CASPIAN SEA. The Caspian Sea, which was known under that name to the Greeks and Romans (Herodotus having given a generally accurate account of it, stating that it is an inland sea having no connection with the ocean), is the largest of those salt lakes or closed inland seas which may be considered as " survivals " of former oceanic areas ; and it is the one whose physical and biological conditions have been most fully studied. These conditions are in many respects extremely peculiar; and tolerably certain conclusions of great interest may be drawn from them, in regard to the past history of the large extent of low steppes that lie—chiefly in Asia, but partly in Europe also—to the east, north, and west of its present area. These will be most fitly considered after a general survey has been taken of the existing basin of the Caspian, and of its relations to the surrounding land.

The general form of the Caspian may be described as a broad band, with sides almost straight and parallel, except near its northern end, where it turns sharply round to the east. The general direction of its axis is about N.N.E. and S.S.W., ranging from lat. 47° 20' to 36° 40' N.,—its most northerly point nearly coinciding with the mouth of the River Ural, and its most southerly being about half-way between the towns of Reshd and Astrabad. The distance in a straight line between these two points is about 740 miles in a straight line. The average breadth of its middle portion is about 210 miles, but the eastern extension of its northern portion into the Bay of Mertvy Kultuk increases the width of that part to 430 miles ; and its southern portion also widens to nearly 300 miles. The total area is estimated at about 180,000 square miles.

The most important fact in the physical geography of the existing Caspian is that its surface is 84 feet below that of the Black Sea, which may be considered as not differ-ing much from the general oceanic level.

Sketch Map of Caspian Sea.

The basin of the Caspian may be considered as consist-ing of three distinct parts,-—the northern, the middle, and the southern. The northern portion is extremely shallow,— its bottom, which is nowhere more than 50 feet below the surface, being a continuation of the almost imperceptible slope of the steppe, so that there is no definite shore-line. It is into this portion that the Volga, the Ural, and the Kuma discharge themselves ; and the deposit of alluvium which these rivers bring down is gradually raising its bottom, and will in time convert it into a salt marsh. Along the north-western border of this basin, from the delta of the Volga to that of the Kuma, a space of 250 miles, the shore is gashed with thousands of narrow channels, termed limans, from 12 to 30 miles in length, separated by chains of hillocks called bugors, which pass landwards into the level ground of the steppes. In the neighbourhood of the mouths of the Volga and Kuma, the excess of water which these rivers bring down at the tune of the melting of the snows passes into these channels, and tends to keep them open ; so that when the inundation is over, the sea again flows into them. But along the inter-vening part of the coast, the channels, like the intervening hillocks, are not continuous, but form chains of little lakes, separated by sandy isthmuses. Although these channels run nearly parallel to each other, yet they have a some-what fan-like arrangement; their centre of radiation being the higher part of the isthmus which separates the basin of the Caspian from the north-east portion of the Black Sea, —a fact, as will be seen hereafter, of no small significance. The coast-line of the Bay of Mertvy Kultuk, on the other hand, is formed by a chain of low calcareous hills, constituting the rampart of the plateau of Ust Urt or Turkoman Isthmus, which divides the Caspian from the Sea of Aral; and it is between head-lands of this high plain that the long extension of this bay termed the Karasu (or Black Water) runs inland, the town of Novo Alexandrovsk being situated near its entrance.

