THE TESTING OF MILK AND CREAM 



111 



stoppers, cork stoppers, metal caps, or screw tops may be used 

 for this purpose. Bottles with paper caps and jelly glasses 

 with tin lids do not furnish tight seals ; they should not be used 

 for this purpose. 



A separate jar is used for each patron, and each jar must bear 

 the respective patron's number. The jars should be thoroughly 

 clean and, in order to guard against errors, they should be ar- 

 ranged on convenient shelves near the weigh can in numerical 

 order, grouping the jars of patrons of the same route together. 

 Taking composite samples (Fig. 10). 



Correct composite samples may be obtained by the use of a 

 milk thief or a graduated pipette. If the milk thief is used, it 

 is inserted into the weigh can of the entire 

 delivery of one patron. The milk in the 

 tube rises to the level of the milk in the 

 weigh can. The milk thief is then emptied 

 into the sample jar. In case the graduated 

 pipette is used, a certain quantity of milk is 

 taken for every pound of milk delivered by 

 the patron (usually about .1 c.c. for every 

 pound of milk delivered). The milk thief 

 is the handier instrument of the two, but 

 where the amount of milk delivered by 

 different patrons varies considerably, the 

 samples of milk from the larger milk pro- 

 ducers are often too large to be practical. 



Other so-called composite samples are FIQ 1Q _ McKay 

 taken by using the same measure for all milk sampler or thief . 

 milk receipts. In this case a small dipper 

 holding about one ounce is generally used. With this dipper 

 a sample of milk is taken daily from the weigh can of each pa- 

 tron's milk and transferred into the sample jar. This method 

 of composite sampling is not mathematically correct and the 

 results tend to be less reliable, although experimental data 



