206 A. F. GUSTAFSON 



concentrations; the same relationship holds at both concentrations when we 

 compare the nitrate recovered from the fine and very fine grades receiving, 

 respectively, 15 and 17 cc. of nitrate solution. This tends to show that the 

 difference is due to a surface effect. 



Let us consider the very fine sand grade in tables 9 and 10. There was 

 incorporated in this grade 8.5 mgm. of nitrate. The first extraction removed 

 7.187 mgm. of nitrate in the 233.46 cc. of drainage (table 10). Had the entire 

 250 cc. of diluted nitrate solution been a homogeneous mixture, this quantity 

 of solution could have contained only 7.695 mgm. of the original 8,5 mgm. 

 incorporated. 



In the second extraction of this grade, 0.48 mgm. of the 1.313 mgm. actually 

 remaining in the quartz, was removed in the 232.85 cc. of drainage. Had this 

 second solution been homogeneous, 0.515 mgm. of nitrate would have been 

 recovered, which still leaves 0.798^ mgm. of nitrate to be accounted for. 



How shall we account for the apparent failure of 0.805 mgm. of nitrate to 

 go into solution in the first, and 0.798 mgm. in second extraction? To the 

 writer there seems but one answer to this question, that of King (46). When 

 the nitrate solution came in contact with the quartz particles, a definite attrac- 

 tion due to the force of adhesion was set up. The quantity of the more con- 

 centrated solution added was greater than the sand could hold by adsorption, 

 so that when the distilled water was supplied the nitrate solution held me- 

 chanically only, formed a homogeneous mixture with the water when stirred. 

 The 16.54 cc. of solution remaining in the quartz after the first extraction, if 

 of the same composition as the 233.46 cc. removed, would have contained 

 0.508 mgm. of nitrate. This is 0.028 mgm. in excess of the 0.48 mgm. actually 

 removed in the second extraction, or, 0.007 mgm. less than would have been 

 contained in the 250 cc. of solution had it been of the same composition as 

 that draining from the sand. 



Since the dilution is so much greater in the second extraction, it may be 

 expected that the outer part of the film, less strongly adhering to the particles 

 would tend to form a homogeneous mixture with the water, this together with 

 the nitrate solution held mechanically only, and that coming by diffusion from 

 the strongly held film, furnishes the whole of the nitrate removed in the second 

 extraction. King's data, table A (p. 177), shows that after six extractions 

 the nitrate recovered is that coming from the strongly adhering film by 

 diffusion. 



Five months after the two extractions were made, the air-dry quartz samples 

 which had stood in the laboratory in beakers covered with paper were them- 

 selves treated with phenoldisulfonic acid and nitrates determined in the usual 

 way. The nitrate so determined is shown as milligrams and per cent of the 

 total in the latter part of table 10. In no case did the samples which had been 

 heated in the previous extractions show the presence of nitrates, while all 



•8.5 mgm. total; actually removed by first extraction, 7.187 mgm., or 1.313 mgm. 

 remained; 1.313 — 0.515 = 0.798 mgm. 



