122 ANALYSIS OF MILK. 



of finely-powdered antydrous sodium sulptate. Tte sodium 

 sulphate absorbs the water contained in the milk, and, after 

 well stirring, the residue is quite dry; this is transferred to 

 a flask holding 100 c.c. and 25 c.c. of ether added ; after shaking 

 for five minutes, an aliquot portion is withdrawn by a pipette, 

 over the point of which a piece of cotton-wool is wrapped, and 

 the fat estimated by evaporation and weighing. 



Froidevaux precipitates the casein and fat with a solution 

 containing 35 grammes of calcium phosphate and 6 c.c. acetic 

 acid per litre ; 90 c.c. of this solution are mixed with 10 c.c. of 

 milk, and the fat determined as in the Eitthausen process. 



(2) "Volumetric Methods. 



These, being suitable for use in the rapid testing of milk, will 

 be more conveniently considered in the chapter on " The Testing 

 of Milk." 



(3) Indirect Methods. 



Estimation of Cream. — One of the earliest and simplest 

 methods of estimating the fat in milk is to allow the milk to 

 stand, and to measure the volume of cream thrown up. For 

 this purpose a creamometer or cylindrical vessel, the upper 

 portion of which is divided into spaces, each representing the 

 y^^th part of the total volume up to the highest line, is employed. 

 It is filled to the mark, and allowed to stand at rest for some 

 time — six, eight, twelve, or twenty-four hours — and the volume 

 of cream measured. A good milk should throw up about 10 per 

 cent, of its cream in eight hours. 



The method is of very slight value for a determination of the 

 fat in milk, as comparatively slight variations in the conditions 

 make enormous variations in the volume of cream. Thus, the 

 author has found that a milk — freshly drawn and not cooled — 

 containing 5 '3 per cent, of fat threw up 25 per cent, of cream in 

 six hours, while another milk with the same percentage of fat, 

 which had been raised to the boiling point and cooled, only 

 threw up 2 per cent, of cream in the same time. These, of course, 

 are extreme instances, and it is found in a majority of cases that 

 the percentage of cream thrown up in six to eight hours divided 

 by 3 will give an approximation to the percentage of fat. 



It has been proposed to modify this method by raising the 

 cream by centrifugal force, and apparatus have been made to fit 

 on to the spindle of cream separators ; though more concordant 

 results are thus obtained, these methods have not the accuracy 

 of many of the volumetric methods, by which they have been 

 superseded as practical methods. 



