188 
compared to the number of those that are formed. When 
this is done it is found that it is still necessary to apply a 
high j^otential in order to extract all the ions from the gas. 
For example, in air at atmospheric pressure an electric force 
of 25 volts to the cm. will only extract about 80 % of the 
ions which are obtained when the force is increased to 1,000. 
The following examj^le will serve as an illustration : — The 
width of the ionisation chamber is 4 mm., the upper electrode 
being a metal plate, the lower a sheet of gauze- A thin layer 
of radium is placed 6*2 cm. below the sheet, and a particles 
emitted from RaC cross the chamber and ionise the air, 
which is at atmospheric pressure. The area of the plate on 
which the rays fall is about 18 cm. The capacity of the elec- 
trometer to which the upper plate is connected is about 150 
cm, and a potential of '125 volts applied direct to the electro- 
meter causes a deflection of 722 divisions on the scale : ten 
divisions — 1 mm. When the lower plate is raised to 400 
volts positive, so that the electric force is 1,000 volts per 
cm., there is a deflection of 982 divisions in 10 seconds, under 
the influence of the a rays. When a potential of 10 volts is 
applied, giving a force of 25 volts per cm-, there is a deflec- 
tion of 772 in 10 seconds. 
In the latter case the charge Q received per sq.cm. of 
electrode in one second, measured in electrostatic units, is — 
772 X 150 
= 8-5 X KM 
10x722x8x300x18 
The number of ions falling on each sq.cm. of electrode per 
second is therefore 1*2 x 10'' nearly. 
The velocity of ions at this potential gradient is nearly 
25 X 1"5, or 37 cm. per second. 
Thus, if n be the number of ions in a cubic centimetre, 
37w= 1-2 X 10'"', and therefore w = 3-2xl0^ Hence, the num- 
ber of recombinations taking place in a second in the space 
between two opposing square centimetres of the electrodes is 
equal to a x 4 x (3-2 x 10^)^ If we take the value of a to be 
3,400 X 3 X 10~^^ we find this number to be nearly 420. 
Finally, therefore, the number of ions recombining in each 
second' is 420, whilst the number received is 1-2 x 10'',andthus 
only 1 /3,000th of the ions are lost in this way. 
But the current at 25 volts is only 772/982, or about 
80%, of the current at 1,000 volts. 
It is clear from this example that there is some cause 
which prevents the current attaining its full value other 
than general recombination between positive and negative 
ions. 
Now, it is possible that ions newly formed might be 
specially liable to recombine with each other. Such a possi- 
