BACTERIOLOGICAL ANALYSIS OF AIR. 641 



In some of the methods which provide for the filtra- 

 tion of bacteria from the air by means of liquid sub- 

 stances a measured volume of air is aspirated through 

 liquefied gelatin ; this is then rolled into an Esmarch 

 tube and the number of colonies counted, just as is 

 done in water analysis. This is the simplest procedure. 

 An objection sometimes raised against it is that organisms 



FIG. 97. 



Petri's apparatus for bacteriological analysis of air. The tube 

 packed with sand is seen at the point a. 



may be lost, and not come into the calculation, by pass- 

 ing through the medium in the centre of an air-bubble 

 without being arrested by the fluid an objection that 

 appears to have more of speculative than of real value. 

 Filtration through porous substances appears, on the 

 whole, to give the best results. Petri recommends as- 

 piration of a measured volume of air through glass 

 tubes into which sterilized sand is packed. (Fig. 97.) 

 When aspiration is finished the sand is mixed with 

 liquefied gelatin, plates are made, and the number of 

 developing colonies counted, the results giving the 

 41 



