EXPERIMENT STATION REPORTS. 147 



2. When slie has added different percentages of lactose and peptone 

 to her agar medium, her average results are inconclusive, and also lack 

 logical sequence. 



3. In the case of temperature the larger percentage is in favor of 

 20°C; in the study of the most suitable acidity, 15° acid represents 

 the most favorable amount in percentage. Hovi^ever, these averages 

 must be subject to more or less fluctuation. 



In general, in statistical evidence of this nature, the highest per- 

 centage does not represent the most desirable method of testing, but 

 rather the lamentable weakness of our present system. The statistical 

 methods of obtaining laboratory results is reprehensible unless all con 

 ditions and agents are the same; vary these conditions and agents, they 

 will lose their force. The averages Avhich have been cited are misleading 

 because of their dissimilarity, hence are at best crudely approximate. 



RECAPITULATION. 



We may assume from our observation that: 



1. Milks vary in their bacterial content both as to 



a. Numbers. 



b. Species. 



2. This variation is dependent upon the 



a. Source of milk. 



b. Method of milking. 



c. Cleanliness in handling. 



d. Temperature maintained. 



e. et cetera. 



3. Milks in the light of tlie above develop their germ content. 



a. With the same rapidity or 



b. In the same relative manner 



and therefore cannot respond alike to even ideal and perfect methods of 

 testing. 



4. Methods of testing are only incidental to the real problem. They 

 should be employed to indicate and assist in control. 



5. While composition of media, means of dilution, the time of plat- 

 ing, temperature for plate development, and other factors are 

 highly important in the execution of these tests, the most essential 

 factor at this time is uniformity of methods that the variable minor 

 discordant elements may to a large extent counteract each other through 

 the accumulation of evidence. 



6. It follows that the purpose of the ''Committee on Standard 

 Methods of Bacterial Milk Analysis" is justified, even .hough many of 

 its detailed decisions are arbitrary and perhaps unwarranted should 

 uniformity of methods be established. 



7. However, much would be sacrificed, if laboratory men forget the 

 real purport of such analysis and use them simply to estimate the milk 

 per se instead of the conditions which are indicated by such analysis. 



In closing, I desire to suggest that perhaps we shall make more rapid 

 progress and secure more enduring results, were we to devote our ener- 



