G38 



DATA OP GEOCHEMISTRY 



to repair the obvious deficiencies in the data of river 

 chemistry. They are based principally on the work of 

 L'vovich (1945), and appear to be substantially correct, 

 except for a few spelling mistakes and two more im- 

 portant ones. The area of the Niger drainage basin is 

 given as 216,000 square miles, whereas it is actually 

 about 800,000 square miles, and the estimated total 

 discharge for the continent of North America seems to 

 be about 2,000,000 cfs too high. 



Additional information was obtained from the 

 "Oxford Atlas" (Lewis and others, 1951) and the 

 "Encyclopedia Britannica" (Yust, 1949) as well as 

 from some manuscript notes made from L'vovich's 

 paper. The original was not available while the com- 

 putations were being made. For the United States 

 some information was obtained from "Large rivers of 

 the United States" (U.S. Geol. Survey, 1949a). 



NORTH AMERICA 



For half of North America there are sufficient 

 chemical and discharge measurements to permit a 

 direct computation of the amount of dissolved sub- 

 stance carried by the large rivers. This yields a figure 

 of 92 metric tons per square mile per year. It would 

 be possible to obtain an estimate for the entire conti- 

 nent by taking this as a representative sample, but a 

 more accurate mean can be obtained by weighting these 

 large rivers in proportion to the part of the entire con- 

 tinent that they represent, instead of in proportion to 

 their own drainage areas. The difficulty in making 

 this kind of estimate is that there are some parts of 

 the continent that are climatically very different from 

 any part whose rivers are known, so that a few data 

 from other parts of the world will have to be used. 



The data are presented in table 75, supplemented by 

 estimates of conditions in places where they are lacking. 

 The two biggest gaps are in the Arctic regions and in 

 Mexico and Central America. These have been filled 

 by assuming that various parts of the areas concerned 

 were similar to parts of Alaska and South America. 

 A weighted mean of the information in this table leads 

 to an estimate of 85 metric tons per mile being carried 

 each year by the rivers of the North American conti- 

 nent. When proper allowance is given for the way in 

 which bicarbonate is expressed, this figure is about 8 

 percent above that obtained by Clarke (1924a, b). A 

 slight further correction might be made because this 

 figure is a mean for the amount delivered to the sea 

 by the entire land surface, including closed basins, but 

 it is evident that the agreement between this estimate 

 and the previous one is fairly good. Further informa- 

 tion for arctic and tropical North America will permit 

 a more exact estimate of chemical denudation of 



Table 75. — Discharge and chemical denudation of North America 



North America but the present one is certainly of the 

 correct order of magnitude. 



EUROPE 



The chemical denudation of Europe is not easy to 

 estimate because the discharge of that continent is 

 divided among a multiplicity of small rivers. The 

 principal rivers for which data are available are listed 

 in table 76, but they account for less than a quarter 

 of the total discharge. The Volga basin, of course, 

 contributes nothing to the sea, but there is a substantial 

 part of western Europe, particularly Iceland, Fennos- 

 candia, and the British Isles, that must have a heavier 

 runoff than the rivers listed. This area has been 

 estimated at 500,000 square miles, with a discharge 

 of 700,000 cfs. There is no firm base to use for com- 

 puting the chemical composition of this water, for 

 most of the rivers that have been analyzed are small 

 ones draining very soluble sedimentary rocks in south- 

 ern England. It may be assumed that the composition 

 is represented by the three rivers in Sweden for which 

 data are available, although this will probably lead 



