DETERMINATION OF TOTAL NITROGEN. 43 



tained for an hour and a half. In this way a clear yellowish 

 or colorless solution is obtained which contains all of the 

 nitrogen of the milk in the form of ammonium sulphate. 

 The flask used in this process should be long-necked 

 (about 15 cm.) and should have a capacity of 150 c. c. 

 to 200 c. c. Care should be taken in digesting the sam- 

 ple with sulphuric acid. Sudden heating is liable to cause 

 bumping and spurting and consequent loss of material. 

 Heating on a sand^'bath is not to be recommended, be- 

 ,cause the flask wilj-more easily check or crack and be- 

 cause the temperatijre is difficult to regulate. Instead of 

 mercury, 0.5 gram of anhydrous copper sulphate may be 

 used although, according to Munk, it is necessary in such 

 cases to digest with sulphuric acid for 20 hours, in order 

 to completely convert the nitrogenous matter into am- 

 monium sulphate.* When mercury is used the transfor- 

 mation may be effected in one to one and a half hours. 

 Many analysts recommend heating the substance with 

 sulphuric acid and sulphuric anhydride, or with sulphuric 

 acid and phosphoric anhydride, or with sulphuric acid and 

 potassium bichromate. Numerous experiments have 

 shown, however, that for the determination of nitrogen 

 in milk, pure sulphuric acid is entirely satisfactory. 



When the digestion with acid is complete the con- 

 tents of the flask are cooled, and transferred to a liter 

 distillation flask. The small flask is rinsed several times 

 with distilled water in order to remove every trace of 

 sulphuric acid. To prevent bumping while distilling add 

 a few fragments of pumice stone. Then add 10 c. c. of 

 a 50 per cent solution of potassium sulphide to decompose 

 mercuro-amides (see note below), and finally enough 

 pure sodium hydrate solution to produce a pronounced 

 alkaline reaction. 



*If copper sulphate be used, the addition of potassium sul- 

 phate later is unnecessary. 



