33 



the slit was in a plane perpendicular to the axis of the ioniza- 

 tion chamber, and a small sliding door cut in what has been 

 so far called the lower gauze of the chamber enabled tha 

 cell to be introduced when necessary. The air was carried 

 across the radium cell and through the ionization chamber 

 with a velocity of 7 cm. per sec. A thin aluminium sheet, 

 sufficient to stop all the a particles, could be moved over the 

 slit when required, and the difference between the readings 

 without and with this sheet over the slit gave a measure of 

 the ionization produced by the a particles. 



In a particular experiment with a piece of tinfoil cover- 

 ing the radium, equivalent to about 1*5 cm. of air at 

 atmospheric pressure, and working under a pressure of two 

 atmospheres, the saturation current with the radium in the 

 ionization chamber corresponded to a deflection of 80 scale 

 divisions in 10 sec. With the radium 8 mm. from the gauze, 

 and with a draught of 7 cm. per sec, the current was 53 

 scale divisions per 10 sec. Thus the loss of ions in the first 

 "115 sec. after the ionization was produced = 34 per cent. 

 By increasing the pressure and determining the saturation 

 current with the radium placed directly in the ionization 

 chamber it was possible to determine the greatest density 

 of ionization which had been produced in the previous ex- 

 periment. An increase of pressure equivalent to a shortening 

 of the range of the a particles by '34 mm. gave a decrease 

 in current corresponding to 30 scale divisions per second. 



The section at the opening of the slit was *08 cm. x 1 cm., 

 the capacity of the system 125 cm., and the electrometer 

 gave a deflection of 340'0 scale divisions per volt. Thus, 

 4*1 X 10® ions are produced per cub. cm. per second just be- 

 yond the slit. The greatest density produced, therefore, in 

 the air as it is blown across the slit is 4*7 x 10^ ions per 

 cub. cm. Now, if we suppose all the ionization to be of this 

 same density, the number of ions which will disappear by 

 ordinary recombination in the period '115 sec. is 36 per 

 cent. The actual number as found was 34 per cent. ; it is, 

 therefore, seen that to account for the 34 per cent, 

 difference in ionization which has occurred after "115 sec, if 

 we are to suppose it due entirely to ordinary recombination, 

 a value equal to that of the maximum density must be 

 assumed for the whole of the ionization. Now it is clear, 

 not only from the geometry of the beam of radiation which 

 is used, but also from the fact that the ionization proceeding 

 from a small heap of radium bromide decreases with the dis- 

 tance from the radium, that the density of ionization in the 

 beam must reach values very much smaller than the maxi- 



