EHFFECT OF TEMPERATURE ON BACTERIA. 31 
be placed upon the differentiation of species. In some cases, 
where, as noted in the tables, the dilution was too high, and in 
some cases, where it was too low, the results are of less value 
than in other cases. ‘This discrepancy has thus far been una- 
voidable, and although it makes the results somewhat irregular, 
nevertheless, since in each series only a few of the tests were 
thus unreliable, the conclusions as to the general tendency of 
the results obtained were not materially vitiated by them. In the 
tables given below it will be seen that some of the analyses are 
wholly omitted because untrustworthy, and that considerable 
irregularities are shown in other cases. Hence, no individual 
analyses can be strictly relied upon, although the general course 
of development of species is probably reliable; a fact proved 
by the general agreement of the different experiments. In our 
experiments we planned to make at least three plates from each 
sample tested, and these three were always made with different 
dilutions. By this means one of the three was usually of a 
proper dilution, even if the others were not. Sometimes all three 
plates were usable, at other times only one was satisfactory. 
The culture medium used for analyses was that described in 
our previous report. It was usually a mixture of ordinary 
peptone gelatin, containing milk sugar, and gelatin made from 
whey as previously described. A mixture of half peptone and 
half whey gelatin was found to be most satisfactory. These 
materials were prepared as described in a previous report, with 
a little higher percentage of the ingredients than usual in such 
media, so that when subsequently diluted with one-fourth 
their bulk of litmus solution, the result was to obtain a medium 
with ordinary percentages. Hight cubic centimeters of this 
more highly concentrated culture medium was placed in test 
tubes and sterilized. At the time of using, two cubic centi- 
meters of sterile blue litmus solution was added to each test tube. 
- The result was a deep blue gelatin, upon which the acid bacteria 
were readily differentiated, and which, as described ina previous 
report, gave a possibility of considerable differentiation among 
the colonies, sufficient at least to differentiate ¢yfes of bacteria, 
though not, as a rule, accurate enough to separate species. 
The plates thus made were allowed to develop, if possible, at 
room temperature for five or six days, or a week, inasmuch as 
the differentiation of the bacteria from the appearance of their 