The northern basin of the Caspian may be considered to terminate on the west side with the Bay of Kuma, and on the east with the hilly peninsula of Mangishlak, on which the town of Novo Fetrovsk is situated. To the south of the line joining these points, in the parallel of 44° 10' N. lat., the western shoreline is higher, and the water deepens considerably,-—thus forming the middle basin of the Caspian, which may be considered to extend as far south as Cape Apsheron, the south-eastern termina-tion of the great Caucasian range. This middle basin receives the large river Terek, which discharges itself by several mouths (some of them entering the Bay of Kuma) through an alluvial delta; and several smaller streams flow into it from the slopes of the Caucasian mountains through the low plain which intervenes between their base and the border of the Caspian. Near the most consider-able of these, the Kabir Yalama, a rocky spur of the Caucasus comes down nearly to the sea ; and a narrow pass is thus formed, which has been fortified from very ancient times, being formerly known as the Albanice or Caspice Pylos, and now as the Pass of Derbend, this being a small town built on the declivity in which the range terminates. The eastern shore of this portion of the Caspian is formed by the plateau of Ust Urt, or "high plain," a very remarkable plateau from 550 to 727 feet above the level of the Caspian, which extends from its eastern shore to the sea of Aral, rising abruptly from t oth seas, aud ranging about 400 miles in the north and south direction; its north and south borders are formed by a precipitous face or cliff, which has much the appearance of an ancient sea-margin. As it is composed of later Tertiary strata, its elevation must have occurred at a time not geologically remote. The headlands of the Ust Urt form an abrupt coast-line along the northern part of the eastern border of the middle basin, with occasional bays into which several small streams from the plateau discharge themselves. Further south, however, the plateau recedes, and the land shelves off more gradually; and here an extensive but shallow basin presents itself (of which more will presently be said) almost entirely cut off from that of the Caspian, termed the Karaboghaz, or Black Culf. To the south of this the coast-line rises again ; and a peninsula is formed by an extension of the Balkan Mountains, which may be considered as forming the southern termination of the middle basin. Except along the shore-lines, the depth of this basin everywhere excedes that of the northern,—being greatest in its middle portion, where over a small area it reaches 400 fathoms, whilst it shallows again towards the south, where there is a sort of ridge between Cape Apsheron and the Balkan peninsula, at the average depth of 30 fathoms, that separates it from the southern basin.

The southern basin ranges from the Balkan Peninsula on the east and Cape Apsheron on the west to the shore-line formed by the base of the great Elburz range of mountains, which curves round its low and swampy border, from the mouth of the Kur to Astrabad, at an average dis-tance of about 40 miles, rising in the peak of Sawalan near Tabreez to 15,800 feet, and in the snow-capped summit of Demavend, on whose southern slope Teheran is situated, to 18,600 feet. These mountains are composed of granite and porphyry, and are covered with recent volcanic deposits. South of Cape Apsheron, this basin receives the large river Kur, which drains the southern slopes of the Caucasian range ; and this is joined, at no great dis-tance from its mouth, by another large river, the Aras or Arax (the ancient Araxes), which forms the boundary between Russian Trans-Caucasia and Persia. The joint channel discharges its water by several mouths, part of them opening into the Gulf of Kizil-Agatch, which is the most considerable extension of the southern basin. From the mouths of the Kur to the Gulf of Enzeli, which resembles the Karaboghaz on a smaller scale, there is no considerable stream ; but not far to the east of the town of Reshd of which Enzeli is the port, the Sefid or White River discharges itself, this being formed by the confluence of the Kizil-Uzen with another considerable river, the two together draining a large portion of the slopes of the western division of the Elburz range, and of its extension towards the Caucasus. The southern border of the Caspian, between the mouth of the Sefid and Astrabad, receives numerous small streams from the northern slopes of the Elburz, but no considerable river; the Bay of Astrabad, however, receives at its northern end the Attruk, a river of considerable importance, which drains an extensive valley enclosed by the mountain ranges that form the southern border of the desert plains of Khiva. On the eastern coast, opposite to the Gulf of Kizil-Agatch, are the Balkan Bay and the Adji-Bojur Bay, which lie between extensions of the Balkan Mountains. One or both of these bays, it may now be pretty confidently stated, formerly received the mouths of the ancient Oxus (now Amou-Daria), when it discharged itself into the Caspian, instead of into the Sea of Aral; and there is further reason to believe that a communication here at one time existed between the Caspian and the Sea of Aral, through a furrow which lies along the southern border of the Ust Urt, and which terminates in what was formerly known as the Gulf of Aboughir, a southern extension of Lake Aral now dried up. The depth of the southern basin of the Caspian is for the most part considerable, ranging in its central portion between 300 and 500 fathoms.

