32 REPORT— 1875. 



The gravimetric method of precipitation by nitrate of uranium is employed 

 by a few chemists, and is very well spoken of. 



Molyhdic-Acid Method. 



Of all methods, Sonnenschein's process of precipitation with molybdic acid 

 appears to be I'egarded as the most accurate. All the chemists who refer to 

 it speak of it as extremely accurate, and consider that it is preferable to any 

 other in presence of much iron or aluminium ; but comparatively few use it 

 habitually. The causes of this unpopularity are the time required and the 

 expensive nature of the reagent. 



As a very large excess of molybdic acid is required above that which is 

 actually precipitated as " phospho-molybdate of ammonium," it becomes 

 an important matter to recover the molybdic acid from the solution. Unfor- 

 tunately no very simple process of effecting this appears to have been 

 devised. 



The yeUow precipitate obtained, containing as it does less than four per 

 cent, of anhydrous phosphoric acid, becomes very bulky and unmanageable 

 when the weight of phosphoric acid present exceeds -1 or -2 gramme. This 

 fact necessitates the employment of very small quantities of the phosphate ; 

 and as the yellow precipitate has to be subsequently redissolved and 

 precipitated with magnesia mixture in the ordinary way, the error liable to 

 occur from the use of an unusually small weight of the sample, together 

 with the loss of time and expense incident to the use of the process, seem to 

 have combined to render the method unijopular for every-day work *, while 

 its value is generally admitted when the above considerations are of secondary 

 importance. 



J. Macagno has veiy recently proposed to reduce the yellow precipitate 

 with zinc and acid, and titrate the solution so obtained with standard 

 permanganate. The test experiments show a maximum error of '5 per cent. 

 of the phosphoric acid present. 



Eggertz^s Molyhdic-Acid Metliod. 



Metallurgial chemists are well aware that M. Eggcrtz has proposed to 

 weigh the yellow precipitate of phospho-molybdate of ammonium instead of 

 redissolving it and converting it into ammouio-phosphate of magnesium in 

 the ordinary manner. 



The modified plan has the advantage of speed ; and the fact that the 

 precipitate contains less than four per cent, of phosphoric anhydride would 

 render the results extremely accurate. 



Unfortunately it seems improbable that the precipitate has a constant 

 composition ; and any sensible variation in the proportion of phosphoric acid 

 contained in it would render it worthless as a method of estimation, at least 

 as far as manures arc concerned. 



M. Eggcrtz estimates the anhydrous phosphoric acid contained in the yel- 

 low precipitate, obtained under the conditions prescribed by him, at 3-72 per 

 cent. 



* Mr. A. Sibson writes : — " The molybdlc-acid process is, in my opinion, not suitable 

 for pbospbatic minerals, altbougli invaluable for soils, limestones, &c. containing small 

 proportions only of pbosplioric acid. The large excess of molybdic acid necessarily 

 employed in the former case is itself a source of error with no adequate advantage, 

 inasmuch as the magnesia precipitate has still to be employed ; and it is in the manipula- 

 tion of this precipitate that the differences in analyses chiefly arise." 



