bacteria. The dry cakes had few moulds, but had a large number of bac- 
teria. All the dry cakes had alum present, not in sufficient quantity to 
act on the bread, but the alum evidently had been used as an antiseptic in 
the mash. Slack sponges were also made of the yeasts, and showed 
marked variations as regards time and extent of fermentation. This is a 
test that could be made easily by the housekeeper. : 
Pure cultures were then made of the yeasts. For this there were used 
eight moist cakes, six dry cakes, six yeasts from the air (four of these 
being red yeasts), two from cider, one from flour, four from fruits (grape, 
guava, persimmon and apple), two separated from corn smut and two 
from beer, making 31 in all. The pure yeasts were tested (1) for gas 
production, for which purpose beer wort was used in fermentation tubes; 
(2) for their action in solutions of sugars—sucrose, dextrose, lactose and 
maltose; (3) to determine the death limit of temperature for young and 
old cells; (4) for their fermentative action in slack sponges and _ stiff 
doughs, the latter being baked so as to determine the flavor imparted to 
the bread. 
The yeasts were examined microscopically, and it was found that ma- 
terial differences existed among some, while others were so much alike 
as to be indistinguishable from one another. The wild yeasts and those 
from the dry cakes were, in general, smaller than those from the moist 
cakes, and they also developed a film sooner. The yeasts from the moist 
cakes were large and resembled the beer yeasts. The appearance of the 
yeasts when grown on solid media will sometimes show variations that 
aid in the determination. 
In gas production there was also much difference, the moist cakes and 
beer yeasts producing more gas and in a shorter time, as a rule, than the 
others. Some of the air yeasts produced no gas. 
In the sugar solutions there were peculiarities appeared, in that certain 
yeasts would grow vigorously in-one sugar and not in another, the yeasts 
showing different preferences. 
For determining the death limit the yeasts were taken when five days 
old, 21 days and 30 days, and were tested, beginning at 65° C., for three 
minutes, then running up until the death point was reached. The death 
point varied between 80° C. for five minutes and 95° €. for 15 minutes, 
so that every one of them would be killed in the baking, in even the short 
time required for biscuit. 
