PHYSICAL AND CHEMICAL EXAMINATION OF MILK 171 



"Section 3: Optional methods: The mercury and cork, alcohol 

 and buret, and alcohol and brass plunger methods may be em- 

 ployed for rapid testing of Babcock bottles, but the accuracy 

 of all questionable bottles shall be determined by the official 

 method. 



"Section 4: The official method for testing pipets and acid 

 measures shall be calibration by measuring in a buret the quan- 

 tity of water (at 20 C.) delivered. 



"Section 5: The limits of error (a) for Babcock bottles shall be 

 the smallest graduation on the scale, but in no case shall it exceed 

 0.5 per cent., or for skimmed milk bottles 0.01 per cent. 



"(6) For full quantity pipets it shall not exceed 0.1 c.c., and for 

 fractional pipets 0.05 c.c. 



"(c) For acid measure it shall not exceed 0.2 c.c." 



Milk-testing bottles can be used for cream testing by dividing 

 the charge between two or more bottles. Enough water to bring 

 the volume in each bottle to about 17.5 c.c. should be added. 

 The readings of the bottles are then added to obtain the percentage 

 of fat. 



Sampling Cream for the Babcock Test. As the sample should 

 be representative of the true composition of the cream to be tested, 

 it is important to mix the cream thoroughly before sampling. 

 This is best accomplished by pouring the cream from one vessel 

 to another. Ordinary milk cans are not suitable for this purpose, 

 as cream is liable to be spilled on account of the small openings 

 in such cans. Specially constructed cans with wide apertures 

 should be used for this purpose. 



After the cream has been mixed a sampler can be used to take a 

 suitable amount, but care should be taken to prevent mixing the 

 cream adhering to the outside of the sampler with the sample. 

 If a sampling tube is used it should be lowered slowly into the can 

 so that it fills as rapidly as it goes down. When several samples 

 are taken from different cans the sampler must be thoroughly 

 cleaned for each sample. 



Frozen cream must be completely thawed. Frozen or partly 

 churned cream should be rubbed through a sieve to break up the 

 clumps of fat. Rich milk, as Jersey milk, for example, may be 

 partly churned during transportation if the cans are not filled to 

 the top. The small lumps of fat are difficult to distribute evenly 

 through such cream. 



It is probably safer to test the cream each time it is sampled 

 rather than collect composite samples. However, when com- 

 posite sampling is practised the cream must be preserved in tightly 

 closed bottles to prevent evaporation. Unless this is provided for, 

 a tough dry film forms on the surface and thorough mixing can be 



