2 3 6 22 



ity of admission was also varied. In the first column 

 of the following table the quantity of air operated on is 

 expressed by the number of inches which the mercury 

 gauge of the air-pump sank when the air entered. In 

 the second column the condition of the platinum tube is 

 mentioned, and in the third the state of the air which 

 entered the experimental tube. 



Quantity of Air. State of Platinum Tube. State of Experimental Tube. 



15 inches . . Cold . . Full of particles. 



15 " . . Red-hot . . Optically empty. 



15 " . . Cold . . . Full of particles. 



15 " . . Red-hot . . Optically empty. 



15 " . . Cold . . . Full of particles. 



15 " . . Red-hot . . Optically empty. 



The phrase " optically empty " shows that when the 

 conditions *f perfect combustion were present, the float- 

 ing matter totally disappeared. It was wholly burnt up, 

 leaving not 9 trace of residue. From spectrum analysis, 

 however, we know that soda floats in the air ; these or- 

 ganic dust particles are, I believe, the rafts that support 

 it, and when they are removed it sinks and vanishes. 



When the passage of the air was so rapid as to ren- 

 der imperfect the combustion of the floating matter, in- 

 stead of optical emptiness a fine blue cloud made its ap- 

 pearance in the experimental tube. The following 

 series of results illustrate this point : 



The optical character of these clouds was totally dif- 

 ferent from that of the dust which produced them. At 

 right angles to the illuminating beam they discharged 