Drainage Area.—The drainage-area of the Caspian is much more extensive on the north and west than on the east and south. The Volga is estimated to drain an area of 527,500 square miles, and the Ural an area of 85,000 square miles,—these two rivers together probably bringing down more water than the Danube and the Don pour into the Black Sea. When to these we add the Kuma, the Terek, the Arax and Kur, the Sefid, and the Attruk, it is obvious that the total amount of river water annually discharged into the basin of the Caspian must be almost, if not quite, the equal of that which is discharged into the basin of the Black Sea. Yet the whole amount of fresh water returned by rain and rivers to the basin of the Caspian is only sufficient to compensate for the loss by evaporation from its surface,—as is shown by the fact that its present level remains constant, or, if it changes at all, rather sinks than rises. Now that the level of the Caspian was formerly about the same as that of the Black Sea, although at present 84 feet below it, is shown by the erosion of the rocks that formed the original sea-shore of the southern basin, which, at the height of from G5 to 80 feet above the present level, have been furrowed out into tooth-shaped points and needles ; and if the water were again to rise to that level, it would overflow many hundred thousand square miles of the southern steppes, extending the area of the basin as far as Saratov. Now supposing the Caspian to have been formerly in communication with the general oceanic area (which will be hereafter shown to be almost a certainty), a reduction of its level and a contraction of its area would follow as a necessary consequence, whenever that communication was cut off. For, as the evaporation-area would have then been much greater than it is at present, whilst the drainags-area would have been the same, there must have been a great excess of loss by evaporation over the water returned by rain and rivers ; and this excess, pro-ducing a reduction of level, would have reduced the area of the northern shallow portion, until it contracted itself within its present limits. That this reduction was rapid, is indicated by two sets of facts;—first, the absence of any erosion of rocks between the level of the old erosion and the present level; and second, the fan-like arrangement of the limans and intervening bugors on the north-west shore, which makes it difficult to suppose that these channels can have been formed except by the furrowing of the soft soil during the sinking of the water, corresponding to that which is seen on a small scale on the muddy banks of a reservoir in which the water is being rapidly lowered by the opening of a sluice-gate.

Salinity of the Water of the Caspian.—It might have been anticipated that such a reduction in the volume of the Caspian water as must have taken place in this lower-ing of its level, would have shown itself (as in the Dead Sea) in an increase of its salinity ; whereas the fact is that the proportion of salt in the water of the Caspian, though varying in different parts of the basin, and also at different seasons, is generally much less than the proportion in oceanic and even in Black Sea water.





In the northern portion, whose shallowness causes the enormous amount of fresh water brought down by the Volga, the Ural, and the Terek to exert the greatest diluting influence, the salinity is so slight (especially when the ordinary volume of these rivers is augmented by the melt-ing of the snows) that the water is drinkable, its specific gravity not being higher than 1_016. In the central and southern basins, on the other hand, which contain a body of salt water too large to be thus affected, the salinity is stated by Von Baer to be about one-third that of ordinary sea-water, the average sp. gr. being about 1 '009. This re-duction from what may be presumed to have been its original amount seems fully explained by Von Baer, who traces it to the number of shallow lagoons with which the basin is surrounded, every one of them being a sort of natural salt-pan for the evaporation of the water and the deposit of its saline matter in the solid form. The process may be well studied in the neighbourhood of Novo Petrovsk, where what was formerly a bay is now divided into a large number of basins presenting every degree of saline con-centration. One of these still occasionally receives water from the sea, and has deposited on its banks only a thin layer of salt; a second, likewise full of water, has its bottom covered by a thick crust of rose-coloured crystals like a pavement of marble ; a third exhibits a compact mass of salt, on which are pools of water whose surface is more than a yard below the level of the sea; and a fourth has lost all its water by evaporation, the stratum of salt left behind being now covered with sand. A similar concentration is taking place in the Karasu ; for notwith-standing the proximity of the mouths of the Ural and Volga, the proportion of salt there rises to such a degree (the sp. gr. being l'057j that animal life is almost, if not entirely, suppressed. In the Peninsula of Apsheron, again, there are ten salt lakes, from one of which 10,000 tons of salt are annually obtained.

