12 



off the cream adhering to the outside of the pipette. When 

 the top of the column of cream has been lowered to coincide 

 with the mark, remove the drop hanging from the point of the 

 pipette and discharge the cream into a clean dry bottle or 

 conical flask of about '600 c.c. capacity. The same pipette should 

 then be filled with clean water exactly to the mark, and the 

 contents delivered into the cream. Again fill the pipette with 

 clean water and deliver into the cream. This must be repeated 

 till the mixture of cream and water contains less than 9 per 

 cent, of fat. The measuring of the water into the cream will, 

 of course, ensure that no cream remains in the pipette. Gene- 

 rally, it may be stated that for a 25 per cent, cream two pipette- 

 fuls of water will be sufficient ; for a 25 per cent, to 35 per cent, 

 cream three pipettefuls ; for a 35 per cent, to 45 per cent, cream 

 four pipettefuls, and so on. Excess of water above that re- 

 quired to dilute the cream to below 9 per cent, should be 

 avoided. 



Extreme care is now required in taking from the diluted 

 cream the 11 c.c. with which to make the test. The need for 

 care arises from the fact that the fat in a mixture of cream and 

 water rises very rapidly, while a violent mixing gives rise to 

 air bubbles which completely vitiate the results. The best 

 method is to first thoroughly shake the mixture, allow it to 

 stand for a few minutes, and then mix again by a gentle rotary 

 motion of the vessel, and immediately draw off the 11 c.c., 

 which should then be delivered into a test bottle, to which 10 c.c. 

 of sulphuric acid has previously been added, as in ordinary 

 testing. The further procedure is identical with that followed 

 in the case of milk. 



The percentage of fat should be read quickly at 65 C. The 

 bottles are then immediately placed in the centrifuge, heated, 

 and again read at 65 C. If the two readings do not agree the 

 sample should be whirled and read a third time. 



It is absolutely necessary to make two separate dilutions of 

 the sample from each consignment of cream, and two tests of 

 each dilution. Thus, four estimations of each cream are ob- 

 tained, and, provided that the extreme difference between these 

 is not greater than 1 per cent, the average of them may be 

 taken to represent the percentage of fat in the diluted sample. 



To obtain from these data the percentage of fat in the original 

 cream multiply the average of the four tests by a number greater 

 by one than the number of pipettefuls of water added, and 

 correct the figure thus found from the following table : 



[TABLE. 



