1318 Wisconsin Academy of Sciences, Arts, and Letters. 
The average was 27 and the percentage of variations 152% One 
c.c. of the same milk at the same time was put with 99 c.c. of 
sterile water, thoroughly shaken, and from this four plates were 
made which gave the following results: 
1) Developed . 78 colonies 
2) Developed . 76 colonies 
3) Developed . 65 colonies 
4) Developed . 77 colonies 
The average number of colonies was 74, and the percentage of 
variations was 17%. 
Comparing the two findings it is seen that the ring method de¬ 
veloped 27 against 74, that is, only about a third as many as the 
standard method, and whereas the standard method showed a 
variation of 17%, the plates in the ring method showed a 152% 
variation among themselves. 
The reason why the rings take up practically .01 of a gram of 
milk, but deliver only one-third of this amount, judging from 
the bacterial count, is because the rings are hollow and hold back 
the milk, and also because it is difficult to wash out of them all 
of the bacteria. 
3. Incubation of Cultures. The question of the best tem¬ 
perature at which milk plates should be incubated is, it seems to 
me, still an open one. The New York Commission, the Milk 
Committee of the American Public Health Association, as well as 
the Committee on Standards of the American Association of- 
Medical Milk Commissions, have, nevertheless, all decided in 
favor of the 37° C. count. The chief reason for this is, no doubt, 
that the results of an analysis can, in this way, be secured some 
days earlier than they can at a lowed* temperature. It is my ex¬ 
perience. however, that a larger number of bacteria always de¬ 
velop at a lower temperature, sav 21° C., than at 37° C. 
In the present investigation the temperature of the 'room 
(21°C.) has been employed exclusively. Any results not so ob¬ 
tained are definitely marked. The plates were counted after 
they had been grown for at least 5 days. The temperature was 
regulated quite closely, see chart, Fig. 12, as an example of the 
temperature variations. The reason for the use of this tempera¬ 
ture was a desire to get the maximum count. A strong argument 
for the use of a low temperature in milk analyses, where it is im¬ 
portant to determine the general character of the, bacteria, is 
