SUBTERRANEAN WATERS. 467 



the ocean, and some remains, tliougli even liere not in ever mo- 

 tionless repose, to fill deep cavities and underground channels. 

 In every case the aqueous vapors of the air ai*e the ultunate 

 source of suj^plj, and all these hidden stores are again returned 

 to the atmosphere by evaporation. 



The proportion of the water of precipitation taken up by direct 

 evaporation from the surface of the ground seems to have been 

 generally exaggerated, sufficient allowance not being made for 

 moisture carried downwards or in a lateral direction, by infiltra- 

 tion or by crevices in the superior rocky or earthy strata. Ac- 

 cording to Wittwer, Mariotte found that but one-sixth of the 

 precipitation in the basin of the Seine was delivered into that 

 sea by the river, " so that five-sixths remained for evaporation 

 and consumption by the organic world." * Maury estimates the 

 annual amount of precipitation in the valley of the Mississippi 

 at 620 cubic miles, the discharge of that river into the sea at 107 

 cubic miles, and concludes that " this would leave 513 cubic miles 

 of water to be evaporated from this river-basin annually." f In 



few important affluents, that evaporation alone would suffice to exhaust all 

 the water which passes under the bridges of Paris." 



This supposes a much greater amount of evaporation than has been usually- 

 computed, but I believe it is well settled that the Seine conveys to the sea 

 much more water than is discharged into it by all its superficial branches. 

 Babinet states the evaporation from the surface of water at Paris to be twice 

 as great as the precipitation. 



Belgrand supposes that the floods of the Seine at Paris are not produced by 

 the superficial flow of the water of precipitation into its channel, but from the 

 augmented discharge of its remote mountain sources, when swollen by the 

 rains and melted snows which percolate through the permeable strata in its 

 upper course. — Annales det Fonts et Chaitssees, 1851, vol. i. 



* PhysicaliseJie Oeographie, p. 286. It does not appear whether this infer- 

 ence is Mariotte's or Wittwer's. I suppose it is a conclusion of the latter. 



According to VaU^s, the Seine discharges into the sea thirty per cent, of the 

 precipitation in its valley, while the Po delivers into the Adriatic two-thirds 

 nnd perhaps even three-quarters of the total downfall of its basin. The differ- 

 ences between the tributaries of the Mississippi in this respect are remarkable, 

 the Missouri discharging only fifteen per cent., the Yazoo not less than ninety. 

 The explanation of these facts is found in the geographical and geological 

 character of the valleys of these rivers. The Missouri flows with a rapid cur- 

 rent through an irregular country, the Yazoo has a very slow flow through a 

 low, alluvial region which is kept constantly almost saturated by infiltration. 



t Physical Geography of the Sea. Tenth edition. London, 1861, § 274. 

 20 



