10 EFFECT OF WATER ON BOCK POWDERS. 



wet grinding, while at the same time the binding power is increased. 

 In the case of such rocks as do not contain alkaline or alkaline earth 

 bases, this test will of course give negative results, but as these bases 

 are usually present in the igneous rocks the samples which do not 

 immediately respond to the phenolphthalein test are the exception 

 rather than the rule. From the figures published by F. W. Clarke 

 for the average composition of all crystalline rocks, computed from a 

 great number of analyses, it is found that lime and magnesia are 

 present to the extent .of about 9 per cent, while the alkaline bases, 

 soda and potash, average about 6.5 per cent. That is to sa} r , in 1 ton 

 of average crystalline rock we might expect to find over 120 pounds 

 of the soluble alkalis alone. 



From the results of these qualitative tests the desire arose to follow 

 quantitatively, if possible, the extent of the reactions carried on by 

 water during the \vet grinding process. To this end a pure orthoclase 

 feldspar of known composition was selected and samples were ground 

 both wet and dry. The d^ ground orthoclase yielded the phenol- 

 phthalein reaction only ver}^ faintly; the wet ground sample on the 

 other hand showed a distinct reaction. Equal weights of the two 

 thoroughly air-dried powders were mixed with equal quantities of 

 recently boiled and cooled distilled water thoroughly shaken for the 

 same length of time and filtered through a Pasteur-Chamberland filter. 

 The intention was to titrate the two filtrates with N/100 acid; there- 

 fore especial precautions were taken to insure the neutrality of the 

 filter tube by pumping quantities of boiled water through and titrat- 

 ing blanks. When the two filtrates were ready for titration, however, 

 it was somewhat disconcerting to find that in neither case were they 

 alkaline to phenolphthalein in the cold, although before filtration they 

 had both appeared distinctly so. On prolonged boiling the pink color 

 appeared and required in each case about 2.5 cc of N 100 acid to dis- 

 charge the color. Unfortunately, the impossibility of obtaining a defi- 

 nite end point in the unfiltered mixtures prevented any comparison of 

 the relative amount of free alkali present before and after filtration. 



The experiment, as described above, was tried again and again with 

 the same sample and with various filters. A great many rock pow- 

 ders were then tried, and as a result of all these experiments the fol- 

 lowing fact appears: TJie majority of rock powders when acted on by 

 distilled water, whether the water has been freed from carbonic acid or 

 not., give immediate alkaline reactions to phenolphtlwilein in the cold; 

 whereas, if the water is freed from the rock powder by passage through 

 any sort of efficient filter the alkaline reaction in the cold is either les- 

 sened or entirely inhibited in the filtrate. 



The first thought which suggests itself in the effort to find an 

 explanation of this is that carbonic acid of the air is alone responsible 



U. S. Geol. Survey Bui. No. 228. 



