167 



placed over the radium, and stops all a particles which move 

 in any direction which is not almost vertical. Thus the par- 

 ticles cross the narrow chamber at right angles to its greater 

 dimensions, and all spend 3 mm. of their paths in the air 

 within it. The resulting ionisation being plotted against the 

 distance from the radium to the middle of the chamber, we 

 obtain an "ionisation curve," as in 

 Fig. 1, where ordi nates represent dis- 

 tances and abscissae represent ionisa- 

 tion currents. Each reading of cur- 

 rent is the difference between two 

 others, one measured when a very thin 

 copper screen is placed over the ra- 

 dium, and one when it is not. 



In this curve the portion .1 K is 

 due to the /? rays only, and repre- 

 sents the effect of such /? radiation 

 as is intercepted by the screen : the 

 chamber is out of range of a rays. 

 Let EA be produced to meet the axis 

 of X in D. The portion ABP repre- 

 sents part of the effect of the a ^^^- ^• 

 particles from RaC. If no other radio-active substances were 

 present, the curve would show a continuation of the portion 

 BP down to the axis of r, in some such manner as PC. 



If the ionisation curve were comj^leted in this way, the area 

 ABPCD would represent the total ionisation due to the a 

 radiation from RaC. If now the air were removed, another 

 gas substituted for it^ and the area re-measured, a comparison 

 of the values obtained would give the result which this re- 

 search aims at. We may call it the specific ionisation of the 

 gas. But the complete determination of the boundaries of 

 the area is so long and complicated a process as to render this 

 procedure impracticable. It can, how^ever, be shown that the 

 product of the co-ordinates of a certain point on the curve 

 may be taken as a measure of the area of the curve, provided 

 certain assumptions are made. The point is at the intersec- 

 tion of the top portion of the curve representing the effects of 

 RaA with the side of that showing the effects of RaC. The 

 co-ordinates of this point are comparatively easy to obtain. 



Now, it might appear that it would be better to measure 

 at one time the whole of the ionisation produced by the 

 particle, rather than to determine the ionisation point by 

 point along its path : since, if this were done, it would no 

 longer be necessary either to find the exact form of the ionisa- 

 tion curve, or to depend upon tlie validitv of assumptions. 



