continents were raised from the sea, the lake-basins had been already formed, and came up, therefore, brimful of water. In the northern and eastern part of the continent, where the supply from rain and snowfall exceeds the loss by evaporation, the salt, being continuously carried away through their outlets, has become so diluted as to be an imperceptible quantity. In arid regions, as the Pacific slope and the country about the Caspian, where the evaporation was in excess of the supply, the water-level of the lakes continuously sank until, on account of the diminished extent of surface, the equilibrium of loss and gain was attained. Hence the exceeding saltness of Great Salt Lake, the Dead Sea, etc. For a like reason the water of the Mediterranean contains more salt relatively than that of the ocean. Evaporation exceeding the supplies from the rivers and rainfall, it requires a constant current through the Strait of Gibraltar. The same is true of the Red Sea, causing a like current through the Strait of Bab-el-Mandeb. Other salt or brackish lakes probably owe their saltness to the supplies from the land. Water being the most general of all solvents, the rains gather up the chloride of sodium from the soils and the disintegrating rocks, and where the streams fall into lakes whose only outlet is evaporation, the land itself must be a constant source of saline supply, and their waters must become more and more salt, until their capacity as a solvent has been reached.
The Utah Basin must once have been filled to its brim with ocean-water. The outlet has been evaporation. The lake, receding to its present level, has left many evidences of its former extent.
To the drying up of salt lakes is probably due the presence of rock-salt, often found in great quantities in regions of little rainfall.
I come, lastly, to the trend of the North American lakes. A good map, and especially one on the Mercator projection, will show that lakes are not dotted promiscuously here and there, with no regard to system. They have with each other a trend of direction often as well defined as mountain-ranges, or the coast-lines of continents. As already shown, the great American depression bifurcates at about the fortieth parallel, and nearly at right angles, into northeastern and north-western branches, whose lines of direction lie respectively in approximate parallelism with the far-off Appalachian and Rocky Mountain ranges, and with the still farther-off Atlantic and Pacific shore-lines.
Geologists, and especially physical geographers, have noted the fact that the mountain-ranges, the shore-lines of continents, and the islands with each other, have lines of trend mostly northeastward or northwestward. The lakes of North America have similar trends of direction, and therefore form an integral part of the great system upon which the planet itself is built. This is as should be expected. That the line of greatest depression should have an approximate parallelism with the adjacent greatest upheaval is but a physical necessity.
