222 G. C. SupPpLer, W. A. WHITING, AND P. A. Downs 
towns in the United States, and found that 189 of these provided for a 
legal limit for bacteria in milk sold within the municipality. The limits 
allowed by these cities ranged from 50,000 to 5,000,000 to the cubic centi- 
meter, with approximately one-half of the cities permitting a limit of 
500,000. The necessity of fixing legal limits for bacteria in cream seems 
to have been regarded as much less important, since only 30 of the 409 
cities had established legal limits for this product. The bacteria allowed 
in the latter case varied from 50,000 to 1,000,000 to the cubic centimeter. 
These municipai regulations must of necessity imply provisions for their 
enforcement and for penalties for failures in their observance. Such 
provisions immediately bring into prominence the difficulty of application 
and enforcement of numerical bacterial standards. Unfortunately, the 
inherent inaccuracies of present methods of enumerating bacteria are too 
great to permit their results to be relied upon with the certainty of exact- 
ness which their fixed numerical standards would seem to warrant. 
REVIEW OF PREVIOUS INVESTIGATIONS 
The American Public Health Association (1915), recognizing the wide 
variations obtained by the ordinary, plating technique, have formulated, 
through their Laboratory Section, the following uniform method for deter- 
mining bacteria in milk. This procedure, known as the “ Standard 
Methods of Bacterial Analyses of Milk,’ has been of considerable value 
in securing uniform technique in different laboratories, and the results 
are comparable, since a uniform interpretation can be given to them. That 
the purpose of the Standard Methods is for securing uniform results 
rather than accurate counts, in the minimum length of time, is evident 
from the fact that 37° C. for forty-eight hours on plain agar is the only 
incubation temperature and medium recognized. In the routine exami- 
nation of milk samples, the short incubation period has certain distinct 
advantages. 
Conn (1915) compiled the results obtained from an exhaustive series 
of comparative determinations made from the same milk by four labor- 
atories. This work, involving many thousand platings made under uni- 
form procedure, nevertheless failed to give uniform and consistent results 
under those particular conditions. Viewed from the standpoint of abso- 
lutely accurate determinations of all bacteria present, there are numerous. 
