silted up by the sedimentary deposits brought down by the rivers Volga, Ural and Terek. The western shore, from the delta of the Volga to the mouth of the Kuma, a distance of 170 m., is gashed by thousands of narrow channels or lagoons, termed limans, from 12 to 30 m. in length, and separated in some cases by chains of hillocks, called bugors, in others by sandbanks. These channels are filled, sometimes with sea-water, sometimes with overflow water from the Volga and the Kuma. The coastline of the Gulf of Mortvyi Kultuk on the north-east is, on the other hand, formed by a range of low calcareous hills, constituting the rampart of the Ust-Urt plateau, which intervenes between the Caspian and the Sea of Aral. On the south this gulf is backed by the conjoined peninsulas of Busachi and Manghishlak, into which penetrates the long, narrow, curving bay or fjord of Kaidak or Kara-su. (2) South of the line joining the Bay of Kuma with the Manghishlak peninsula, in 44° 10' N. lat., the western shore is higher and the water deepens considerably, being over one-half of the area 50 fathoms, while the maximum depth (between 41° and 42° N. lat.) reaches 437 fathoms. This, the middle section of the Caspian, which extends as far as the Apsheron peninsula, receives the Terek and several smaller streams (e.g. Sulak, Samur), that drain the northern slopes of the Caucasus. At Derbent, just north of 42° lat., a spur of the Caucasus approaches so close to the sea as to leave room for only a narrow passage, the Caspiae Pylae or Albanae Portae, which has been fortified for centuries. The eastern shore of this section of the sea is also formed by the Ust-Urt plateau, which rises 550 ft. to 750 ft. above the level of the Caspian; but in 42° N. lat. the Ust-Urt recedes from the Caspian and circles round the Gulf of Kara-boghaz or Kara-bugaz (also called Aji-darya and Kuli-darya). This subsidiary basin is separated from the Caspian by a narrow sandbar, pierced by a strait 11 m. long and only 115 to 170 yds. wide, through which a current flows continuously into the gulf at the rate of 11 to 5 m. an hour, the mean velocity at the surface being 3 m. an hour. To this there exists no compensating outflow current at a greater depth, as is usually the case in similar situations. The area of this lateral basin being about 7100 sq. m., and its depth but comparatively slight (31 to 36 ft.), the evaporation is very appreciable (amounting to 3.2 ft. per annum), and sufficient, according to von Baer, to account for the perpetual inflow from the Caspian. South of the Kara-Boghaz Bay the coast rises again in another peninsula, formed by an extension of the Balkhan Mountains. This marks (400 N. lat.) the southern boundary of the middle section of the Caspian. This basin may be, on the whole, considered as a continuation of the synclinal depression of the Manych, which stretches along the northern foot of the Caucasus from the Sea of Azov. It is separated from (3), the southern and deepest section of the Caspian, by a submarine ridge (30 to 150 fathoms of water), which links the main range of the Caucasus on the west with the Kopet-dagh in the Transcaspian region on the east. This section of the sea washes on the south the base of the Elburz range in Persia, sweeping round from the mouth of the Kura, a little north of the Bay of Kizil-agach, to Astarabad at an average distance of 40 m. from the foot of the mountains. A little east of the Gulf of Enzeli, which resembles the Kara-boghaz, though on a much smaller scale, the Sefid-rud pours into the Caspian the drainage of the western end of the Elburz range, and several smaller streams bring down the precipitation that falls on the northern face of the same range farther to the east. Near its south-east corner the Caspian is entered by the Atrek, which drains the mountain ranges of the Turkoman (N.E.) frontier of Persia. Farther north, on the east coast, opposite to the Bay of Kizil-agach, comes the Balkhan or Krasnovodsk Bay. In the summer of 1894 a subterranean volcano was observed in this basin of the Caspian, in 38° 10' N. lat. and 52° 37' E. long. The depth in this section ranges from 300 to 500 fathoms, with a maximum of 602 fathoms.
