116 THK FLOATING-MATTER OF THE AIR. 



was thick and coherent the Bacteria died, or became 

 dormant, and fell to the bottom as a sediment. The 

 growth of mould and its effect on the Bacteria are very 

 capricious. The turnip-infusion, after developing in 

 the first instance its myriadfold Bacterial life, fre- 

 quently contracts mould, which stifles the Bacteria and 

 clears the liquid all the way between the sediment and 

 the scum. Of two tubes placed beside each other, one 

 will be taken possession of by Bacteria, which success- 

 fully fight the mould and keep the surface perfectly 

 clean ; while another will allow the moidd a footing, the 

 apparent destruction of the Bacteria being the con- 

 sequence. This I have proved to be the case with all 

 infusions, fish, flesh, fowl, and vegetable. At the pre- 

 sent moment, for example, of three tubes containing an 

 infusion of sole, placed close together in a row, the two 

 outside ones are covered by a thick tougli blanket of 

 mould, while the central one has not a single speck 

 upon its surface. The Bacteria which manufacture a 

 green pigment appear to be uniformly victorious in 

 their fight with the Penicillium. 



These observations enable us, I think, to draw some 

 interesting conclusions. From the irregular manner in 

 which the tubes are attacked we may infer that, as 

 regards quantity, the distribution of the germs in the 

 air is not uniform. A single tube will sometimes be a 

 day or more in advance of its neighbours. The singling 

 out, moreover, of one tube of the hundred by the par- 

 ticular Bacteria that develop a green pigment, and 

 other cases just adverted to, shows that, as regards qual- 

 ity, the distribution is not uniform. This has been 

 further illustrated by the following observations : — Of 

 five-and-twenty tubes of different animal infusions ex- 

 posed in groups of five, in the middle of November, and 



