302 EVOLUTIONAL GEOLOGY. 



rivers, although only approximate, it is no doubt sufficiently exact; 

 at all events such alterations as it is likely to undergo will not 

 greatly affect the hnal result. When, however, we pass to the last 

 quantity to be determined, the chemical composition of average river 

 water, we find that only a very rough estimate is possible, and this is 

 the more unfortunate because changes in this may very materially 

 affect our conclusions. The total quantity of river water discharged 

 into the sea is, as we have stated, 6,524 cubic miles. The average 

 composition of this water is deduced from analyses of 19 great rivers, 

 which altogether discharge only 488 cubic miles, or 7.25 per cent 

 of the whole. The danger in using this estimate is twofold; in the 

 first place, 7.25 is too small a fraction from which to argue to the 

 remaining 92.75 per cent, and, next, the rivers which furnish it are 

 selected rivers, i. e., they are all of large size. The efi'ect of this is 

 that the drainage of the volcanic regions of the earth is not sufficiently 

 represented, and it is precisely this drainage which is richest in sodium 

 salts. The hivas and ashes of active volcanoes rapidly disintegrate un- 

 der the energetic action of various acid gases, and among volcanic 

 exhalations sodium chloride has been es])ecially noticed as abundant. 

 Consequently, we find that while the proportion of sodium in Professor 

 Joly's average river water is oidy 5.73 per million, in the rivers of the 

 volcanic island of Hawaii it rises to 24.5 per million (Walter Maxwell, 

 Lavas and Soils of the Hawaiian Islands, p. 170). No doubt the 

 area occupied by volcanoes is triHing compared with the remaining 

 land surface. On the other hand, the majority of volcanoes are situa- 

 ted in regions of copiovs rainfall, of which the}'^ receive a full share, 

 owing to their mountainous form. Much of the fallen rain percolates 

 through the porous material of the cone, and, richh^ charged with alka- 

 lies, finds its way by underground passages toward the sea, into which 

 it sometimes discharges by submarine springs. 



Again, several considerations lead to the belief that the suppl}^ of 

 sodium to the ocean has proceeded, not at a uniform, but at a grad- 

 ually diminishing rate. The rate of increase of temperature with 

 descent into the crust has continuously diminished with the flow of time, 

 and this must have had its influence on the temperature of springs, 

 which furnish an important contribution to riv^er water. The signifi- 

 cance of this consideration may be judged from the composition of the 

 water of gej\sers. Thus Geyser, in Iceland, contains 884 parts of 

 sodium per million, or nearly one hundred and sixt}^ times as much as 

 Sir John ]\Iurray estimates is present in average river water. A mean 

 of the analyses of six gej'sers in different parts of the world gives 400 

 parts of sodium per million, existing partly as chloride, but also as 

 sulphate and carbonate. 



It should not be overlooked that the present is a calm and quiet 

 epoch in the earth's history, following after a time of fiery activity. 