Drainage Area and Former Extent.—The catchment area from which the Caspian is fed extends to a very much greater distance on the west and north than it does on the south and east. From the former it is entered by the Volga, which is estimated to drain an area of 560,000 sq. m., the Ural 96,000 sq. m., the Terek 59,000 sq. m., the Sulak 7000 sq. m., the Samur 4250 sq. m.; as compared with these, there comes from the south and east the Kura and Aras, draining the south side of the Caucasus over 87,250 sq. m., and the Sefid-rud and the Atrek, both relatively short. Altogether it is estimated (by von Dingelstedt) that the total discharge of all the rivers emptying into the Caspian amounts annually to a volume equal to 174.5 cub. m. Were there no evaporation, this would raise the surface of the sea 51 ft. annually. In point of fact, however, the entire volume of fresh water poured into the Caspian is only just sufficient to compensate for the loss by evaporation. Indeed in recent times the level appears to have undergone several oscillations. From the researches of Philippov it appears that during the period 1851–1888 the level reached a maximum on three separate occasions, namely in 1868–1869, 1882 and 1885, while in 1853 and 1873 it stood at a minimum; the range of these oscillations did not, however, exceed 3 ft. 61 in. The Russian expedition which investigated the Kara-boghaz in 1896 concluded that there is no permanent subsidence in the level of the sea. In addition to these periodical fluctuations, there are also seasonal oscillations, the level being lowest in January and highest in the summer.
The level of the Caspian, however, was formerly about the same as the existing level of the Black Sea, although now some 86 ft. below it. This is shown by the evidences of erosion on the face of the rocks which formed the original shore-line of its southern basin, those evidences existing at the height of 65 to 80 ft. above the present level. That a rapid subsidence did take place from the higher level is indicated by the fact that between it and the present level there is an absence of indications of erosive energy. There can be no real doubt that formerly the area of the Caspian was considerably greater than it is at the present time. Nearly one hundred and fifty years ago Pallas had his attention arrested by the existence of the salt lakes and dry saline deposits on the steppes to the east of the Caspian, and at great distances from its shores, and by the presence in the same localities of shells of the same marine fauna as that which now inhabits that sea, and he suggested the obvious explanation that those regions must formerly have been covered by the waters of the sea. And it is indeed the fact that large portions of the vast region comprised between the lower Volga, the Aral-Irtysh water-divide, the Dzungarian Ala-tau, and the outliers of the Tian-shan and Hindu-kush systems are actually covered with Aralo-Caspian deposits, nearly always a yellowish-grey clay, though occasionally they assume the character of a more or less compact sandstone of the same colour. These deposits attain their maximum thickness of 90 ft. east of the Caspian, and have in many parts been excavated and washed away by the rivers (which have frequently changed their beds) or been transported by the winds, which sweep with unmitigated violence across those wide unsheltered expanses. The typical fossils unearthed in these deposits are shells of species now living in both the Caspian and the Aral, though in the shallow parts of both seas only, namely (according to Ivan V. Mushketov [1850–1902]) Cardium edule, Dreissena polymorpha, Neritina liturata, Adacna vitrea, Hydrobia stagnalis, in the Kara-kum desert, and Lithoglyphus caspius, Hydrobia stagnalis, Anodonta ponderosa and the sponge Metchnikovia tuberculata, in the Kizil-kum desert. The exact limits of the ancient Aralo-Caspian sea are not yet settled, except in the north-west, where the Ergeni Hills of Astrakhan constitute an unmistakable barrier. Northwards these marine deposits are known to exist 80 m. away from Lake Aral, though they do not cross the Aral-Irtysh water divide, so that this sea will not probably have been at that time connected with the Arctic, as some have supposed. The eastern limits of these deposits lie about 100 m. from Lake Aral, though Severtsov maintained that they penetrate into the basin of Lake Balkash. Southwards they have been observed without a break for 160 m. from Lake Aral, namely in the Sary-kamysh depression (the surface of which lies below the level of the Caspian) and up