806 METHODS OF INVESTIGATING BACTERIA. 



with an india-rubber cork, the centre of which is per- 

 forated, and contains a glass tube plugged with cotton 

 wool ; this glass tube is connected with the aspirator, 

 the other end is covered with a caoutchouc cap, in which 

 there is a central hole through which the air enters. 

 The whole tube is placed on a stand in a horizontal 

 position. 



The germs present in the air fall usually shortly 

 after the entrance of the air into the tube on the 

 gelatine, and develop there to form isolated colonies. 

 By this method we often obtain very instructive appear- 

 ances, but even this method does not give accurate 

 comparative results. For one thing, it is difficult to 

 ascertain the proper rapidity of the current of air, so 

 that no germs shall pass through the tube, and yet that 

 they shall not be too numerous near the entrance of 

 the tube ; again, the dry surface of the gelatine is 

 not a suitable place for the commencement of develop- 

 ment; and, lastly, a mass of organisms gives rise to 

 colonies which are equally well isolated as those which 

 develop from a single individual. 



Other Other attempts to determine quantitatively the germs 



of the air have also as yet failed in giving entirely satis- 

 factory results. Yon Sehlen's attempts to employ agar 

 nutrient solutions are open to the error that rapidly grow- 

 ing saprophytes may enter these nutrient materials while 

 the air is being slowly drawn through them ; and further, 

 that it is difficult to retain the germs in the fluid, and 



K ault8 x f . also to vary the composition of the nutrient substratum. 



the methods as J 



yet employed, It seems best to employ indifferent substrata for the re- 

 tkms for their ception of the germs of the air. This can be best done by 

 improvement, the use of cotton wool, glass wool, and similar substances. 

 The material employed for the filtration of the air, and 

 laden with the air germs, is then put into nutrient jelly, 

 and after this has been shaken up for about half an hour 

 (in order to break up any colonies that may be present) 

 it is poured out on plates. If we wish to employ a 

 variety of nutrient substrata the material used for 

 the filtration is in the first place shaken up with salt 

 solution, and equal known quantities of this solution are 



