Xciv PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [May I909, 



By dividing the total number of tons of sodium by the total 

 number of cubic miles, we obtain the number of tons of sodium, 

 namely, 46,200, in one cubic mile of river-water. The number 

 obtained from Murray's estimate is, as we have seen, 24,100 ; 

 and it is upon this basis that Joly's estimate of 90 millions 

 of years depends. By substituting the number just obtained, we 



arrive at 90 millions x -?lil^ =47 millions of vears. 

 46,200 J 



The total rainfall of North America, as estimated by Murray, 

 amounts to 56] x 10 12 cubic feet or 3811 cubic miles. If the 

 average run-off may be taken at 0-222, this gives us in whole 

 numbers 846 cubic miles of river-water. The volume of river- 

 water on which our result is based is 235 cubic miles, or 0*278 of 

 the total amount. This might afford a sufficiently trustworthy 

 average, if the rivers on which it is based were fairly distributed ; 

 but the necessary omission of the more northerly rivers, to some 

 extent impairs its value. 



The total area of the North American continent is about 

 9 x 10 6 square miles, and the area drained by the rivers that we 

 have taken into account is a little more than 1,808,710 square 

 miles, or about one-fifth of the total. Thus rather more than 

 one-fourth of the river-water is derived from one-fifth of the 

 area of the continent. The difference is in the right direction, 

 since the greater part of the remaining area belongs to the more 

 northerly part of the continent, where the rainfall and run-off are 

 both less, and it includes besides extensive desert-regions. 



The average composition of the northern rivers might differ 

 from the average that we have already found. Fortunately, a 

 single analysis has been made of one of these rivers, the Nelson, 

 which drains a fairly large area in Northern Canada. The water 

 of this contains near its mouth 45,850 tons of sodium to the cubic 

 mile — a remarkably close approximation to our average. 



which in several cases is situated some distance above the mouth ; in making 

 the calculation its value was carried to the fifth decimal place : the sodium 

 is given in tons. The analyses from which the amount of sodium was 

 determined are given by F. W. Clarke, in the 'Data of Geochemistry': 

 St. Lawrence p. 61 (G), Androscoggin p. 62 (C), Merrimac p. 62 (D), 

 Croton p. 62 (H), Mohawk p. 62 (G), Potomac p. 63 (L), James p. 63 (M), 

 Cahaba p. [63 (C), Mississippi p. 65 (G), Rio Grande del Norte p. 69 (B), 

 Pecos p. 69 (O), Colorado p. 69 (D), Cedar p. 70 (B), Snake p. 70 (C), Sacra- 

 mento p. 70 (U). The other data have been taken from the Water-Supply 

 Papers of the U.S. Geological Survey. 



