66 THE FLOATING-MATTEK OF THE AIR. 



the merest speck of a vegetable infusion containing 

 Bacteria infects all animal infusions, and vice versa. 

 The bursting of a bubble infects an infusion reached by 

 the spray. It is the envelope, and not the gas of the 

 bubble, which produces this result. 



Other experiments on hay-infusions, acid, neutral, 

 and alkaline, placed in contact with air purified in 

 various ways, yielded in 1875 the same negative re- 

 sult. 



§ 9. Infusion of Sole. 



The fish was cut up and digested for three hours in 

 water kept at 120° Fahr. On the 17th of November it 

 was introduced into a case containing three test-tubes, 

 and boiled there for five minutes. Three other tubes 

 hung outside the case were exposed to the ordinary 

 laboratory air. 



The three exposed tubes were feebly but distinctly 

 cloudy on the 19th. On the 22nd they were all thickly 

 turbid. Scattered spots of Penicillium then appeared 

 on two of them, while the third tube, which stood be- 

 tween these two, kept the Penicillium down. This 

 central tube contained the pigment-forming Bacteria, 

 which have frequently shown a singular power in pre- 

 venting the development of mould. For nearly two 

 months the central tube successfully withstood this de- 

 velopment, while its two neighbours were covered by a 

 matted layer of Penicillium. 



During the whole of this time the protected infusion 

 continued as clear and colourless as distilled ivater. 



§ 10. Liver-Infusion. 



On the 10th of November the infusion was prepared 

 by the process of digesting already so often described. 



