BAKING POWDERS. 595 



cipitate of iron and aluminium phosphates is filtered oft', washed, and 

 redissolved in the smallest quantity of hydrochloric acid. The resultant 

 solution is poured into an excess of an aqueous solution of pure caustic 

 soda contained in a platinum or nickel vessel. After heating for some 

 t ime, the liquid is considerably diluted and filtered. The filtrate is acid- 

 ulated with hydrochloric acid, ammonium acetate and a few drops of 

 sodium phosphate added, and then a slight excess of ammonia. The 

 liquid is kept hot till all smell of ammonia is lost, when it is filtered, 

 a iid the precipitated aluminium phosphate washed, ignited, and weighed. 

 Its weight multiplied by 3.713 gives the ammonia alum (hydrated), or 

 by 3.873 the potash alum in the .5 grams of sample taken. 



In the phosphate and alum powders the above method gave a fairly 

 good separation of the alum, but the following separation by means of 

 molybdenum was found to be more exact, and at the same time much 

 more convenient of application. It was adapted to the powders by Mr. 

 K. P. McElroy. 



Weigh out 5 grams into a platinum dish, char, treat with strong nitric 

 acid, and filter into a 500 cubic centimeter flask. After washing the 

 residue slightly, transfer filter and all back to the platinum dish and 

 burn to whiteness. To the ash add mixed carbonates and fuse. Take 

 up with nitric acid, evaporate to dry ness, acidify again with nitric acid, 

 and wash all into the 500 cubic centimeter flask. Nearly neutralize the 

 contents of the flask with ammonia, and add molybdate of ammonium 

 sufficient to precipitate all the phosphoric acid present. Allow to stand 

 some time, make up to the mark, shake thoroughly, and filter off 100 cu- 

 bic centimeters through a dry filter. This is exactly neutralized with 

 ammonia, keeping the solution as cool as possible to avoid the deposi- 

 tion of molybdic acid. Filter and wash the precipitate, redissolve in 

 dilute nitric acid with the addition of a little ammonium nitrate, and 

 precipitate as before. Filter through a paper filter, burn, ignite, and 

 weigh as A1 2 O 3 . The alumina and phosphoric acid may be determined 

 in the same sample by the above method, modifying it as follows: When 

 the solution, ash, etc., have all been brought into the graduated flask, 

 make up to the mark without adding molybdate, filter and take 100 

 cubic centimeters, nearly neutralize with ammonia, add ammonium ni- 

 trate and molybdate of ammonium, digest and filter. The filtrate con- 

 tains the aluminium and may be precipitated with ammonia as above, 

 while the phosphoric acid is all contained in the precipitate, which may 

 be redissolved in ammonia and precipitated with magnesia mixture. 



ESTIMATION OF CALCIUM. 



This determination was made as follows : Weigh out 5 grams of the 

 sample, transfer to a 500 cubic centimeter flask, add 40 or 50 cubic cen- 

 timeters of water, and then 20 or 30 cubic centimeters of strong hydro- 

 chloric acid. Make up to the mark with water, shake thoroughly, and 

 set aside to allow starch to settle. Filter through a dry filter, and take 



