238 THE PHYSICAL GEOGRAPHY OF THE SEA. 



665. In tliat great inland basin of Asia which holds the Caspian 

 Sea, and embraces an area of one million and a half of geograph- 

 ical square miles, we see the water-surface so exquisitely adjusted 

 that it is just sufficient, and no more, to return to the atmosphere 

 as vapor exactly as much moisture as the atmosphere lends in 

 rain to the rivers of that basin — a beautiful illustration of the fact 

 that the span of the heavens was meted out according to the 

 measure of the waters. 



666. Thus we are entitled to regard (§ 639) the Mediterranean, 

 the Red Sea, and Persian Gulf as relays, distributed along the 

 route of these thirsty winds from the continents of the other hem- 

 isphere, to supply them with vapors, or to restore to them that 

 which they have left behind to feed the sources of the Amazon, 

 the Niger, and the Congo. 



667. The hypothesis that the winds from South Africa and 

 America do take the course through Europe and Asia which I have 

 marked out for them (Plate VII.), is supported by so many coin- 

 cidences, to say the least, that we are entitled to regard it as prob- 

 ably correct, until a train of coincidences at least as striking can 

 be adduced to show that such is not the case. 



668. Returning once more to a consideration of the geological 

 agency of the winds in accounting for the depression of the Dead 

 Sea, we now see the fact palpably brought out before us, that 

 if the Straits of Gibraltar were to be barred up, so that no water 

 could pass through them, we should have a great depression of wa- 

 ter-level in the Mediterranean. Three times as much water (§ 648) 

 is evaporated from that sea as is returned to it through the rivers. 

 A portion of water evaporated from it is probably rained down and 

 returned to it through the rivers ; but, supposing it to be barred 

 up — as the demand upon it for vapor would exceed the supply by 

 rains and rivers, it would commence to dry up ; as it sinks down, 

 the area exposed for evaporation would decrease, and the supplies to 

 the rivers would diminish, until finally there would be estabhshed 

 between the evaporation and precipitation an equilibrium, as in the 

 Dead and Caspian Seas. But, for aught we know, the water-level 

 of the Mediterranean might, before this equilibrium were attained, 

 have to reach a stage far below that of the Dead Sea level. 



