1914] 
NOLTE—SALIVARY ORGANISMS AND AIR POLLUTION 67 
The volume of air exhausted from the bulb at each pressure 
was determined as follows: Тһе bulb, filled with water, was 
weighed. Pressure was then applied, forcing out the water, 
after which the bulb was again weighed. The difference in 
weight in grams is approximately the volume of air in cc. ex- 
hausted by a similar pressure. In the calibrations the results 
varied but slightly. By placing the fingers on the bulb in a 
certain fixed position each time, it was found that the bulb could 
be made to deliver 300 cc. of air at each exhaustion and, conse- 
quently, to receive 300 сс. of air at each release of pressure. 
Tt was, of course, necessary to have all joints air-tight, this being 
accomplished by making all connections with rubber tubing and 
glass and using plenty of overlap. 
Тһе sand filter, after being plugged at both ends with cotton, 
was sterilized for 30 minutes at 15 pounds pressure. Тһе rubber 
stopper support was allowed to fit very loosely into the tube dur- 
ing sterilization in order to prevent setting of the rubber. After 
the apparatus was removed from the autoclav, the stopper was 
immediately fitted in tightly, thus rendering the connection air- 
tight. Тһе sand filter was always used within 24 hours after 
sterilization. It was connected to the support as shown in 
plate 2. 
When operated in public buildings or conveyances, the sup- 
port, with the filter, was wrapped in stiff paper in such a manner 
as to permit of the easy operation of the pinch cock and exhaust. 
Тһе apparatus thus wrapped was held in the left hand and from 
it heavy rubber tubing passed down the left coat sleeve and then 
diagonally across to the right coat pocket where it was connected 
to the bulb. "This rendered the whole apparatus inconspicuous. 
'The bulb was operated with the right hand, the pinch cock with 
the left. А test tube with a sterile cotton plug was always 
carried, the latter being used to replace the plug which was 
removed from the intake of the filter at the beginning of the 
experiment. 
The plating was always carried out within 30 minutes after 
the sample was obtained. Тһе sand from the filter was carefully 
poured into a 100 се. flask containing 15 cc. of sterile distilled 
water. Тһе bolting cloth, which had a tendency to stick to the 
rubber stopper, was removed with sterile forceps and introduced 
