toorfan are 140 leagues apart, inthe direction of east and west. About forty leagues westward of the meridianof Ho-chow, at the foot of the gigantic Bokhda-ula, is the great solfatara ofUroomtsi. At 140 leagues north-west of this, in a plain adjoining the banksof the Khobok, which flows into the small lake of Darlay, rises a hill, “theclefts of which are very warm, though they do not exhale smoke (visible va-pours): the sal ammoniac in these crevices is sublimed into so solid a coating,that the stone is obliged to be broken in order to get it.” These four places hitherto known, namely, Pïh-shan, Ho-chow, Uroomtsi,and Khobok, which exhibit evident volcanic phenomena, in the interior ofAsia, are 130 or 140 leagues to the south of the point of Chinese Zungaria,where I was at the beginning of 1829. Aral-tubeh, the conical and insularmountain of lake Ala-kul, which has been in a state of ignition in historicaltimes, and which is mentioned in the itineraries collected at Semipolatinsk, isin the volcanic territory of Bishbalik. This insular mountain is situated tothe west of the ammoniac-cavern of Khobok, and to the north of Pïh-shan,which still emits light, and which formerly discharged lava, and at a distanceof sixty leagues from each of these two points. From Lake Ala-kul to LakeZaisang, where the Russian Cossacks of the line of the Irtish exercise theright of fishing, by connivance of the Mandarins, the distance is reckoned atfifty-one leagues. The Tarbagatay, at the foot of which is situated Choogon-chak, a town of Chinese Mongolia, and where Dr. Meyer, the learned andenterprizing companion of M. Ledebour, fruitlessly essayed, in 1825, to pro-secute his researches in natural history, extends to the south-west of LakeZaisang towards the Ala-kul. ‡ We are thus acquainted, in the interior ofAsia, with a volcanic territory, the surface of which is upwards of 2,500

---

* Ho-chow, a city, now destroyed, was a league and a half to the east of Toorfan.† M. Rémusat calls the volcano of Pïh-shan, to the north of Kucha, the volcano of Bishbalik. Fromthe time of the Mongols in China, all the country between the northern slope of the Tëen-shan and thelittle chain of the Tarbagatay has been called Bishbalik.‡ I do not wish to express any doubt respecting the existence of the Ala-kul and the Alaktugul-noor,lakes in the vicinity of each other; but it appears singular to me, that the Tatars and Mongols, who|238| square leagues, and which is distant 300 or 400 leagues from the sea: it occu-pies a moiety of the longitudinal valley situated between the first and secondsystems of mountains. The chief seat of volcanic action seems to be in theTëen-shan. Perhaps the colossal Bokhda-ula is a trachytic mountain likeChimborazo. On the side north of the Tarbagatay and of Lake Darlay theaction becomes weaker; yet Mr. Rose and I found white trachyte along thesouth-western declivity of the Altaï, upon a bell-shaped hill at Ridderski, nearthe village of Butachikha. On both sides of the Tëen-shan, north and south, violent earthquakes arefelt. The town of Aksu was entirely destroyed by a convulsion of this kindat the beginning of the eighteenth century. Professor Eversman, of Casan,whose repeated travels have made us acquainted with Bokhara, was told by aTatar, who was a servant of his, well acquainted with the country betweenLakes Balkashi and Ala-kul, that earthquakes were very common there.In eastern Siberia, to the north of the fiftieth parallel, the centre of the circleof shocks appears to be at Irkutsk, and in the deep basin of Lake Baikal,where, on the Kiachta road, especially on the banks of the Jeda and theChekoy, basalt is found with olivine, cellular amygdaloid, shabassie, andapophyllite.* In the month of February 1829, Irkutsk suffered greatly fromviolent earthquakes; and in the month of April following, convulsions werealso felt at Ridderski, which were perceived at the bottom of the mines, wherethey were very severe. But this part of the Altaï is the extreme limit of thecircle of shocks; further to the west, in the plains of Siberia, between theAltaï and the Ural, as well as along the entire chain of the latter, no motionhas hitherto been observed. The volcano of Pïh-shan, the Aral-tubeh, to thewestward of the caverns of sal ammoniac of Khobok, Ridderski, and the por-tion of the Altaï which abounds in metals, are situated for the most part in adirection which but slightly deviates from that of the meridian. Perhaps theAltaï may be comprehended within the circle of the convulsions of the Tëen-shan, and the shocks of the Altaï, instead of coming only from the east, orfrom the basin of the Baikal, may also come from the volcanic country ofBishbalik. In many parts of the new continent it is clear that the circles ofshocks intersect each other, that is, the same country receives terrestrial con-vulsion periodically on two different quarters.

