126 ANALYSIS OF MILK. 



using litmus or cochineal as indicator. An amount of acid 



■equivalent to the alkali used is subtracted from the amount 



of acid originally added ; the difference represents the acid 



neutralised by the ammonia distilled. From this figure should 



be subtracted the figure obtained in a blank experiment — i.e., 



an experiment performed without the addition of milk, but in 



N 

 other respects exactly the same. Each c.c. of — acid neutralised 



by the ammonia produced is equal to 0"0014 gramme of nitrogen. 

 The percentage of nitrogen roultiplied by 6 "38 will give the 

 percentage of total proteins. 



Another form of apparatus gives equally good results if care be 

 used ; the caustic soda and sulphide solution are poured carefully 

 ■down the side of the distilling flask, so that the alkaline solutions 

 do not mis with the acid ; the bulb tube rapidly inserted, and the 

 contents of the flask mixed. No condenser is used, but the flask 

 containing the standard acid may be placed in cold water (Fig. 11). 



This method may be modified in many ways. The potassium 

 bisulphate may be omitted, but the digestion then takes longer. 

 ■Copper oxide or sulphate may be substituted for the mercury ; 

 if this be done, the potassium sulphide solution may be replaced 

 by an equal bulk of Soxhlet's alkaline tartrate solution (see 

 Estimation of Millc-Sugar). 



The tube with two bulbs can be replaced by any other form of 

 apparatus having for its object the prevention of splashing ; for 

 the copper flask a Jena or other glass flask, or even a tin bottle, 

 may be used. If a tin flask is employed, it must be remembered 

 that " rust " contains ammonia, and the tin must be boiled with 

 strong soda solution before use ; for this reason a tin flask cannot 

 be recommended. 



For the estimation of nitrogen the methods of Varrentrap and 

 Will and of Dumas may be employed, but they are not so gene- 

 rally convenient. In large laboratories where many deter- 

 minations are made the method of Dumas is perhaps more 

 convenient than that of Kjeldahl ; these methods will be found 

 described in manuals devoted to analytical chemistry. The 

 method of Kjeldahl is, however, the most generally employed in 

 milk analysis. 



Ritthausen's method is performed as follows : — Ten grammes 

 ■of milk are diluted to about 100 c.c. and 5 c.c. of Soxhlet's copper 

 sulphate solution (see Estimation of Milk Sugar) added ; a solution 

 of caustic soda (25 grammes per litre) is added, drop by drop, till 

 the solution is nearly neutral. The precipitate settles rapidly ; 

 an excess of alkali must be avoided, as it prevents precipitation 

 of the proteins. The precipitate is allowed to settle, and the 

 supernatant liquid poured off through a tared filter, or Gooch 