This process of elimination goes on, however, upon its greatest scale in the Karaboghaz, whose nearly circular shallow basin, about 90 miles across, is almost entirely cut off from the Caspian by a long narrow spit of land, communicating with it by a channel which is not more than about 150 yards broad and 5 feet deep. Through this channel a current is stated by Von Baer to be continually running inwards (during the summer months, at least) at an average rate of three miles per hour; this rate being accelerated by westerly and retarded by easterly winds, but never flowing at less than a mile and a half per hour. The navigators of the Caspian, and the Turkoman nomads who wander on its shores, struck with the constant and unswerving course of this current, have supposed that its waters pass down into a subterranean abyss, through which they reach either the Persian Gulf or the Black Sea,— an hypothesis for which there is not the least foundation, and which is directly negatived by comparison of levels. The current is really due to the indraught produced by the excess of evaporation from the surface of the basin, which is exposed to every wind and to intense summer heat, and which receives very little return from streams. The small depth of the bar seems to prevent the return of a counter-current of highly saline water, such as, in the Strait of Gibraltar, keeps down the salinity of the Mediter-ranean (see MEDITERRANEAN), none such having been detected by the careful investigations of Von Baer. And thus there is a progressively increasing concentration of the contents of the basin of the Karaboghaz, so that seals which used to frequent it are no longer found there, and its borders are entirely destitute of vegetation. Layers of salt are being deposited on the mud at the bottom ; and the sounding-line, when scarcely out of the water, is covered with saline crystals. Taking the lowest estimates of the salinity of the Caspian water, of the width and depth of the channel, and of the speed of the current, Von Baer has shown that the Karaboghaz daily withdraws from the Caspian the enormous quantity of 350,000 tons of salt.

Now, if such an elevation of the bar were to take place as should cut off the basin of the Karaboghaz from that of the Caspian, the former would quickly diminish in ex-tent, and the concentration of its waters would cause an increased deposit of salt to take place on its bottom. According to the proportion between the evaporation from the area so reduced and the return of fresh water by rain and streams, the Karaboghaz would either be converted into a shallow lake of extremely salt water, or into a salt marsh, or might altogether dry up and disappear, leaving behind it a thick bed of " rock-salt" resembling the deposits contained in the Saliferous strata of various geological periods. These several conditions all obtain at the present time in different parts of the great area of the steppes of Southern Russia. There are several small salt lakes which receive water enough from rain, snow, and streams to compensate for the loss they sustain by evapora-tion ; these especially occur in the Kirghiz steppes, which lie to the north-east of Astrakhan, between the Volga and the Mongodjar Hills that form the southern extremity of the Ural range; the most notable of them being Lake Elton, which lies about 200 miles to the north of the present border of the Caspian, and from which large quantities of salt are annually procured. There are large tracts of these steppes, again, which are alternately muddy and white with salt, according as they are moistened by rain or dried up by the heat of the sun; one of these, lying between Lake Elton and the River Ural, occupies a depressed area about 79 feet below the present level of the Caspian, and more than 160 feet below that of the Black Sea. Everywhere the sand of these steppes contains an admixture of salt; and there are various local accumula-tions of salt, often associated with marl, having shells and fish-bones imbedded in them, and thus clearly marking the sites of lakes which survived for a time the reduction of level and recession of the northern border of the Caspian, but which are now entirely dried up.

