552 Mr. A. Wood on Spontaneous Ionization of 



was too short-lived to reach the measuring vessel, and, indeed, 

 this seems a very probable explanation. Other indirect 

 evidence of the existence of an emanation will be cited towards 

 the end of the paper ; and it will be seen that, although the 

 research cannot claim to have proved that ordinary matter is 

 radioactive, it has at least established a strong probability in 

 favour of that view. 



The vessel in which the ionization was originally measured 

 was of tin, about 13 cm. high and 6*25 cm. in diameter. The 

 top of the tin was perforated to admit an ebonite plug about 

 2 cm. in diameter. Through this plug there passed a brass 

 tube about 1 cm. in diameter containing a sulphur plug which 

 carried the electrode. The vessel was kept charged to a 

 potential of about 250 volts by means of a battery of small 

 storage-cells, one terminal of which was connected to the 

 vessel, while the other was connected to earth. The 

 brass tube was also connected to earth, and as it projected 

 right through the ebonite plug it prevented all possibility of 

 leakage from the vessel to the electrode across the insulation. 

 The electrode was a stout brass wire one end of which passed 

 down almost to the bottom of the vessel, while the other was 

 connected to the electroscope. As the electroscope has been 

 very fully described by Wilson*, it is unnecessary to give the 

 details of its construction here. It was found possible to 

 make it extremely sensitive, but great sensitiveness was found 

 to introduce compensating disadvantages which more than 

 outweighed the gain Thus minute changes in the potential 

 of the cells made the gold-leaf very unsteady and rendered 

 it difficult to take an accurate reading. The apparatus also 

 became very sensitive to small differences of temperature, 

 the convection currents caused inside the electroscope moving 

 the gold-leaf through several divisions of the microscope- 

 scale. For those reasons it was found most convenient to 

 work with a sensitiveness of from 30 to 50 divisions per volt. 

 This sensitiveness was sufficient to enable one to measure 

 accurately the amount by which the electrode system became 

 charged up in an interval of twenty minutes. To one end of 

 a stout brass wire, the other end of which carried the gold- 

 leaf of the electroscope, was attached a mercury-cup into 

 which dipped the electrode- wire. Both wires were screened 

 throughout their entire length with earthed metal screens so 

 as to prevent induction effects from neighbouring charged 



* Proc. Camb. Phil. Soc. vol. xii. part ii. (1903). 



