-11 



c. inokganic constituents. 



1. Methoiis of Chemical Analysis. 



The drift samples froin the three glacial types were subjected to a 

 complete or rock anal\sis. In the case of the individual field samples, 

 determinations of only carbon dioxide and i)hosphoric acid were made. 



The methods of analysis employed were those used in the labora- 

 tory of the United States Geological Survey (19) except in the case of 

 phosphoric acid. In determining this extremely important constituent 

 by the method used in that laboratory such discordant results were 

 obtained that the main investigation was halted until a satisfactory 

 method for the determination of total phosphoric acid could be devel- 

 oped. The author (28) has already published an account of this sub- 

 sidiary investigation. 



The method for the determination of phosphoric acid thus dexel- 

 oped is briefly as follows : One gram of dry soil is weighed into a ])lat- 

 inum dish of appropriate size and ignited in the muffle at dull red heat 

 a sufficient length of time to insure the com]^lete oxidation of the 

 organic matter. After cooling, 10 cc. of distilled water, 10 cc. of nitric 

 acid and 5 cc. of hydrofluoric acid are added, the mixture well stirred, 

 and the contents of the dish evaporated on the steam bath until ai)proxi- 

 mately 5 cc. remain, when an additional 5 cc. of hydrofluoric acid are 

 added, the mixture again well stirred, and the evaporation continued to 

 complete dryness. Evaporation with small quantities of nitric acid is 

 repeated two or three times. After the final evaporation the residue is 

 dried at 110 degrees C. in the air-bath for an hour or two in order, as 

 Washington (32, p. 163) states, "to render insolul)le any silica which 

 might otherwise come down with the phosphorus." When cool the 

 residue is taken up with 3 cc. of concentrated nitric acid and 7 cc. of 

 distilled water, boiled gently for a few minutes, and after cooling 

 somewhat, is filtered and washed : the phosphorus in the filtrate is pre- 

 cipitated with ammonium molybdate. and finally weighed as magnesium 

 pyrophosphate. With soils poor in phosphorus a larger sample of soil 

 should be taken and the amounts of the reagents ])roportionately 

 increased. 



It w'as later ascertained tliat it about 1.^ cc. of boiling hot water be 

 added after treating the residue with 3 cc. of nitric acid and the con- 

 tents of the dish stirred and set aside for 5 or 10 minutes, instead of 

 bringing this to lioiling. the time of filtering could be considerably 

 shortened without lessening the accuracy of the method. 