Climate of the Caspian.—The temperature of the Caspian area is remarkable for its wide range, both geographical and seasonal,—the difference between the mean winter temperatures of its northern and southern extremities being very great, whilst over its whole extent a high summer temperature prevails. The January isotherm of 15° skirts its northern border ; that of 20° crosses it at the line of division between its northern and middle basins; that of 30° crosses it between its middle and its southern portions ; and that of 40° skirts its southern border. Thus between the mean winter temperatures of the northern and southern extremities of the Caspian there is a geographical range of 25°. These means, however, do not indicate the extremely low temperatures which prevail over the whole region of the steppes during the prevalence of north-east winds; the thermometer then sinking to - 20°, or even lower, on the level areas, whilst on the elevated plateau of Ust Urt a temperature of - 30° is nothing remarkable. The July isotherm of 75°, again, crosses the middle basin of the Caspian, nearly coinciding with the January isotherm of 25°; and that of 80° skirts the southern border of the sea, nearly coinciding with the January isotherm of 40°; so that the mean annual range is 50° over the northern portion of the basin, and 40° at its southern extremity. These sum-mer means give no truer indication than the winter means of the extremes of temperature occasionally reached ; thus Major Wood saw the thermometer mark 110° in the shade on the bank of the Oxus, recalling to his recollection the intense heat of Annesly Bay in the Abyssinian expedition.

The shallow northern basin of the Caspian is frozen during the entire winter, and the ice sometimes extends to the middle basin ; the deep southern basin on the other hand, is never frozen over.

The prevalent winds of the Caspian are the south-east, which usually blow between October and March, and the north or north-west, which are common between July and September. They sometimes continue with great violence for days together, rendering navigation dangerous, and inundating the shores, wherever these are low and flat, against which they blow. The same cause tends to disturb the general level of the water, svhich is raised or lowered by from 4 to 8 feet at the north or the south end of the basin, according to the direction of the wind; and when this changes suddenly, as it often does, strong currents are generated. There are no perceptible tides in the Caspian ; and the changes of level occasionally observed without any wind to account for them seem attributable on the one hand to inequality between the evaporation and the return of water by rain and rivers, and on the other to differences in atmospheric pressure between one part of the area and another, such as alter the level of the Baltic (see BALTIC). It was stated by Colonel Monteith (Royal Geographical Journal, vol. iii.), that during his residence in that part of Asia from 1811 to 1828, the Caspian, " as well as every other lake in Persia, had sensibly decreased in depth ;" but according to the information given him by the inhabitants of Enzeli, there is a rise and fall of several feet in periods of thirty years; and Von Baer, by whom the question was carefully examined, could not obtain an evidence that any continuous reduction of level is at present in progress. There is indeed reason to believe that the level of the Caspian was once much lower than it is at present; for at Derbend, whose foundation is assigned to Alexander, masonry has been ascertained to exist at a depth of 50 feet below the present surface level; and as it is recorded that the Khorasmians made an offer to Alexander to conduct his army to Colchis, it would seem as if the ridge at the southern end of what is now the middle basin could then have been crossed dry-shod. This does not appear very improbable, if, as ancient geographers and historians ex-plicitly state, the Volga flowed in their time, not into the Caspian, but into the Sea of Azoff,—a condition which seems to have persisted as late as the 5th century. The channel of its lower part would then have been that of the present River Don, which at one part approaches so closely to that of the Volga, that the two are united by a canal of less than 50 miles' length.

