632 APPENDIX 



carbonate are fused as directed in Fresenius' " Quantitative Analysis," 

 page 426, and by Hillebrand in Bulletin 305 of the United States Geo- 

 logical Survey, where an illustration of the apparatus is given. The 

 fused mass is transferred to a porcelain dish, slacked with hot water, 

 finely ground with an agate pestle and transferred to a filter. After 

 washing with about 600 cc. hot water, the filtrate and washings are run to 

 dryness in a Jena beaker, taken up with hot water and again filtered, 

 acidified with hydrochloric acid, concentrated to about 10 cc., and i^- cc. 

 of a platinum chlorid solution (10 cc. containing i g. platinum) added. 

 This is then evaporated to a sirupy consistency, taken up and washed 

 about fifteen times with 80 per cent alcohol, three times with ammonium 

 chlorid solution, and again fifteen times with alcohol. The precipitate 

 is then washed through the filter with hot water into a platinum dish^ 

 evaporated on the steam bath to dryness and heated in an air oven at 

 110 C. for an hour, cooled in a desiccator, and weighed. Duplicate 

 samples should not differ more than 1.5 mg. in the final weight. 



A correction must be made for the amount of potassium in the reagents 

 which is found by making a blank determination, using no soil. 



(Ammonium chlorid solution is made by dissolving 200 g. NH 4 C1 in 

 1000 cc. water and saturating with K 2 PtCl 6 .) 



Calcium. Five grams of soil (or less if high in calcium) are decom- 

 posed by heating with 10 g. of sodium peroxid in an iron crucible, 

 taken up with water and hydrochloric acid and made up to 500 cc., as 

 for phosphorus. After being allowed to settle over night, 200 cc. of the 

 supernatant solution are heated to boiling and precipitated from the 

 hot solution with ammonia. The precipitate is filtered out on a 15 cm. 

 filter and washed with hot water until but" a slight test for chlorids is 

 given with silver nitrate. The filtrate is again evaporated to dryness 

 and heated (to dehydrate any remaining silica), taken up with water 

 and hydrochloric acid, brought to a boil, and ammonia added to pre- 

 cipitate any remaining aluminum. The precipitate is filtered out on a 

 small filter and washed with hot water. It should not be washed more 

 than necessary to remove the chlorids, as the wash water carries alumi- 

 num through into the filtrate. On heating this filtrate and allowing it 

 to stand over night, more aluminum may be found to precipitate out. 

 All of the aluminum must be removed by repeated precipitations. The 

 solution is then made slightly alkaline with ammonia, brought to a boil, 

 and to it is added slowly, while it is being stirred, enough concentrated 

 ammonium oxalate solution to precipitate the calcium and to change 

 the magnesium to the oxalate. After boiling until the precipitate has a 

 granular appearance, it is allowed to stand three hours or longer, de- 



