1352 Wisconsin Academy of Sciences , Arts, and Letters. 
equally divided among the ten small test-tubes, is put under an¬ 
aerobic conditions, and incubated for 48 hours. The tubes show¬ 
ing the characteristic reaction are counted. If all of the tubes 
are positive, the result is recorded as 10. All gradations between 
0 and 10 will be encountered. 
Turning now to the results obtained in this investigation (see 
Table VIII,) the raw milks used for pasteurization in the bulk 
nearly always, contained the organism under discussion, the 
number ranging from 0 to 10. The average for the three sup¬ 
plies were 2.5, 3.1, and 7, respectively. Raw milks produced 
for direct consumption were much less frequently and less 
heavily infected. The milks supplied to a small town, F x and F 2 , 
proved to be free from this germ, but F 3 , which is a bottling 
concern is infected in three out of five samples. Milks K and L, 
which had a high count, likewise have Bact. Welchii present. 
Inspected milks, are especially free, but occasionally, and es¬ 
pecially the poorer grades, do contain these germs. The num¬ 
ber varies from 1.7 to 3.4. The better grades of this class, 
and all of the certified milks are free from Bact. Welchii. 
In pasteurized milks the germ under discussion appears in 
practically the same frequency with which it occurs in the raw 
milk which has been treated. A 2 , B 2 , C 2 , show averages of 2, 
3 and 6.2, while 2.5, 3.1 and 7 are the averages of the raw milks 
(A x , B x and C x ). The milks pasteurized in the bottle (A 3 , and 
A 4 ,) have averages of 0.6 and 0.7, while these same milks in the 
raw condition (A 5 ) contain 1.8. 
There seems to be quite a striking correspondence between the 
distribution of Bact. Welchii and the lactose fermenters (coli-ae- 
rogenes group.) This organism is rarely found, and apparently 
never in any quantity in raw milks unless the coli-like organisms 
are also present in considerable quantities. Positive tests for 
both of these organisms is quite conclusive proof of manurial con¬ 
tamination. If only a single test could be made on a milk a posi¬ 
tive reaction of this organism would indicate more serious con¬ 
tamination than the presence of B. coli. 
In pasteurized milk this organism may be present in consider¬ 
able quantity and B. coli may be absent. This would mean that 
the original milk was badly contaminated, but that the pasteuriz¬ 
ing process had been well carried out. 
It seems then that this test is capable of being used to a very 