Fauna of the Caspian.—-The animal life of this inland sea presents a remarkable admixture of marine and fresh-water types. The presence of seals and herrings seems an unmistakable indication of its former communication with the ocean,—and this rather northwards with the Polar Sea than westwards through the Black Sea and Mediterranean. Again, the Caspian abounds in salmon, a fish that may be considered essentially marine, though resorting to rivers to breed. And among its most notable and valuable inhabi-tants are four species of Sturionidoz—the sturgeon, the sterlet, the sevriouga, and the beluga—which are essentially estuary fish, ascending rivers from their mouths. The fisheries are extremely valuable,—a very large amount of fish being salted for transmission to distant parts, while the Sturioiiidse afford the principal supply of caviare (prepared from their roe) and of isinglass (their swim-bladders cut into strips) for the whole world. The Molluscan fauna is not by any means proportionally numerous or varied. It principally consists of these wide-spread marine forms which are able to adapt themselves to a variety of conditions, and especially to a reduction in the salinity of the waters they inhabit, which (as in the parallel case of the Baltic) tends to dwarf the races of mollusks subjected to it.





Naphtha and Petroleum Springs.—Various parts of the shore of the Caspian abound in naphtha and petroleum. This is especially the case with the Peninsula of Apsheron, and with the Island of Tchilehon or Naphthalia, which lies near the opposite coast, off the Bay of Balkan. The whole soil of Apsheron is said to be saturated with naphtha, which rises wherever a hole is bored ; and round the town of Baku there are nearly a hundred bituminous springs, from many of which considerable supplies of naphtha are drawn. Some of these are constantly burning; and one of them, termed the "burning field," was formerly a celebrated " shrine of grace " to the Ghebers or Parsees, multitudes of pilgrims resorting to it, as Mahometans do to Mecca.

Former Extent of the Caspian.-—From what has been stated, there can be no reasonable doubt (1) that the area of the Caspian must have formerly been much more exten-sive than at present, and (2) that it must at some time have had free communication with the ocean. It was long since pointed out by Pallas that the presence of salt lakes, dry saline deposits, and sea shells of the same species as those now inhabiting the Caspian, over a very large extent of the steppes to the east, north, and west of the present basin, can only be accounted for on such an hypothesis; and he traced out what may probably be regarded as a northern shore-line, along the base of the Mongodjar hills. Further, the fauna of the Caspian corresponds so remark-ably with that of the Black Sea on the one side, and with that of the Sea of Aral on the other, that it can scarcely be doubted they were formerly in free communication with one other; and the lines of this communication can be pretty certainly traced out by the peculiar lowness of the levels. Thus between the Caspian and the Black Sea, or rather the Sea of Azoff, it would have lain across the low-lying portion of the steppe, which is at present a receptacle for the drainage of the surrounding area, forming the long and shallow Lake Manytsch. And between the Caspian and the Aral Sea it probably followed both the northern and the southern borders of the Ust Urt, which would have thus formed an insulated platform. If the elevation of level were sufficiently great to raise the water in Lake Aral to the height which it had in former times (as is shown by various clearly discernible landmarks), it would have overflowed a large area to the south also ; and of this again, some parts of the coast-line are traceable. A very slight further elevation would bring it into com-munication with the Arctic Sea.

There is much to support this view, not only in the writings of ancient geographers and in the incidental notices which have been gleaned from the records of early travel, but also in the physical relations of the three basins now forming separate seas. For if the outlet of the Bosphorus were closed, the progressive accumulation of the excess of fresh water which at present escapes from the Black Sea by that channel (see BLACK SEA), would in no long time cause an overflow into the basin of the Caspian ; since, although the Black Sea proper is separated from the southern portion of the Caspian by the mountainous region of the Caucasus, yet between the Sea of Azoff and the northern portion of the Caspian there is only the low steppe inhabited by the Don Cossacks and the Kalmucks ; and, according to Major Wood, au elevation of the Black Sea to no more than 23 feet above its present level would cause it to overflow into the basin of the Caspian by the line of the Manytsch. The continuance of such an overflow would in time raise the Caspian to the same level, and would thus produce (as already shown) an immense extension of its area. For although that area would be prevented by the interposition of the Ust Urt from directly spreading towards the Sea of Aral, yet a continued rise of the Caspian would enable its water to find its way along the north and south of that plateau, so as to extend itself over a large part of the Aralo-Caspian depression, including what is now the isolated Sea of Aral, and completely surrounding the Ust Urt, which would rise as an island in the midst of it. A rise of 158 test above the sea would bring it up to the level of the Sea of Aral; and it is considered by Major Wood that a further rise of about 62 feet, making 220 feet in all, of which there is distinct evideuce in horizontal water-marks, would cause this Asiatic Mediterranean to overflow its northern boundary into the watershed of the Tobol, one of the tributaries of the Obi, through which its water would be discharged into Polar Sea. And it is a fact of no little interest, that the existence of such a communication between the Aralo-Caspian basin and the Northern Ocean was most distinctly affirmed by Strabo and other ancient geographers.

