46 REPoRT OF THE BACTYERIOLOGIST OF THE 
occur through the distribution of the residue after filling the last 
hoop in any vat or through a confusion in placing the cheese 
upon the shelves. Constant care was exercised, and we believe 
that the data as given are substantially correct. 
The omissions during the first and the last weeks in September 
were accidental, and the latter is particularly to be regretted, 
Since it occurred just at the time of an outbreak. The only pre- 
disposing cause which could be found within the factory at this 
time was a change in the method of manufacture, which left 
considerable free moisture in the curd and allowed the small 
number of germs present to do a maximum amount of damage. 
During a considerable portion of the season the red points were 
so small and scattering that they would have been overlooked 
without a magnifying glass, and even with this aid they could 
not always be found. At no time were they sufficiently evident 
to cause financial loss except in one vat on August 2 and in a 
half dozen vats during late September. 
These favorable results were obtained notwithstanding the fact 
that during this time the cheese contained a high percentage of 
moisture and was uncolored, making the conditions very favor- 
able for the formation and observation of blotches. 
A single set of tools was used in all the vats, and the marked 
Showing of spots in the product of one vat, while they were 
practically absent in the others, suggested that the inoculation of 
a considerable quantity of Bacillus rudensis was necessary in 
order to result in any considerable showing in the cheese. 
The same observations regarding the difference in the product 
from different vats on the same day, which is well illustrated on 
August 2, could be best explained by the assumption that there 
was an outside source of contamination which our heating process 
did not reach and which at irregular intervals furnished a heavy 
inoculation of Bacillus rudensis in certain vats. 
For the purpose of testing this assumption samples were col- 
lected in steamed quart fruit jars from the milk of 21 patrons on 
October 24 and each manufactured into a cheese with the least 
possible exposure to contamination. When these small cheeses 
were examined on November 5 red spots were fairly plenty in 
two and a few were found in two others. | 
