Composite Samples of Milk. 



165 



are provided as there are patrons; there is no need of using any 

 preservatives for milk in this case. Fig. 55 shows a convenient 

 rack for holding the test bottles used in composite sampling with 

 a * ' one-third sample pipette. ' ' 



Accurate results can be obtained by this method of sampling, 

 if care is taken in measuring out the milk, and if it is not frozen 

 or contains lumps of cream. 

 It is doubtful if the method 

 has any advantage over the 

 usual method of composite 

 sampling. If milk is deliv- 

 ered daily and each lot is 

 sampled with the one-third 

 pipette, twice or three times 

 the number of tests are re- 

 quired as when composite 

 samples are taken in jars 

 and tested once every week 

 or ten days. This method 

 furthermore takes a little 

 more time in the daily sam- 

 pling than the other, as the 

 quantity of milk must be 

 measured out accurately each 

 time. If a test bottle is accidently broken or some milk spilled, 

 the opportunity of ascertaining the fat content of the milk de- 

 livered during the three days is lost; if a similar accident should 

 occur in testing composite samples collected in jars, another test 

 can readily be made. 



188. Accuracy of the described methods of sam- 

 pling. An experiment made at the Wisconsin Dairy 

 School may here be cited, showing that concordant re- 

 sults willl be obtained by the use of the drip sampling 

 method and the Scovell tube. Two composite samples 

 were taken from fifty different lots of milk, amounting 

 to about 6,000 Ibs. in the aggregate. One sample was 

 taken of the drip from a hole in the conductor spout 



FIG. 55. Test-bottle rack for use in 

 crenmeries and cheese factories. 



