2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 56 



In 1887 Sir John Mnrray ^ published a careful discussion of the rela- 

 tions between rainfall and run-off for the entire globe, and incidentally, 

 from the average composition of nineteen rivers he estimated that the 

 saline matter annually carried into the ocean amounted to nearly 5,000,- 

 000,000 tons. The rivers taken for this estimate were not named, but 

 must have been in great part European. American data were at that 

 time very meager, and, except for the Nile, there was little material to be 

 had relative to the rivers of Africa and Asia. Eeade and Murray both 

 were handicapped by the defectiveness of their data; and yet their con- 

 clusions were so important that they have since been made the basis of 

 several attempts to compute geological time. These attempts will be 

 considered in due course later. It is now time to revise some of the 

 fundamental figures. 



Within the past year, 1909, two noteworthy reports have been pub- 

 lished by the United States Geological Survey.^ In one of them Dole and 

 Stabler have summarized a great number of observations upon the dis- 

 charge, drainage areas and salinity of many American rivers, and in the 

 other Dole has given detailed analyses of the waters east of the hun- 

 dredth meridian. These analyses are numbered by thousands, and in each 

 case the composition of a river water represents the average composition 

 during an entire year. For example, the water of the Mississippi, taken 

 just above New Orleans,' was collected daily. Each week the seven 

 samples were mixed and analyzed, and this was repeated regularly during 

 the year. The average of the fifty-two weekly composites gave the mean 

 composition of the river; and, combined with the known discharge, the 

 amount of dissolved matter contributed annually by the entire drainage 

 basin of the Mississippi to the Gulf of Mexico. With evidence of this 

 kind the problem of chemical denudation can now be attacked systematic- 

 ally ; and to do so is the purpose of this paper. 



First of all, let us try to determine the average composition of the 

 inorganic matter held in solution by river waters, taking into account as 

 far as possible the entire surface of our globe. This is a subject of funda- 

 mental importance; for upon its details the estimates of geological time, 

 to which reference has already been made, must be based. For the rivers 

 of the United States the material, although not exhaustive, is adequate ; 

 for Europe a fair average can be computed ; but beyond these areas the 

 evidence is still very imperfect. Something, however, can be done even 

 Avith the scanty data which now exist ; by comparing the analyses of Asia- 

 tic or African waters with those of similar areas elsewhere. The com- 



^ Scottish Geographical Magazine, voL 3, p. 65, 1S87. 

 " Water-Supply Papers 234 and 236. 



^ The river was also studied with j^rreat thovnughness at several other points, and so too were its 

 more important tributaries. 