Now, as there is strong reason to suspect, from the evidence of recent volcanic change in that locality, that the opening of the Bosphorus took place within a period which, geologically speaking, was very recent, it does not seem at all improbable that this event (which some writers identify with the deluge of Deucalion) was the commence-ment of a series of changes, by which the "Asiatic Mediter-ranean " came to be divided into the three separate basins which now constitute its " survivals." Supposing, then, the level and extent of this great inland sea to have been formerly such as just described, the effect of the opening of the Bosphorus would of course be to lower its surface and to contract its area. So long as the Caspian retained its communication with the Black Sea, it would remain at the general oceanic level,—the excess of the river drainage into the western basin (including that of the Volga) supplying what was deficient in the eastern. But if, by a slight elevation of the intervening isthmus, this communication were cut off, the excess of evaporation over the Caspian area (which would have been previously separ-ated from the Aral Sea) would have reduced its level all the more rapidly, when the Volga, which now furnishes its principal supply, was not one of its affluents; and we can thus account for that depression of its surface much below its present level, which seems to have existed in the time of Alexander. By the subsequent deflection of the lower part of the Volga from the Sea of Azoff into the basin of the Caspian, the level of the latter would have been raised again, and its area extended, until that equality came to be established between the evaporation-loss and the river-supply which obtains at the present time.

The changes produced in the eastern portion of the " Asiatic Mediterranean " by the opening of the Bosphorus would have been yet more considerable. In consequence of the greater elevation of the Aralian area, a comparatively slight reduction of level would have served to lay dry a large proportion of it, and to cut off all communication with the Caspian except by a narrow outlet; and the mainte-nance of the level in what thenceforth existed as an isolated basin would depend upon the relation between its evapora-tion and its river-supply. This supply is mainly derived from two principal rivers :—the Syr Darya (the ancient Jaxartes), which takes its rise in the high valleys to the east of Kokand, flows through that khanate in a westerly direction, and now, after passing Khojend, turns suddenly northwards, and then to the north-west, and finally discharges itself into the Sea of Aral near its northern extremity ; and the Amou Darya (the ancient Oxus), which rises in the plateau of Pamir and the high valleys of the Hindu Kush—-its sources being in close proximity to those of the Indus,—-and then, rapidly descending into the great Turcoman Plain, at present continues onwards in a north-west direction to Khiva, after passing which it flows into the southern end of the Aral Sea. A large proportion of the water of both these rivers, however, is withdrawn from them in the latter part of their course,—partly by percolation through the sandy soil (there being no defined river-beds), and partly through the extensive irrigation by which the dwellers along their course render productive the otherwise barren land. The supply which they bring to the existing Aral Sea does not suffice to keep it up to its present level, as is proved by recent exact observation ; and it is clear, there-fore, that even the whole body of water they bring down could not have maintained the level of the far larger area over which it must have originally spread, and that this must consequently have been rapidly reduced. Now there is very distinct evidence, both historical and physical, that the Oxus, wittin a comparatively recent period, flowed westwards across the desert of Khwarezm, near the parallel of 39° N.,and discharged itself into the Caspian basin through the Balkan Bay. And there is also much reason to believe that the Syr Darya also, or a considerable part of it, once flowed westwards where it now takes its northerly bend, crossed the desert of Kizzel Koom, and finding its way into the Uzboy furrow which skirts the southern border of the Ust LTrt, poured its water into the Caspian. Thus the area now occupied by the Aral Sea, deprived of its two main affluents, must either have entirely dried up, or have been reduced to a salt marsh, until a change in their course filled its basin to somewhat above its present level.

