DAIRY PRODUCTS. 55 



O, that is, when the substance contains no pure butter, the value of n 

 may bo taken at .30. We bave, therefore, 



O=a (0.30 -&) from which l=3(}. 



When B is equal to 100, that is, when the butter is pure, as has already 

 been said, the most probable value of w, according to the thirteen anal- 

 yses given, is 13.07, or in round numbers 1445, then we have the 

 equation 



100=a (14.00i 0.45-0.30) 



and from this the value of 



a=7.30i.24. 



The above equations may therefore be condensed into B = (7.30 

 0.24) (n 30), that is, in order to find the probable butter content of a 

 fat mixture subtract from the number of the cubic centimeters of deci- 

 nonnal soda lye used for titration .30 and multiply the remainder by 7.30. 

 The probable error which will be met with by this estimation amounts 

 to i 0.24 (n 0.30). 



Medicus and Scherer 1 examined the method of Eeichert and found it 

 to be quite exact. For pure butter they found the quantity of decinormal 

 soda lye required should be 13cc. ; a mixture of equal parts of butter fat 

 and tallow required 7cc. 



Two parts of butter fat and one of tallow required O.lcc.; three parts 

 of butter fat and one of tallow required lO.lcc. 



The authors call attention to the fact that melted butter fat slowly 

 cooled may separate into portions requiring different quantities of the 

 deciuormal soda for the saturation of the distilled acid which they 

 afford. Two and one-half to 3 pounds of pure butter fat were nsed. 

 This was melted and allowed to cool with continued stirring in order 

 to secure a perfectly homogeneous mass. 2J grams of this mixture, by 

 Keichert's method, required 14cc. of decinormal soda. The fat was now 

 again melted, poured into a large beaker glass, and uncovered allowed 

 to cool without stirring. The solidification took place slowly. After 

 solidification 2J grams from the upper layer required 13.3cc. of soda. 



Allen 2 also highly recommends Keichert's method. He uses it as fol- 

 lows : Weigh out 25 grams ot the clarified butter fat and saponify in a 

 closed ilask (a closed flask has been used in the work of the Chemical 



N 

 Division with butter since 1883) with 25cc. of approximately KOH. 



Transfer the product to a porcelain basin and evaporate the alcohol at a 

 steam heat. Dissolve the residue in water, add some fragments of 

 pomace coiled round with platinum wire, and distil gently until 50cc. 



N 

 have passed over. Titrate the distillate with caustic alkali using 



phenol-phthalein as an indicator. 



'Zeit. Anal. Chem., 1880, pp. 151) ctseq. 2 Analyst, 1885, p. 103, et seq. 



