38 METHODS OF QUANTITATIVE ANALYSIS. 



and 10 c. c. of the above mentioned potassium hydrate so- 

 lution. Continue heating until a clear yellow solution is 

 obtained, and then dilute to lOO c. c. 



D. METHODS WHICH GIVE ONLY APPROXIMATE- 

 LY ACCURATE RESULTS. 



A large number of methods and apparatus have been 

 proposed for the rapid estimation of fats, but most of 

 them are accurate only to a moderate degree. To this 

 class belong the "cremometer," the "lactoscope," the 

 "lactocrite," and others. The first of these depends upon 

 the principle of measuring the layer of cream that sepa- 

 rates out; the others are optical methods. The cream 

 measure (cremometer) would give reliable results, if in 

 each case, equal volumes of cream of the same fat con- 

 tent would separate out when samples of milk of uniform 

 fat percentage are allowed to stand in proper vessels un- 

 der similar conditions. But this is not the case, and 

 moreover, the temperature, the kind of vessel, and the 

 condition of the milk, all have an influence upon, the 

 amount of cream obtainable. Thus for example, milk 

 diluted with water, furnishes proportionately more 

 •-cream than the same milk undiluted, because the fat 

 globules in the watered milk can rise to the surface 

 more easily than in the pure milk, which is specifically 

 heavier. 



For similar reasons the optical methods for the more 

 exact determinations should be rejected. The trans- 

 parency of milk bears a certain relation to the number 

 of fat globules contained in it. That is to say, the larger 

 the number of fat globules, the greater will be the opac- 

 ity of the milk column. If now the number of fat glo- 

 bules were proportional to the fat content, this method, 

 based upon the transparency of milk, could not be crit- 

 i.ized. But this is not the case. 



