572 Mr. A. Wood on Spontaneous Ionization of 



of the arrangement o£ the insulation is given in fig. 3. 

 Apparently when the vessel is charged up the ebonite takes 

 some charge also. This charge acts inductively on the elec- 

 trode, sending electricity of the same sign to the gold-leaf, 

 and so apparently increasing the leak. This would explain 

 why the leak always seemed to be greatest when the field was first 

 applied. If now the ebonite is able to hold more charge when 

 hot than when cold, the observed phenomena admit of a very 

 simple explanation. While the temperature of the ebonite is 

 rising the charge on it will be increasing, and so the apparent 

 leak will be large ; while the temperature is constant the leak 

 will be normal, and while the ebonite is cooling the reverse effect 



Fi£. 3. 



*! 



■ ffff/ISS Gu/lffD ll/g£ 



Sulphur 



fl£CT/r'0O£ 



£<?/7SS 

 Gw?D 76/8 £ 



COffM 



'SULPHO/? 



lTlECT/?0£>£ 



to that observed during heating will take place, and theapparent 

 leak will be abnormally small, or, if the temperature has been 

 high enough, the leak will even appear to be negative. The 

 following experiments may be adduced in favour of this 

 view : — If the field was applied to the vessel for some time, 

 and after the ebonite had had time to charge up the vessel was 

 earthed, heating the vessel produced a large apparent leak. 

 On the other hand, if the vessel had not been charged, then, 

 unless the field were on, no apparent leak could be obtained. 

 The changes in the leak were thus dependent on the charge 

 on the ebonite, and occurred whether the field was on or 

 not. In order to get rid of this source of error, the distribu- 

 tion of the insulation was changed, as shown in the diagram. 



