THE ASH CONSTITUENTS OF PLANTS. 219 



filtrate from the latter with ammonia and sodium phosphate or ammo- 

 nium phosphate. 



In the second aliquot the sulphuric acid is precipitated with barium 

 chlorid, the precipitate being boiled once with dilute hj^drochloric 

 acid. The third aliquot is treated with recrystallizcd ])arium hydrate 

 to remove the ferric oxid, magnesia, and sulphuric and phosphoric 

 acids, and then with ammonium carbonate to remove the barium and 

 lime. The filtrate is neutralized with hydrochloric acid, evaporated to 

 drjniess, and the ammonia driven off, after which the residue is puri- 

 fied Iw taking up in water, filtering, and evaporating, and the com- 

 bined weight of the potassium and sodium chlorids obtained. The 

 potash in this residue is determined with platiuic chlorid as usual, and 

 the soda found b}^ difference. 



The phosphoric acid is precipitated from the nitric acid solution first 

 with molybdic acid and then with magnesia mixture; and the chlorin 

 is precipitated in another portion with silver nitrate. 



To avoid contamination of the solutions it is advisable to employ 

 only dishes of platinum or of Berlin or Meissen porcelain, and Jena 

 glass beakers. The calculations are made by means of the atomic 

 weights of Clarke or of the Berlin atomic weight commission. The 

 results must of course be corrected for the chlorin in combination, 

 deducting its equivalent of oxygen from the total. 



DEFECTS OF THE DESCRIBED METHOD. 



The above described method may possibly be regarded as open to 

 criticism, as it does not take account of all the elements occurring in 

 ash, and does not make an altogether sharp separation of phosphoric 

 acid, lime, and magnesia. It possesses the advantage, however, of 

 being applicable to analysis on a large scale, which can not be said of 

 the theoretically exact methods, as they require too much time, and 

 except in the hands of very expert analysts are liable to lead to error. 



Among the limitations of the method it will be noticed that no notice 

 is taken of manganese, although this element is frequently found in 

 plant ash. Pichard^ found it in considerable amount, especially in 

 fungi, leaves, and grain. It may be estimated in the filtrate from the 

 iron-phosphate precipitate with bromin, the manganese being precipi- 

 tated as MnOg. If no manganese is separated it is mixed with other 

 precipitates. 



The lime precipitate derived by the above method may contain some 

 phosphoric acid and also some magnesia; and, on the other hand, the 

 lime is not wholly insoluble in the acid solution. These errors may 

 counterbalance each other, but some lime may be thrown down later 

 with the magnesium phosphate."^ In very exact analysis, either the 



iCompt. Rend. Acad. Sci. Paris, 126 (1898), p. 550. 



=* Among others see Hornberger, Ztschr. Analyt. Chem., 18 (1879), p. 361; Richard- 

 son, Ztschr. Analyt. Chem., 23 (1884), p. 409; Amer. Chem. Jour., 3 (1881), p. 422. 



