JoLY — Some Experiments on Denudation. 25 



overwhelming amounts o£ sodium and potassium already present, 

 could not be eiiected with sufficient accui'acy. The lime was in those 

 cases estimated where the nature of the mineral rendered its solution 

 probable. A pai'tial analysis of the sea-water used was also carried 

 out under like conditions to those obtaining in the case of the salt- 

 water analyses. 



The procedure in analyses was the usual one. The lime was 

 weighed as oxide by ignition of the oxalate, no attempt being made to 

 separate fui'ther possible impurities, which may in the case of sea- 

 water, as pointed out by Dittmar, amount to as much as 9 per cent, of 

 MgO, jS'aaO, &c. The presence of TiOa in the silica precipitate was 

 not sought for, the weighing after the usual precipitation with HCl 

 and ignition being entered as silica. 



The columns headed i. refer to the basalt dealt with according to 

 the second method of experiment as described, the substance being 

 comparatively coarse-grained, and subjected to an alternate flow of 

 water in opposite directions. In all cases the quantities in the columns 

 headed " Salt " have already been reduced by the amounts of dissolved 

 silica, alumina, and lime detected in the unused sea- water, as given in 

 the last column. 



The mean temperature prevailing during the progress of Experi- 

 ments ii., in., IV., and v., was 7° C. ; and during Experiment i., 12° C. 

 At the conclusion of the experiments the sea- water, in each case, 

 showed a distinctly increased alkaline reaction towards litmus ; the 

 fresh water also showed a very faint alkaline reaction towards litmus. 



Consideration of Results. — The principal issue which led to the 

 foregoing experiments is the broad and somewhat complex one, as to 

 whether the water of the sea is a more active solvent denuding agent 

 than fresh water. It seems to have been left an open question up to 

 the present. Daubree's well-known experiments ^ with chloride of 

 sodium and water acting on orthoclase exposed to violent attrition in 

 a rotating vessel admittedly applies to the activity of the one dissolved 

 substance only. Moreover, the negative result which Daubree appar- 

 ently obtained is not in agreement with the results obtained byBeyer,^ 

 that the felspars decompose rapidly in water containing sodium 

 chloride. But Daubree does not appear to have gone beyond investi- 

 gating the final reaction, whether alkaline or not. Further direct 



' "Geologie Experimentale, " vol. i., p. 237. 



* Quoted by G. P. Merrill in " Eocks, Rock -Weathering, and Soils," p. 178. 



