AGRICULTURAL CHEMISTRY. 935 



subject and reports detai Ifil studies of the conditions affecting the accuracy of the 

 magnesium, molybdomagnesium, molybdic, and citrate methods. 



He concludes that cither gravimetric or volumetric methods depending upon a con- 

 stant relation of phosphorus to molybdenum in ammonium phosphomolybdate are 

 not reliable because the composition of the precipitate varies with the conditions of pre- 

 cipitation. The proportion of phosphorus to nitrogen in theammonium phosphomo- 

 lybdate is also n inch too variable to serve as a means of determining phosphoric acid. 

 The solubility of magnesium-ammonium phosphate in ammoniacal water can be deter- 

 mined with reasonable accuracy even \\ hen ammonium chlorid, ammonium citrate, 

 as well as magnesium chlorid are present. The equilibrium between magnesium 

 chlorid, ammonium chlorid, and ammonia at aboul LOO' < '. can be determined with 

 sufficient accuracy. 



In the precipitation of phosphoric acid by means of magnesia mixture in cold solu- 

 tion it is not possible to obtain a precipitate of constant composition either by direct 

 precipitation or by the use of the molybdic or citrate methods. By precipitation in 

 hot solution the conditions of precipitation may be varied considerably without any 

 variation in the proportion of phosphorus to magnesium, and in this way 1 part of 

 phosphoric acid in 1,000 parts of solution may be determined with accuracy. 



The molybdomagnesium method worked oul by the author on this basis gives 

 accurate results even in the presence of large amounts of strong inorganic acids, 

 silicic acid, and citric acid, as well as of calcium, iron, and aluminum compounds. 

 The citrate method which he proposes, however, is not applicable in case of solutions 

 containing a large amount of silicic acid and can be used with safety only when small 

 amounts of calcium, iron, and aluminum compounds are present. The molybdo- 

 magnesium method gives highly accurate results with fertilizers and crude phosphates. 

 The citrate method is recommended only for the determination of water-soluble 

 phosphoric acid in superphosphates. 



The molybdomagnesium method proposed is in brief as follows: The solution of 

 phosphate is heated for 10 minutes to about 50° ('., shaken, and allowed to cool. A 

 slight excess of molybdic solution is added (formulas for the calculation of the 

 amount of molybdic solution to be used are given), the precipitation and collection 

 and washing of the precipitate being done in the usual way. The precipitate is dis- 

 solved in the smallest possible amount of 2.5 per cent ammonia solution, and the 

 solution is heated nearly to the boiling point when neutral magnesia mixture is 

 added drop by drop to precipitate the phosphoric acid. While the solution is cool- 

 ing it is frequently shaken in order to cause the formation of a crystalline precipitate. 

 After standing 4 hours the precipitate is collected on a filter, ignited, and weighed 

 as usual. 



In the citrate method proposed the phosphate solution is mixed with 25 to 30 cc. 

 of neutral ammonium citrate and 25 cc. of 10 per cent ammonia solution, heated to 

 boiling in a covered flask, ami the phosphoric acid precipitated with 30 to 1(1 cc of 

 neutral magnesia mixture. By repeated stirring or shaking the precipitate separates 

 in crystalline form, ami after standing at least 4 hours is collected on a filter, 

 ignited, ami weighed as usual. 



A list of 269 references to literature of this subject is given. 



Citrate solubility of phosphoric acid in fertilizers, J. K. S. Dixon (Jour. 

 Agr. Scl., i (1906), No. ;, />/>. 480-448). — The author reports studies of the solubility 

 of the phosphoric acid of steamed hones, raw bones, fish meals, and guano in 

 alkaline, neutral, and acid ammonium citrate, and 0.1, 1, and 2 per cent citric acid. 

 The results are somewhat inconclusive, but seem to warrantthe following conclusions: 



"(1) The order of the solvent power of the three solutions of each of the two 

 classes remains the same throughout in the citrate solutions the order i- (descend- 

 ing), (a) neutral, (b) acid, (c) alkaline solution, with only two exceptions, viz., the 



31080— No. 10—06 2 



