[Vor. 1 
74 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
reactions on all three differential media. Тһе presence of the 
single unidentified coccus is again noted. Хо salivary coccus 
forms were isolated, from which fact it appears that the air in 
the particular location from which the sample was taken was 
“safe.” 
Experiment 11.—This sample was taken on the balcony of 
the building, about 10 feet from the rear wall. Every seat was 
occupied. As indicated in table тх, two plates were examined. 
On the first, 7 colonies developed—3 streptothrix, 3 bacillus, and 
1 coccus. Оп the second plate 3 colonies appeared, all of which 
were of bacilli. The reaction of the coccus was negative on the 
three differential media, thereby indicating that the sample of air 
taken was free from salivary coccus forms and therefore “safe.” 
In summing up the results of the experiments carried out in 
the vaudeville house it is to be noted that in both cases no sali- 
vary coccus forms were found. Table rx further shows that 
the total number of organisms found per unit volume of air 
was smaller than in the street car experiments. 
EXPERIMENTS 4 AND 7 
These samples were obtained in the basement of a local 5 and 
10 cent store. The ceiling was rather low, being only about 
9 feet from the floor level, the entire basement having a volume 
of about 72,000 cubic feet. Тһе samples in these experiments 
were taken in the midst of a crowd gathered to listen to a singer 
advertising songs. Little attention was given to the matter 
of ventilation until after the results of the experiments were 
obtained. Subsequently, however, investigation revealed the 
fact that ample provision had been made for ventilation. 
Transoms at the level of the sidewalk provide openings to the 
outside; along the inside wall and near the ceiling are revolving 
fans about 20 feet apart. These keep the air in circulation 
until it is drawn out by a suction fan situated in one corner, 
about 2 feet from the ceiling. Тһе same sampling apparatus 
was used as in the preceding experiments. As before, a total 
of 10,800 ec. of air was drawn through the sand filter in each 
sampling at the rate of 900 cc. per minute. Тһе samples were 
plated as shown in table rx. 
Experiment 4.—The air sample for this experiment was taken 
