RELIABILITY OF BACTERIAL COUNTS 387 
the numbers of samples analyzed. If five samples of the same milk are tested 
the results may be relied upon as fairly accurate, and always sufficiently accurate 
to place any particular milk supply unhesitatingly in Grade A, B, or C. The 
object of bacterial tests of milk samples for the numbers of bacteria should be 
primarily to determine the sanitary character of the milk supply from which 
the sample is taken, rather than the character of a single sample of milk. It is 
strongly urged by this commission that no grading of milk should be made upon 
the analysis of single samples, and that no prosecutions or court cases should be 
brought upon the bacterial analysis of a single sample of milk. 
Interpretation of Bacterial Tests. The commission has put its opinions on 
this subject in the form of resolutions, as follows: 
Whereas milk is one of the most perisnable foods, being extremely suscep- 
tible to contamination and decomposition; and 
Whereas the milk consumer is justified in demanding that milk should be 
clean, fresh, and cold, in addition to having the element of safety; and 
Whereas milk which is from healthy cows and is clean, fresh, and which has 
been kept cold, will always have a low bacterial count; and 
Whereas milk that is dirty, stale, or has been left warm, will have a high 
bacterial count; therefore it is resolved: 
First. That the health officer is justified in using the bacterial count as an 
indicator of the degree of care exercised by the producer and dealer in securing 
milk from healthy cows and in keeping the same clean, fresh, and cold; and 
Second. “That the health officer is justified in condemning milk with a high 
bacterial count as being either unhealthy or decomposed, or containing dirt, 
filth, or the decomposed material as a result of the multiplication of bacteria 
due to age and temperature. 
Third. That the health officer is justified in ruling that large numbers of 
bacteria are a source of possible danger, and that milk containing large numbers 
of bacteria is to be classed as unwholesome, unless it can be shown that the bac- 
teria present are of a harmless type, as, for example, the lactic acid bacteria in 
buttermilk or other especially soured milks. 
Grading by the Bacterial Count. Concerning the number of tests which 
should be made in order to determine the grade of a milk supply, the commission 
recommends that the grade into which a milk falls shall be determined bac- 
teriologically by at least five consecutive bacterial counts, taken over a period 
of not less than one week, nor more than one month, and that at least four out 
of five of these counts (80 per cent) must fall below the limit or standard set for 
the grade for which classification 1s desired. 
The grading of milk has necessarily been based on its sanitary character, 
primarily as determined by the bacterial test. The enforcement of grading, 
therefore, requires the application of the bacterial test in a manner sufficiently 
comprehensive to fairly determine the sanitary character of milk so that it may 
be assigned to the grade in which it belongs. Such an administrative system 
greatly modifies the former conception of milk inspection by public health 
officials. The inspection service under the grading system becomes subordinate 
