207 



quart, put into the jar; at the end of a week or ten 

 days the milk in the jar is tested, and the figures thus 

 obtained, together with the record showing the number 

 of pounds of milk brought, furnish the data for making 

 a calculation of the amount of butter fat supplied the 

 creamery by each patron. 



One of the first plans proposed was to add some chem- 

 ical to keep the composite sample of milk sweet, and as 

 this milk was to be used for testing only, corrosive 

 sublimate was proposed. A very small quantity of this 

 deadly poison will prevent lactic fermentation from 

 developing in milk, and as it can be added in the form of 

 a powder mixed with some aniline dye, this increase in 

 the volume of the milk is not sufficient to be taken into 

 consideration when the milk is tested. 



The color given to the mixture by the aniline is sup- 

 posed to act as a guard against tasting or drinking the 

 poisoned milk. 



Flo wever satisfactory this method may be, the use of 

 the poison makes it objectionable, especially as the 

 persons who wish to test composite samples of milk are 

 generally unaccustomed to handling poisons. A great 

 many other chemicals have been tried as substitutes for 

 the corrosive sublimate, but they have proved more or 

 less impracticable. For example, I found that by 

 passing a current of sulphuretted lwdrogen gas through 

 milk it kept in excellent condition for testing and there 

 was undoubtedly no danger of anyone drinking the 

 milk on account of its foul odor. 



After some investigation I have discovered and 

 adopted the following method : Collect the composite 

 samples as before suggested, and make no attempt to 

 keep the milk sweet, but allow it to sour as much as it 

 will. The souring of milk for a week does not percep- 



