302 Mr. Norman Campbell on Delta Rays. 



effects. If a large difference of potential (say 100 volts) is 

 applied to the terminals and then reversed, the current does 

 not attain a steady value for about a minute, but similar 

 changes have often been noticed in ionization currents. The 

 value of the steady current appears to be the same within 

 1 per cent, for either direction of the P.D. with any dif- 

 ference of potential between the terminals from 0*01 volt 

 upwards. Moreover, the resistances showed no lag when 

 rapidly changing potentials of the order of 0"1 volt were 

 applied to the ends. 



Within a range f O01 volt to 400 volts the resistances 

 appear to obey Ohm's law perfectly within the experimental 

 error. The temperature coefficient has not been investigated 

 directly, but determinations of the conductivity made during 

 three months, when the temperature of the laboratory must 

 have varied 5°, agree within 1 per cent. No change of the 

 resistance with the time after the first 24 hours of prepara- 

 tion has been noted. 



It will be seen that these liquid resistances have many 

 advantages over those of Bronson constructed with uranium 

 or polonium. They appear able to be constructed with a 

 resistance over a much wider range ; the range mentioned 

 could certainly have been extended widely below the lower 

 limit. They are inexpensive and easy to construct, and they 

 do not (like polonium resistances) change with the time. 

 Moreover, the wide range over which Ohm's law is appli- 

 cable is an immense advantage ; it not only increases their 

 usefulness for general purposes, but also renders their re- 

 sistance extremely easy" to determine. The resistance of a 

 tube intended to be used for very low differences of potential 

 can be measured by means of a high potential battery and a 

 standardized galvanometer. Indeed, it appears that if such 

 resistances were carefully standardized they would provide a 

 far more convenient and much cheaper means of reducing 

 electrostatic measurements of current to absolute units, than 

 the standard capacities which are ordinarily used or quartz 

 piezo-electric. They would probably not be quite so accu- 

 rate, but for many purposes a loss of accuracy would be 

 unimportant. 



Leeds, April 1911. 



