Necomhination of lon.'i in Air and Teniperatnre of Air. 659 



instiint the rays ceaso, and n is the number per c.c. remaining 

 at the end of a time t after the cessation of the rays, while a 

 is the coefficient of recombination. Now one of the objects 

 of this investioation was to test whether this equation repre- 

 presents the law of recombination at various temperatures 

 of the oas. To test this the three quantities to be measured 

 are /;, N, and t. 



When the levers L and L' were placed directly opposite 

 each other, as in the diagram (fig. 1), then the pendulum in 

 swinging along the path xy struck the two levers at the 

 same instant, and thus the rays were stopped and the electro- 

 motive force a})plied to the electrode E at the same instant. 

 Now if, when the ionization in the cylinder has reached a 

 ^steady state, the source of ionization be suddenly cut oiff and 

 a large electromotive force be applied to one of the electrodes, 

 the other electrode will receive a charge proportional to the 

 number of ions existing; in the gas between the electrodes. 

 Also if the electromotive force be applied to the electrode at 

 a given interval after the cessation of the rays, the other 

 electrode will receive a charge proportional to the number of 

 ions which are left in the gas. 



In the actual experiments, the observations were taken as 

 follows : — When the gas had reached the temperature to be 

 investigated, and the temperature had become steady, the 

 rays were started and allowed to run for an interval of ten or 

 fifteen seconds, so that the ionization might reach a steady 

 state. During this time both the electrodes E and E', as 

 well as the electrometer-quadrants, were connected to earth. 

 At the end of that time the electrode E' was insulated by 

 opening the key K, and the pendulum was released, which 

 broke the contacts, stopping the rays and applying the electro- 

 motive force to E at the same instant. In order to avoid 

 the violent kick of the electrometer-needle which would 

 occur, due to the inductive effect when the large E.M.F. was 

 applied to E, if the quadrants of the electrometer were insu- 

 lated at the same time as E, the electrometer was disconnected 

 bv the opening of the key K before the pendulum was re- 

 leased, and was kept to earth. Immediately after the passage 

 of the pendulum the lever L' was replaced in contact with Q', 

 so that the induced charge was counterbalanced by putting 

 E to earth again ; then the quadrants of the electrometer 

 were insulated and connected to E' by closing the key K, 

 and the deflexion of the electrometer-needle observed. This 

 deflexion should therefore be proportional to the maximum 

 number of ions N in the given volume of the gas. This then 

 frave a measure of the quantity N. 



