322 



MILK AND CREAM TESTING 



difficult and unsatisfactory to try to carry on composite samples 

 of cream for the purpose of knowing the pounds of fat delivered, 

 by means of a single test, is because the test of the several lots 

 varies so greatly that unless a sample be taken in exact propor- 

 tion to the quantity of cream delivered, or received, the effect of 

 this method of delivery upon the sample will be out of propor- 

 tion. To illustrate, suppose there be delivered to the factory, 

 four batches of cream as follows : 



= — 7 



b — * 



= —s 



E— z 

 «---= — 



Fia. 111. — In milk testing the fat should be read from a to b, not to c nor to d, but 

 in cream from a to d. 



490 117 127.4 



The false average test, found by dividing 117 by 4, is 29.25 per cent. 



The average test, found by dividing 117 by 4, is 29.25 per 

 cent. There will have been delivered 490 pounds of cream. 

 The quantity of fat delivered may readily be found by multiply- 

 ing the pounds of cream by the test in each and every case, and 

 by adding the quantities of fat. Thus we see in this case 127.4 

 pounds of fat were delivered. If a composite sample had been 

 made of these four lots of cream and the quantity of cream taken 

 for the sample had been the same in each case the test of the 



