170 THOMAS COOKSEY. 



the same as that previously described. The precipitated 

 phosphates are dried, and can be removed from filter paper, 

 the latter burnt off in platinum crucible at low temperature 

 and the remainder of precipitate added. The whole is then 

 heated by Bunsen burner to constant weight. It is perhaps 

 advisable not to raise to too high a temperature. The iron 

 can be separately estimated either in the original solution 

 or by dissolving up the phosphates with a few drops of 

 strong hydrochloric acid and making use of the iodine and 

 thio-sulphate method. 



We have now the weight of the total iron and aluminium 

 phosphates, the amount of iron, and the amount of phosphate 

 left in solution. The amount of iron phosphate correspond- 

 ing to the oxide of iron found, is subtracted from the total 

 weight of phosphates, the difference gives the aluminium 

 phosphate. By subtracting the known quantity of phos- 

 phoric acid added, from the sum of the three amounts of 

 phosphoric acid found, that is, that combined with the iron, 

 that combined with the alumina, and the amount left in 

 filtrate, we obtain the original phosphoric acid in solution. 



Examples — To the solution to be tested were added in 

 each case 10 cc. normal phosphoric acid solution equivalent 

 to '2366 gm. Po0 5 . 



I. II. 

 Iron oxide, found by iodine method in the ppt. *0384 *0386 



Phosphoric acid in filtrate '1413 '1406 



Iron and aluminium phosphate ... ... ... *1772 '1795 



Iron phosphate calculated from amount of iron 



found '0724 '0728 



Aluminium phosphate by difference ... ... *1048 '1067 



AleOs -0438 '0446 



I. FOUND. II. 



Fe 2 3 '0381 P 2 5 '0340 Fe 2 3 *0386 P 2 5 *0342 



A1 2 3 -0138 P 2 5 '0609 Al 3 O a '0446 P 2 5 '0620 



In filtrate P 2 5 '1413 In filtrate P 2 5 '1406 



Total ... P 2 5 '2362 Total ... P 2 0, '2368 



