DETERMINATION OF ASH. 29 



feres with the accuracy of the ash determination which 

 usually follows the estimation of total solids. 



By prolonged heating the albuminoids of milk acquire 

 a brown color (Renck), and according to Cazeneuve and 

 Haddon and Renck the lactose becomes caramelized, so 

 that the dry residue is always yellowish brown in color. 

 It has been shown that this is not a source of error. 



The dry matter is very hygroscopic and therefore the 

 dish must be tightly covered while being weighed. 



IV. DETERMINATION OF ASH. 



To determine the ash or mineral substances of milk 

 one may use to advantage the residue of dry matter just 

 obtained. This residue is heated to dull redness in the 

 dish loosely covered. After igniting for one to two 

 hours the ash becomes pure white. The crucible and 

 contents are then cooled in the desiccator and weighed. 

 If phosphoric acid in the ash is to be determined, not 

 less than 20 c.c. of milk should be incinerated. The ash 

 is dissolved in very dilute nitric acid and treated with an 

 excess of ammonium molybdate solution. (At least 40 

 parts of molybdate to one part of phosphoric acid). ^V 

 yellow precipitate of ammonium phosphomolybdate is 

 obtained. The whole is then allowed to stand for twelve 

 hours at a temperature of about 40 degrees C, and is 

 then filtered through a small filter and the filtrate exam- 

 ined. An addition of ammonium molybdate to the fil- 

 trate should cause no further precipitation, even on long 

 standing. The precipitate on the filter is washed with 

 a 10 per cent, solution of ammonium nitrate until the 

 washings cease to react acid and is then dissolved by add- 

 ing a little warm dilute ammonia. Rinse the filter thor- 

 oughly with very dilute ammonia. The filtrate and 

 washings containing the ammonium phosphomolybdate 