Thus it would appear that the condition of the Aralo-Caspian area must have undergone very considerable alterations within the historic period ; and it is main-tained by Major Wood (The Shores of Lake Aral, 1876) —who has recently investigated the whole subject both physically and historically,—that these alterations may have taken place without any such geological disturbances as some physical geographers have supposed necessary. Some of these changes, he argues, may be fairly attributed to human agency, which can be shown to have exerted a con-siderable influence, not only on the amount of water carried along by the two great rivers of the Aralian area, but even, it is probable, on their course.

But the hypothesis of an Asiatic Mediterranean will not of itself account for the facts which indicate that its basin was formerly in free communication with the general oceanic area. For as the water of this great inland sea must have risen to 220 feet above its present level, to have escaped across the ridge that formed its northern boundary, into the watershed of the Obi, only an outward or overflow current could have passed that ridge, and no sea-water could have entered the basin from the outside. Hence the saltness, not only of the water of the Caspian and Aral seas, but of that of the numerous lakes still remaining in the most depressed spots formerly covered by the Asiatic Mediterranean, together with the large admixture of salt in the sand that covers what is now its dried-up bed, can only be accounted for on the supposition that this Asiatic Mediterranean was itself a " survival" of an extension of the oceanic area properly so called,— retaining not only much of its salinity, but a portion of its characteristic fauna. And this conclusion derives confirma-tion from the fact (ascertained by the researches of the Russian naturalist, Bogdanoff) that the polar fauna may be traced through the succession of salt lakes lying to the north of the Aral Sea, and that its proportion increases as we approach the Polar Ocean. Now it is certain that the whole of this area was submerged during the Cretaceous period,—what is now the North Atlantic Ocean having then extended (with little interruption of its continuity) from the American continent to Siberia. The general rise of the Asiatic and European part of its sea-bed, which took place at the end of the Secondary period, may not improbably have cut off the Asiatic Mediterranean, enclos-ing it within the limits already pointed out, and at the same time elevating it above the general level of the sea. Under these conditions it would have for some time retained much of its original saltness ; and this seems the explana-tion of the fact that the marine shells which are now scattered over the ancient sea-bed, and are occasionally found accumulated in masses, are much larger than the shells of the same species now inhabiting the weakly-saline Caspian. If the river-drainage into this area were more than sufficient to equalize its loss by evaporation, it may have remained without any essential alteration of its con-ditions, until the opening of the Bosphorus initiated a new succession of changes, which in the case of the Aral Sea appear to be still in progress. In this later succession, such alterations in the courses of the two great rivers of the Aralian area as are distinctly indicated by historical as well as physical evidence must have exerted a very impor-tant influeuce ; and a due appreciation of the results of these alterations seems (as already shown) to afford the clue to the differences in the accounts that have been given of the Aral Sea within the historic period.

Bibliography.—In addition to the writings of Professor Von Baer and Major Wood, of which special mention has already been made, the student of the physical geography of the Aralo-Caspian area should refer to the discussion between Sir Roderick Murchison and Sir Henry Rawlinson in the Journal of the R. Geog. Soc. for 1867; the paper of Professor Eichwald in the same journal; the Aralseefrage of Roesler (Vienna, 1873) ; and the learned Das alte Bett des Oxus of Professor Goeje (Leyden, 1875). (w. B. C.)


Footnotes

See especially the "Kaspisehe Stuclien " of Prof. Von Baer, in the St Petersburg Bull. Sci., and in Erman's Archio. Buss., 1855, 1856.

It was here that the expedition of Peroffsky, in 1839-40, lost all hut 200 of its 12,800 camels.