---

traverse these parts so often, and who have been questioned at Semipolatinsk, should only know theAla-kul, and assert that the Alaktugul-noor owes its existence to a confusion of names. M. Pansner,in his Russian map of Inner Asia, which may be implicitly relied on with regard to the country north ofthe course of the Ele, makes the Ala-kul (properly Ala-ghul, or “party-coloured lake”) communicatewith the Alaktugul by five channels. Possibly the isthmus which separates these lakes may be marshy,which causes them to be considered as one. Casim Bek, a professor at Casan, and who is a Persian bybirth, insists that tugul is a Tataro-Turkish negation, and that, therefore, Alaktugul signifies “thelake not variegated,” as Ala-tau-ghul implies “the lake of the variegated mountain.” Perhaps thenames of Ala-kul and Ala-tugul mean merely “lake near the Ala-tau mountain,” which stretches fromTurkestan to Zungaria. The small map published by the English missionaries of the Caucasus, doesnot contain the Ala-kul; there appears only a group of three lakes, the Balkashi, the Alak-tugul, andthe Koorgeh. The hypothesis, however, according to which the vicinity of large lakes produces, in theinterior of Asia, the same effect upon volcanos remote from the sea, as the ocean, is without founda-tion. The volcano of Toorfan is surrounded only by insignificant lakes; and, as it has been alreadyremarked, Lake Temoortu or Yssi-kul, which is less than double the extent of the Lake of Geneva, isthirty-three leagues from the volcano of Pïh-shan.—Humboldt. The Chinese maps represent the two lakes as one, having a mountain in the midst. This lake is calledAla-kul, its eastern portion bears the name of Alak-tugul-nor, and its western gulf that of She-bartu-kholay.—Klaproth. * Dr. Hess, associate of the Academy of Sciences of St. Petersburgh, who resided on the borders ofthe Baikal and to the south of the lake, from 1826 to 1828, gives us reason to expect a geological descrip-tion of a portion of the remarkable country which he traversed. He frequently observed at Verkhnei-Oudinsk granite alternating several times with conglomerates.|239| The volcanic territory of Bishbalik is to the eastward of the great dip of theold world. Travellers who have journeyed from Orenburg to Bokhara, relatethat at Sussak in the Kara-tau, which forms with the Ala-tau a promontory tothe north of the town of Taraz in Turkestan, on the edge of the dip, warmsprings spout up. On the south and on the west of the inner basin we findtwo volcanos still in activity; Demavend, which is visible from Tehran, andthe Seyban of Ararat,* which is covered with vitrified lava. The trachytes,porphyries, and thermal springs of the Caucasus are well known. On bothsides of the isthmus between the Caspian and Black Seas, naphtha springs andvolcanos of mud are numerous. The muddy volcano of Taman, of whichPallas and Messrs. Engelhard and Parrot have described the last fiery eruption,in 1794, from the reports of Tatars, is, according to the very sensible remarkof Mr. Eichwald, “a dependency of Baku, and of the whole peninsula ofAbsheron.” Eruptions take place where the volcanic forces encounter leastopposition. On the 27th November 1827, crackings and tremblings of theearth, of a violent character, were succeeded, at the village of Gokmali, inthe province of Baku, three leagues from the western shore of the CaspianSea, by an eruption of flames and stones. A space of ground, 200 toises longand 150 wide, burned for twenty-seven hours without intermission, and roseabove the level of the neighbouring soil. After the flame became extinct,columns of water were ejected, which continue to flow till the present hour.I am gratified at being enabled to state here, that Mr. Eichwald’s periplus ofthe Caspian Sea, which will soon appear, will contain some very importantphysical and geological observations, more particularly upon the connexion offiery eruptions with the appearance of naphtha-springs and strata of sal gem,on masses of calcareous rock hurled to considerable distances, on the elevationand sinking of the bed of the Caspian Sea, which still continue; on the pass-ing of black porphyry, partly vitrified and containing garnets (melapyres),through granite, red quartzose porphyry, very dark syenite, and calcareousspar, in the Krasnovodsk mountains washed by the bay of the Balkan, to thenorthward of the ancient mouth of the Oxus (Amoo-doria). We shall learnfrom the geognostic description of the eastern shore of the Caspian Sea, wherethe island of Chabekan discovers naphtha-springs the same as Baku and theisles between this town and Salian, what species of crystallized rocks are hiddenbeneath the rocks in horizontal strata in the peninsula of Absheron, where theaction of subterranean fire is always felt, and where it has not yet been ableto reach the open air. The porphyries of the Caucasus, which run fromW.N.W. to E.S.E., a position and a direction which I have already mentionedas the reason of the presumed connexion of this chain with the cleft of theTëen-shan, discover themselves again, traversing all the rocks nearly to thecentre of the great dip of the old world, to the east of the Caspian Sea, in themountains of Krasnovodsk and Kurreh. Recent researches and the traditionsof the Tatars inform us, that the existence of naphtha-springs has always beenpreceded by fiery eruptions. Several salt lakes on the two opposite shores ofthe Caspian Sea have a very elevated temperature; and blocks of sal gem, tra-versed by asphaltum, are formed, as Mr. Eichwald remarks with much shrewd-ness, “by the effect of a sudden volcanic action, as at Vesuvius,† in theCordilleras of South America and in Azarbaijan, or even under our own obser-

