296 PHYSIOLOGICAL CHEMISTRY. 



control of market milk, fat is generally now determined, by separating it 

 from the milk in a centrifugal machine and reading off the volume. A 

 definite quantity of milk is measured out, mixed with a little acid to facili- 

 tate the breaking up of the fat globules, placed in a special bottle with 

 graduated neck or stem and rapidly rotated. The liberated fat collects in 

 the stem and is read off. With the Babcock machine in common use the 

 method is rapid and very accurate. 



Fat is very frequently determined by evaporating milk mixed with 

 broken glass or quartz sand to dryness and extracting with a good sol- 

 vent, preferably light petroleum benzine or perfectly anhydrous ether. A 

 better method is to distribute about 5 to 10 grams of milk from a pipette 

 over the surface of a strip of specially prepared absorbent paper. This 

 is coiled up somewhat loosely, placed in an air oven, dried thoroughly and 

 then transferred to a Soxhlet extraction apparatus, where it is treated 

 with the solvent by percolation through two or three hours. The solvent 

 carries the fat down into a small weighed flask. On evaporation of the 

 solvent the dry fat is left and may be so weighed. 



Sugar. To determine the sugar the fat and proteins must be first sepa- 

 rated, which may be conveniently done by the copper process as illustrated 

 above. 25 cc. of milk is diluted with water to 400 cc. and 10 cc. of the 

 Fehling copper solution added. Then from a corresponding sodium 

 hydroxide solution (containing 10.2 gm. to the liter) alkali is added in 

 amount just sufficient to throw down a bulky precipitate containing all 

 the proteins and fats with the copper. This requires about 7 cc. of the 

 alkali. The mixture is diluted to 500 cc. and a portion is filtered off for 

 tests. If the precipitation was properly made a clear filtrate is secured 

 which contains only a trace of copper, and not enough to appreciably affect 

 the accuracy of titration by the Fehling solution as described in an earlier 

 chapter. The proper factor for lactose must be used in the calculation. 



The protein and fat may be precipitated by use of a solution of lead 

 acetate or mercuric nitrate without dilution. On filtering a clear filtrate 

 is obtained which may be tested by the polariscope. 50 cc. of milk should 

 be used, and after precipitation and filtration made up to 100 cc. for the 

 polarization test. The details cannot be given here. 



The Proteins. If the sugar, fat and ash are accurately found the pro- 

 teins may be estimated by difference; that is by subtracting the sum of 

 these from the total solids found. But this plan should not be followed 

 except as a control. A direct determination of casein may be made in 

 this way : 10 cc. of milk is diluted to 50 and mixed with dilute acetic 

 acid to produce complete precipitation. Something less than 1.5 cc. of 10 

 per cent acid will be needed for this. The precipitate is collected on a 

 Gooch funnel, washed with water, hot alcohol and finally enough ether 

 to remove all the fat. What is left is dried and weighed as casein. From 

 the first filtrate plus the wash water the albumin may be precipitated by 

 boiling. The coagulum is collected on a Gooch funnel, washed with water 

 and alcohol and dried as before. 



