METHODS OF TESTING BUTTER FOR WATER 121 



tial pieces (i) a flask (A, Fig.37b), (2) a receiving- 

 bulb with graduated tube (C, Fig. 37b), and (3) a 

 condensing-jacket (E, Fig. 37b). The receiving-bulb 

 is connected with flask A by means of a rubber stop- 

 per (B). G is a ground glass stopper fitting tightly 

 into the tube above the graduated scale. When more 

 than one apparatus is usecj, each stopper and tube 

 should be properly marked so that each stopper shall 

 always be used with its corresponding tube. Each 

 mark of the graduated-scale represents 0.2 per cent, 

 of water when a lo-gram sample of butter is used. 

 The glass condensing-jacket (E, Fig. 37b) is con- 

 nected to the graduated tube by a rubber stopper 

 (D, Fig. 37b). 



The operation of the test is carried out as follows : 

 (i) Weigh a lo-gram sample by placing on each scale- 

 pan one sheet of parchment paper and balancing; 

 the sample is placed in center of paper on one pan, a 

 lo-gram weight being in the other. (2) The paper 

 with sample is folded in such shape as to slip without 

 loss into flask A (which must always be clean and 

 dry before use). Some prefer to weigh the sample 

 of butter directly into the flask. (3) Then pour 

 6 cc. of amyl reagent into flask A, (4) connect the ap- 

 paratus (which must be clean and dry always before 

 use) in the manner shown in Fig. 37b, fill the condens- 

 ing-jacket with cool water to within i inch of top 

 (F, Fig. 37b) and remove the glass stopper (to pre- 

 vent danger of bursting of tube by pressure of un- 

 condensed water- vapor). (5) Then heat the bottom 

 A over the flame of the burner. The butter 

 lelts, mixes with amyl reagent, and later passes 



