42 METHODS OF QUANTITATIVE ANAL,YSIS. 



IV. ESTIMATION OF ALBUMINOID SUBSTANCES. 

 A. DETERMINATION OF TOTAL NITROGEN. (Ac 

 cording to Kjeldahl.) 



Ten c. c. of milk are placed in a round bottom flask c 

 hard refractory glass,, and treated with 15 c. c. of pur 

 cone, sulphuric acid. The acid should be added grad 

 ually and mixed thoroughly by shaking. To this is adde^ 

 one c. c. of metallic mercury or 0.5 gram of mercuri 

 oxide. The flask is supported in an inclined position 01 

 a wire gauze, and gently heated, at first with a sma 

 flame. Gradually the heat is increased until the content 

 of the flask begin to boil. This temperature is then main 



calcium salts precipitate out as insoluble sulphate. The heat c 

 the reaction with the acid is sufficient to melt the fats, and t 

 keep them in a molten condition during the next step of th 

 process. While still hot the flasks are placed in a centrifug: 

 machine and whirled for about five minutes. The speed of th 

 centrifugal should be maintained at 1,000 to 1,200 revolution 

 per minute. After rotating for five minutes, boiling water i 

 added till the column of fats comes entirely within the range ( 

 graduations on the neck of the flask. The flasks are then quickl 

 returned to the centrifugal and revolved again for one or tw 

 minutes. In this way the molten fats are collected in a clear! 

 defined column in the graduated stem of the flask. By subtrac 

 ing the number on the scale at the bottom of the column froi 

 the number at the top, the percentage of butter fats is show 

 directly. The extreme points of the meniscus at the top and : 

 the bottom are to be considered as the terminals of the columi 

 The reading should be taken as quickly as possible and befoi 

 the fats solidify. 



Experience has shown that in this process, the strength of tl 

 sulphuric acid used must be approximately as stated. If the ac: 

 be too strong it will carbonize the lactose and particles of carbc 

 will rise to the surface of the liquid. On the contrary, if the ac 

 be too dilute, the casein will not be entirely dissolved and part 1 

 the unchanged curd will appear on the surface. In either case tl 

 line of demarkation of the fat column will not be sharply d 

 fined. 



For determining the per cent of fats in skimmed milk and 

 cream, special forms of flasks are used for each. In the case • 

 skimmed milk twice the usual quantity of sample is taken, ai 

 the reading obtained is divided by two. 



For a more complete description of this method see Wilej 

 "Principles and Practice of Agricultural Analysis," Vol. II 

 page 499. — Translators. 



