52 THE FLOATING-MATTER OF THE AIR. 



of September. Each tube was about half-filled with 

 the liquid. The tubes were then immersed in a bath of 

 brine, raised to ebullition, and permitted to boil for five 

 minutes. Aqueous vapour rose from the liquid into the 

 chamber, where it was for the most part condensed, the 

 uncondensed portion escaping, at a low temperature, 

 through the bent tubes at tlie top. Before the brine was 

 taken away little stoppers of cotton-wool were inserted 

 in the bent tubes, lest the re-entrance of the air into the 

 cooling-chamber should at first be forcible enough to 

 carry motes along with it. As soon, however, as the 

 outside temperature was assumed by the air within the 

 case the cotton-wool stoppers were removed. 



The front and back of this chamber were squares of 

 14 inches the side, the depth of the chamber being 8*5 

 inches. It contained, therefore, 1666 cubic inches of 

 air, which had unimpeded access to the liquid in the 

 tubes. No stoppers were employed. The air was un- 

 affected by calcination, or even by filtering. Neither 

 cotton- wool nor hermetic sealing was resorted to. Self- 

 subsidence was the only means employed to rid the 

 ' untortured ' air of its floating matter. 



A second series of eight tubes were filled at the same 

 time with the same liquid, and subjected to the same 

 boiling process. The only difference between the two 

 series was, that these latter tubes were placed in a stand 

 beside the case containing the former ones and exposed 

 to the common air of tlie laboratory. 



For the sake of distinction I will call the tubes 

 opening into the case the 'protected tubes, and those 

 openiiig into the common air tlie exposed tubes. 



On tlie 17th of September all the protected tubes 

 were bright and clear, while all the exposed tubes were 

 distinctly turbid. Specks of mould, moreover, were in 

 every case seen on the smface of the exposed liquid. 



