50 STORRS AGRICULTURAL EXPERIMENT STATION. 
the milk at 50° kept, on an average, about 300 hours before 
it was sufficiently acid to curdle. The difference in the keep- 
ing power of the strained and unstrained samples was not 
sufficiently great to have much significance, although the un- 
strained samples kept in these experiments, on an average, 
about ten hours longer than the strained samples. If the 
figures are compared:a little more closely, the curious fact is 
brought out that the time of curdling of the milk when it is 
kept at 50° ‘appears to be quite unrelated to the number of 
bacteria present at the outset. For instance, from Table 6 it 
will be seen that the sample of Jan. 2 contained nearly 13,000 
bacteria, the largest number in fresh milk in the whole series. 
But at 50° this milk kept 308 hours, somewhat longer than 
the average. The sample of Jan. 16 contained in the fresh 
milk the smallest number of bacteria in the whole series, only 
1,137, but this sample kept the shortest time of all, curdling in 
219 hours. A further study of the tables and a comparison of 
the first column with the last column bring out clearly the 
fact that in milk kept at 50° the number of hours that elapsed 
before the milk was ready to curdle is quite unrelated to the 
number of the bacteria present in the fresh milk. The practi- 
cal conclusion that is to be drawn from this for guidance in the 
handling of milk is that, for the purpose of increasing the keep- 
ing properties of the milk, the question of temperature zs of 
primary importance and has far more significance in determin- 
ing the keeping property of milk than has the extent of the 
original contamination. This result holds for all samples of 
milk where the number of bacteria is moderate, ranging from 
1,000 Or more tO 20,000 or 30,000 per cubic centimeter. Whether 
the same lack of parallelism holds in cases where the milk is 
excessively contaminated or where it is exceptionally clean, is 
studied by experiments to be described in the third section of 
this article. ; 
6. Number of bacteria at the time of curdling.—In the four 
tables it will be seen that about half of the samples curdled in 
the daytime, and consequently the number of bacteria at the 
time of curdling could be determined. ‘The other half it was 
impossible to study, for reasons indicated. Only a few con- 
clusions can be drawn from the figures which are shown in bold 
faced type in the tables. It will be noticed that the number of 



