130 ANALYSIS OF MILK. 



obtained by precipitation is thus raised. In practice, the two 

 errors have a tendency to compensate one another. 



Albumin is estimated by boiling the filtrate from the casein ; 

 it is preferable to use the filtrate from the solution containing 

 salt solution, and this should be brought to a degree of acidity 

 such that 100 c.c. require 2"2 c.c. N alkali for neutralisation; 

 the filtrate is raised just to boiling over a small flame, and digested 

 on the water-bath for fifteen minutes ; the albumin separates 

 in a pulverulent form. It is collected on a tared filter or, pre- 

 ferably, in a Gooch crucible, and dried at 100° C. ; the nitrogen 

 may be estimated by the Kjeldahl method, and multiplied 

 by 6'38. The albumin is precipitated practically in a state 

 of purity, and no correction for the ash need be made, but, owing 

 to the precipitation not being complete, the results are slightly 

 low ; if the casein has not been completely precipitated, a portion 

 may be found with the albumin. 



A small quantity of so-called " lacto-protein " remains in solu- 

 tion after precipitation of the casein and albumin ; this chiefly 

 consists of the unprecipitated portions of casein and albumin. 

 It may be estimated by precipitation as copper salt by Eitt- 

 hausen's method, as described above ; by precipitation by tannin 

 and estimation of the nitrogen in the precipitate by Kjeldahl's 

 method ; or by evaporation of the solution and estimation of 

 the nitrogen. 



Sebelein's method, though more tedious, is preferable to the 

 above ; to 10 grammes of milk, 20 c.c. of a saturated solution of 

 magnesium sulphate are added. Solid magnesium sulphate in 

 the form of powder is then added in small quantities at a time 

 till no more is dissolved. The solution is allowed to stand for 

 twelve hours and filtered ; the precipitate is washed four or five 

 times with a saturated solution of magnesium sulphate, an 

 operation which takes some time. The filter and its contents 

 are dropped into a Kjeldahl digestion flask, 30 c.c. of sulphuric 

 acid added, and the nitrogen estimated, as previously described ; 

 an increased volume of soda solution to neutralise the 30 c.c. 

 of acid used must be employed. The nitrogen multiplied by 

 6"38 will give the weight of casein. 



The magnesium sulphate must be free from sodium sulphate, 

 commercial " Epsom salts " often containing this impurity. If 

 distinct acidity be developed in the milk, this should be neutral- 

 ised previous to the addition of magnesium sulphate. 



The albumin is separated by diluting the filtrate and precipi- 

 tating by the addition of tannin, or phospho-tungstio acid ; the 

 precipitate is collected on a filter, and the nitrogen therein esti- 

 mated by Kjeldahl's method. The albumin may be less exactly 

 estimated by boiling the filtrate after dilution and addition of a 



