L A K L A K that the rains of several wet seasons would have to accumulate before forcing a passage. The result would be a tide of a period corresponding to the recurrence of series of wet or dry years. Were the lake situated at or near the level of the ocean, its equatorial position would give it such a preponderance of rain over the whole year as to keep its outlet constantly open ; but its actual position, 2700 feet above the sea, produces an alteration in climate, equi valent to an increase of latitude, which would place it in the trade wind region rather than in that of equatorial calms and rains. That such is actually the effect is shown by the range of temperature, which is moderate (59 to 83 Fahr.),and the rainfall(27 to 30 inches), which is almost exactly that of London. The Central African lakes, from their immense size and from their equatorial position, possess a peculiar interest for the physical geographer, and it is to be hoped that before long we shall have sufficient soundings to give a general idea of the size of their basins, and also temperature observations to show the effect of a vertical sun on large bodies of water at a moderate eleva tion, and removed from the disturbing influence of oceanic circulation. As might be expected, in salt lakes which have no overflow, the yearly rise and fall is often considerable. In the Great Salt Lake in Utah, the greatest depth of which is 56 feet, changes of level are accompanied by great changes in water surface, and also in saltness of water. In the rainy season the Dead Sea stands 10 or 12 feet higher than in the dry season. The following table shows the chemical composition of the waters of various salt lakes, that of the sea-water in the Suez Canal being added for comparison : Kokonor Sea. Aral Sea. Caspian Sea. Urumieh Sea. Dead Sea. Van Sea. Suez Canal, Ismailiu. Open. Karabugas. 1-00907 1-11 1-09 1-01106 1-30 1-26217 28-6 1-17500 22-28 22-13 1-01800 1-73 1-03898 5-1 Name of Salt. Grammes Salt in 1000 Grammes Water. 0-6804 0-005-3 0-6598 0-2185 1-3499 2-9799 6-2356 0-1145 0-6603 0-1123 0-0014 0-0021 0-9004 3-0855 S-li 63 0-1339 0-0034 0-6115 0-0081 0-0024 61-9^50 83-2840 9-9560 0-2510 129-3770 0-1930 0-7570 13-5460 192 4 i oO 15-4610 0-5990 0- 600 76- 000 23- 000 95- 000 22-4500 2-3100 0-2400 0-4031 5-3976 0-2595 2-5673 0-5363 8-0500 0-0761 0-0072 0-0069 0-6629 1-8593 3-2231 40-4336 0-6231 0-0265 4-7632 0-0779 0-0027 , magnesia 0-0028 0-9324 1-7241 6-90(18 0-2209 0-0055 potassium magnesium Bromide of magnesium 0-0045 0-0098 Silica Total solid matter 11-14C3 10-8987 12-9773 2849960 222-7730 221-2600 17-2899 51-0264 This table embraces examples of several types of salt lake. In the Kokonor, Aral, and open Caspian seas we have examples of the moderately salt, non-saturated waters. In the Karabugas, a branch gulf of the Caspian, the Urumieh, and the Dead Sea we have examples of saturated waters containing principally chlorides. The Van Sea is an example of the alkaline seas which also occur in Egypt, Hungary, and other countries. Their peculiarity consists in the quantity of carbonate of soda dissolved in their waters, which is collected by the inhabitants for domestic and for commercial purposes. The chemical reader will be struck by the quantity of magnesia salt dissolved in water which contains so much carbonate of soda. The analysis in the table is by Abich, quoted by Schmidt iu his interesting "Etudes Hydrologiques, " published in the Bulletin dc I Academic de St Petcrsbourrj. Another analysis by De Chancourt, quoted by Bischof, omits all mention of sulphate of magnesia, but inserts the carbonate. The limits of this article do not admit of the discussion of the many interesting phenomena connected with salt lakes. With regard, however, to a former connexion of the Caspian with the Black Sea, which has been so often suggested, it seems improbable, both on chemical and on physical grounds, that they were ever con nected as seas, that is, in the same way as the Black Sea is con nected with the Mediterranean ; but, if we consider the topography of the Caucasus district, we see that the lowest summit level of the land between the two seas is in the Manytsch valley, 86 feet above the Black Sea. Were the climate of the Caspian to change only very slightly for the moister, its waters might easily rise the 196 feet which would enable it to overflow towards the Mediter ranean, while a relapse towards dryness would be followed by the retreat of the waters, which would be then confined as they are now to the basin of the sea. It is important, therefore, to bear in mind that no terrestrial dislocations are required to produce enor mous changes in the level of salt lakes ; we require only changes of climate, and these very slight. There can be little doubt that, if the climate of the Black Sea extended across the isthmus to the Caspian, the latter -would now stand 200 feet higher, would be fresh, and would overflow into the Sea of Azoff. For other aspects of the subject see GEOLOGY. (J. Y. B.) LAKE DWELLINGS, as their name implies, are habitations constructed, not on the dry land, but within the margins of lakes or creeks at some distance from the shore. The villages of the Guajiros in the Gulf of Maracaibo are described by Goering as composed of houses with low sloping roofs perched on lofty piles and connected with each other by bridges of planks. Each house consisted of two apartments ; the floor was formed of split stems of trees set close together and covered with mats ; they were reached from the shore by dug-out canoes poled over the shallow waters ; a notched tree trunk served as a ladder and the piles were so firmly driven that no shakiness was perceptible even when the houses were crowded with people. In such a climaue the advantages of dwelling in houses so situated are obvious. The custom is common both in the Gulf of Marac.iibo and in the estuaries of the Orinoco and Amazon; indeed the name of the province of Venezuela was given to it from the prevalence of these pile-dwellings along its shores. A similar system prevails in New Guinea. D Urville describes four such villages in the Bay of Dorei, containing from eight to fifteen blocks or clusters of houses, each block separately built on piles, and consisting of a row of distinct dwellings accommodating a number of families. Cameron describes three villages thus built on piles in Lake Mohrya in Central Africa, the motive here being to prevent surprise by bands of slave-catchers. Similar con structions have been described by travellers, among the Dyaks of Borneo, in Celebes, in the Caroline Islands, on the Gold Coast of Africa, and in other places. Historians have referred to the former existence of the custom in Europe and Asia. Hippocrates, writing in the 5th century B.C., says of the people of the Phasis that their country is hot and marshy and subject to frequent inundations, and that they live in houses of timber and reeds constructed in the midst of the waters, and use boats of a single tree trunk. Herodotus, writing also in the 5th century B.C., describes the people of Lake Prasias as living in houses constructed on platforms supported on piles in the middle of the lake, which are approached from the land by a single narrow
