Io8 STORES AGRICULTURAL EXPERIMENT STATION. 



this one. When the pail i.s to be u.sed a few layers of clean 

 cheese cloth are placed across the opening of the lower funnel 

 and the loose funnel is pushed in to hold the cheese cloth in 

 position. The whole apparatus is simple in structure and can 

 be easily cleaned. 



The method of the experiment was to compare the amount 

 of dirt in samples from the two pails. In order to have the 

 conditions as nearly uhiform as possible two cows were chosen 

 which gave about equal quantities of milk and required about 

 the same time to milk. These cows stood side by side in the 

 stable, and were milked by the same man each time. On one 

 day cow No. i was milked into the open pail and cow No. 2 

 into the covered pail, and on the next day the order was re- 

 versed, cow No. 2 being milked into the open pail and cow 

 No. I into the covered pail. 



The milk in each pail was then thoroughly stirred and one 

 liter of it was taken out for a sample and tested for the amount 

 of dirt it contained. For this work it was necessary to devise 

 a method whereby the amount of dirt in a given quantity of 

 milk could be determined. So far as known to the writer no 

 work of this kind has been done in this country. vSome Euro- 

 pean investigators have made quantitative determinations of 

 the dirt in market milk, and tests have been made of the dif- 

 ferent methods of clarifying milk, the most satisfactory- results 

 being obtained by the use of centrifugal machines. The latter 

 method was employed in collecting dirt from milk in these ex- 

 periments. 



It is very difficult to remove all the dirt particles from milk 

 without at the same time taking out some of the ingredients of 

 the milk, such as small masses of ca.sein, fat globules, etc. 

 Several methods were tried, the one giving the most satisfac- 

 tory results being as follows: 



The sample of milk was heated to 90° F., and was then run 

 through a separator the bowl of which had also been warmed 

 to 90° by running through it warm filtered water, the amount 

 and temperature of the water necessary for the purpose having 

 been previousl}^ a.scertained. The milk was poured directlj^ 

 from the sampling flask into the receiving cup of the separa- 

 tor, thus avoiding an}- error that might arise from pouring it 

 through the supply tank. The .sampling flask was then washed 



