3 



of each cheese, in order to have a duplicate in case any accident should befall 

 one of the samples. The cheese borer was cleansed before using to obtain 

 the second specimen; and the test tubes were plugged, packed on ice and for- 

 warded to Kingston, being received about eighteen hours after they were 

 taken. 



At this point the analyses might not be entirely reliable; for while speci- 

 mens usually came in good condition, still on several occasions, the ice in the 

 packing box was completely melted, and the contents of the box were almost 

 at the temperature of the air— which was likely due to the placing of the box 

 in some exposed place by the carriers. The exact effects of such a change 

 in temperature could not be accurately gauged; but when it was considered 

 to be a factor, the results of the analysis were excluded from the tables. 



The samples of cheese taken from the College factory, Guelph, were ob- 

 tained in a similar way, except that it was not necessary to pack them on ice, 

 as the laboratory is only a few minutes' walk from the factory. These sam- 

 ples were promptly taken from the factory to the laboratory and immediately 

 analyzed. 



A source of error in the quantitative bacteriological analysis of cheese is 

 the fact, repeatedly determined in control analyses, that plugs from different 

 parts of the same cheese, of the same age, vary as much as 30 per cent, in 

 their bacterial content. Further, even in the same plug, portions of equal 

 weight sometimes show as high as 20 per cent, of difference in the number of 

 bacteria contained in them. A few examples of this fact may be given. A 

 plug from cheese of July, igo2, age 12 days, gave 144,000,000 per gram. From 

 cheese made in September, 1902, the age being 40 days, one plug gave 27.000,000 

 per gram; and another from a different part of the cheese gave 22,500,000 per 

 gram. From cheese of July, 1902, age 12 days, the upper portion of a plug gave 

 210,000,000 per gram and lower portions of the same plug gave 293,000,000. 



We have also noticed in abnormal cheese, made by adding a culture of a 

 gas-producing germ to the milk, that in the separate particles of curd which 

 unite to make the cheese, the exterior surface of each particle contains a 

 larger number of bacteria than the interior thereof. Thus, in an analysis of 

 cheese made in November, the exterior, or outer surface, of the curd particles 

 gave 456 millions per gram, while the interior thereof gave 51 millions per 

 gram; and again, at a later date, the exterior and interior of the curd particles 

 in the same cheese gave respectively 67 millions and 37 millions per gram. 



These examinations, which are typical of many others which we have made, 

 show that there is not an even distribution of bacteria throughout the sub- 

 stance of a cheese, and it would, therefore, seem necessary to modify some- 

 what our methods of analyses. 



Methods of Anajltsis, Etc. 



Methods followed in the Analysis of Samples. The samples sent from 

 Carp factory to Kingston were all subjected to an examination by the differ- 



