i6o Bulletin of Lahoratofies of Denison University. [Voi.xi. 
Before proceeding to a consideration of our data, it should 
be stated, that the samples were obtained by the use of a cheese 
borer or trier, the core of cheese placed in a sterile glass test-tube, 
and this was sent to the Laboratory in a wooden case, usually by 
mail. Through the kindness of Mr. J. W. Deeker of the Wiscon- 
sin Dairy School, most of the samples were so obtained, together 
with many important data. In all cases, one or more days inter- 
vened between the time of sampling and that of analyzing the 
sample. Although the exterior of the plug was carefully removed 
by means of a sterile knife, and only the center used, yet it is pos- 
sible that the normal distribution of the cheese bacteria may have 
been altered in some degree by the removal of the sample from 
its natural conditions and the consequent drying. Samples 33-36, 
however, were obtained directly from fair sized portions of the 
cheese. It would perhaps have been fortunate, if more of the 
samples had been so taken, so as to admit of a comparison. 
Usually three peptone gelatin Petri plates were made, 
using three dilutions of the cheese. The first contained, approx- 
imately, ^ gr. of cheese added to the tubes of gelatin and 
finely triturated by means of a sterile glass rod ; the second con- 
tained from 3 to 5 drops taken from the first; and the third 3 to 
5 loopfuls taken from the second. In this way a wide differ- 
ence in seeding was obtained, and one of the three plates usu- 
ally presented favorable conditions for counting regardless of 
the age of the cheese. In old cheese, colonies were usually lim- 
ited to plate No. i. New cheese gave the most favorable 
results on plate No. 3, etc. A special advantage in this method 
lay in the fact that moderately scarce germs, which would have 
been missed in Nos. 2 and 3 were brought out in No. i. 
This was particularly the case with the liquefiers or digesters, 
of which there was usually a trace on plate No. i, but none on 
the other two plates. 
Table II comprises 36 analyses of cheese, 26 of which were 
Wisconsin product and 10 from other states — a total of nine 
states being represented, which included a wide range of terri- 
tory, extending from New York and Pennsylvania on the east 
to Colorado on the west and Canada on the north. Within this 
