ANALYSIS OF ASH. 107 



cipitate, after weighing, is dissolved in dilute hydrochloric acid, 

 keeping the bulk small ; ammonia is added to alkaline reaction, 

 and the small precipitate of calcium phosphate collected, ignited, 

 and weighed. Its weight is subtracted from the previous weight, 

 and the difference gives the weight of the calcium carbonate, 

 which, multiplied by 04, gives the calcium, or by 0-56 the lime, 

 contained in it ; the weight of the calcium phosphate multiplied 

 by 0-3871 gives the calcium, or by 0-5419 the lime, contained in 

 it. The total calcium or lime is the sum of the two. 



The filtrate is made strongly ammoniacal by the addition of 

 0-880 ammonia and allowed to stand twenty-four hours. The 

 precipitated magnesium-ammonium phosphate is filtered off, 

 washed with dilute ammonia, ignited, and the magnesium pyro- 

 phosphate weighed. Its weight multiplied by 0-21622 will give 

 the magnesium, and by 0-36036 the magnesia contained in it. 



To the filtrate from this, magnesia mixture is added. The 

 precipitate of magnesium-ammonium phosphate is filtered off 

 after twenty-four hours and treated as above. 



From the total weight of the two quantities of magnesium 

 pyro-phosphate, the phosphoric anhydride is calculated by 

 multiplying by 0-63964 ; to this is added the phosphoric anhy- 

 dride in the calcium phosphate calculated by multiplying the 

 weight by 0-4581. 



The above method has proved satisfactory in the author's 

 hands, though it takes no account of the traces of iron present, 

 which is precipitated with the calcium phosphate, or the mag- 

 nesium-ammonium phosphate. If desired, this may be estimated 

 by dissolving up the precipitate of calcium phosphate and the 

 first magnesium-ammonium phosphate precipitate in dilute hydro- 

 chloric acid, and determining the iron colorimetrically as 

 thio-cyanate, or by dissolving in strong hydrochloric acid and 

 comparing the colour with that of a standard iron solution in 

 strong hydrochloric acid. 



To estimate alkalies, another portion of milk is ignited, as 

 before, and the total ash dissolved in dilute hydrochloric acid 

 and boiled ; a few drops of barium chloride solution are added 

 containing not more than 0-1 gramme of barium to 100 grammes 

 of milk, and the boiling continued for some minutes. After 

 some hours the precipitate of barium sulphate is filtered off, 

 ignited, and weighed ; its weight multiplied by 0-34335 will give 

 the sulphuric anhydride in the milk. If an excess of barium 

 chloride has been added, a little phosphoric acid, or ammonium 

 phosphate, may now be dropped into the filtrate, though it is 

 not necessary if the quantity of barium chloride given above has 

 been employed. A quantity of ferric chloride solution sufficient 

 to colour the solution brown is added, and the filtrate made 