---

* The height of Ararat, according to Parrot, is 2,700 toises (17,718 feet); that of Elbourz, accordingto Kuppfer, 2,560 (16,800 feet) above the level of the ocean.† In an eruption of this volcano in 1805, M. Guy Lussac and I found small fragments of sal-gem inthe lava as it cooled. My Tatar itineraries likewise speak of sal-gem in the neighbourhood of a vol-canic mountain of the Tëen-shan, north of Aksu, between the station of Turpa-gad and Mount Arbab.|240| vation by the slow but continued action of heat.” M. L. de Buch has longdirected his attention to the connexion of the volcanic forces with the massesof enhedral sal-gem, which traverse so often and so many formations of hori-zontal strata. We have already seen that the circles of the terrestrial convulsions, ofwhich Lake Baikal or the volcanos of Tëen-shan are the centre, do not extendin western Siberia beyond the western declivity of the Altaï, and do not passthe Irtish or the meridian of Semipolatinsk. In the chain of the Ural, earth-quakes are not felt, nor, notwithstanding the rocks abound in metals, neitherbasalt or olivine is found, nor trachytes, properly so called, nor mineral springs.The circle of the phenomena of Azarbaijan, which includes the peninsula ofAbsheron, or the Caucasus, often extends as far as Kizlar and Astrakhan. It is the same on the border of the great hollow in the west. If we directour observation from the Caucasian isthmus to the north and north-west, wecome to the country of grand formations in horizontal and tertiary strata,which occupy southern Russia and Poland. In this region, the rocks ofpyroxene pierce the red free-stone of Yekaterinoslav, whilst asphaltum andsprings impregnated with sulphurous gas denote other masses concreted in theform of sediment. It may also be mentioned as an important fact, that inthe chain of the Ural, which abounds so much in serpentine and amphibole,and which serves as a boundary between Europe and Asia, a true amygdaloidalformation appears at Griasnushinskaia, towards its southern extremity. We shall content ourselves here with observing, with reference to the inge-nious opinions recently promulgated by M. Elie de Beaumont, respecting therelative age and the parallelism of systems of contemporary mountains, thatin the interior of Asia likewise, the four grand chains which run from east towest are of a totally different origin from the chains which lie in a directionnorth and south, or N. 30° W., and S. 30° E. The chain of the Ural, theBelor, or Beloor-tag, the Ghauts of Malabar, and the Kingkhan, are probablymore modern than the chains of the Himalaya and the Tëen-shan. Thesystems of different epochs are not always separated from each other by anyconsiderable space, as in Germany, and in the greater part of the new conti-nent. Frequently, chains of mountains, or axes of exaltation, of dissimilardirections, and belonging to epochs totally different, are nearly approximatedby nature; resembling so far the characters on a monument which, crossingdifferent ways, were engraved at different periods, and carry intrinsic marks oftheir own date. Thus, in the south of France, are seen chains and undu-lated swellings, some of which are parallel to the Pyrenees and others to thewestern Alps. The same diversity of geological phenomena is apparent in thehigh land of Central Asia, where isolated portions appear as it were sur-rounded and enclosed by subdivisions, in parallel lines, of the systems ofmountain.