618 BACTERIA IN THE AIR. 
many more than when the room has been undisturbed for some time. 
The painstaking researches which have already been made have es- 
tablished in a general way the most important facts relating to the 
distribution of atmospheric bacteria, but have failed to show any de- 
finite relation between the number of atmospheric bacteria and the 
prevalence of epidemic diseases. In the apparatus of Hesse, Fig. 
188, a glass tube, having a diameter of four to five centimetres and a 
length of half a metre toa metre, isemployed. In use this is sup- 
ported upon a tripod, as shown in the figure, and air is drawn 
through it by a water aspirator consisting of two flasks, also shown. 
The upper flask being filled with water, this flows into the lower 
flask by siphon action, and upon reversing the position of the flasks 
number one is again filled. By repeating this operation as many 
times as desired a quantity of air corresponding with the amount of 
water passed from the upper to the lower flask is slowly aspirated 
through the horizontal glass tube. The microérganisms present are 
deposited upon nutrient gelatin previously allowed to cool upon the 
lower portion of the large glass tube. The air enters through a small 
opening in a piece of sheet rubber which is tied over the extremity 
of the horizontal tube, and before the aspiration is commenced this 
opening is covered by another piece of sheet rubber tied over the 
first. Experience shows that when the air is slowly aspirated most 
of the germs contained in it are deposited near the end of the tube 
through which it enters. The colonies which develop upon the nu- 
trient gelatin show the number and character of living microérgan- 
isms contained in the measured quantity of air aspirated through the 
apparatus. The method with a soluble filter of pulverized sugar, to 
be described hereafter, is preferable when exact results are desired; 
and for the purpose of determining the relative abundance and the 
variety of microédrganisms present in the atmosphere of a given lo- 
cality the exposure of nutrient gelatin in Petri’s dishes is far simpler, 
and, as a rule, will furnish all the information that is of real 
value. 
In his extended researches made at the laboratory of Montsouri, 
in Paris, Miquel has used various forms of apparatus and has ob- 
tained interesting results ; but his method of ensemencements frac- 
tronnés requires a great expenditure of time and patience, and the 
more recent method with soluble filters is to be preferred. 
In his latest modification of the method referred to Miquel used a 
flask like that shown in Fig. 189. From twenty to forty cubic cen- 
timetres of distilled water are introduced into this flask. The cap A 
contains a cotton air filter and is fitted to the neck of the flask by a 
ground joint. This is removed during the experiment. The tube C 
is connected with an aspirator. It contains two cotton or asbestos 
