114 Hon. J. Strutt on the Preparation and Properties oj 



metal than merely to bring it in contact with a solid one. 

 Air bubbled through hot mercury (at al^out 300° 0.) was 

 accordingly tried next. This was found to give a large 

 effect, and was therefore selected for study first. To obtain a 

 strongly active gas the following procedure was adopted. 



A mercurial air-pump was used to circulate the air conti- 

 nuously through a gently sloping tube containing mercury 

 heated to 300° C, in which this bubbling took place. The 

 air, after passing through this tube, was drawn through a 

 stopcock into the pump-head. It passed out from the blow-off 

 of the pump to a vessel connected with the electroscope, in 

 which the electrical leakage could be tested. Thence it 

 passed to the bubbler again. The pressure in the testing-vessel 

 was always atmospheric while an electrical test was being 

 made, the bubbling being stopped meanwhile. 



As the circulation proceeded the rate of leak steadily 

 increased, until after some days it reached the value of 

 250 scale-divisions per hour. The leak with untreated air 

 was only 2 scale-divisions ; so that it had be^n increased 

 125-fold. 



In this experiment the mercury was heated, as I have said ; 

 but this is not essential. 



Active air can be obtained by bubbling air through cold 

 mercury; I am inclined to think that this is not quite so 

 effective. But no exact quantitative comparisons have yet 

 been made. It has been found that after prolonged bubbling 

 the mercury becomes temporarily exhausted, and that it 

 recovers its power on standing. The full description of these 

 experiments is reserved for a future paper. 



A part of the most active sample of gas was placed in a 

 testing-vessel, and allowed to remain some hours. It was 

 then pumped out and replaced by ordinary air. There was 

 found to be still a considerable leak ; it amounted, imme- 

 diately after admission of the air, to about \ that which 

 the radioactive air had given. This leak fell in about 20 

 minutes to half its original value *. There can be little 

 doubt that this activity left in the vessel after the removal of 

 the active air is of the same nature as the " induced radio- 

 activity " discovered by Rutherford in the cases of radium 

 and thorium. This conclusion is further supported by a 

 series of accurate experiments made in order to observe the 

 rate of decay of the activity of the radioactive gas itself. A 

 portion of this gas was inclosed airtight in a vessel, and the 

 rate of leak observed from time to time. 



* The law of decay lias not yet been accurately determined. 



