1^8 



We might spread a layer of radio-active material on the floor 

 of an ionisation chamber, and so arrange the temperature and 

 pressure of the gas in relation to the dimensions of the cham- 

 ber that all the a particles completed their paths within the 

 gas. But the potential gradient required to separate and 

 collect the ions made by the a particle is generally very 

 great. For example, in ethyl chloride at 30 cm. pressure 

 and ordinary temperatures, about 1,000 volts per cm. is desir- 

 able, if saturation is to be certain. With such gas it would 

 be necessary to make the height of the chamber about 4 cm., in 

 order to allow all the a loarticles to complete their ranges ; even 

 if the radio-active material were uranium or polonium. 

 Thus, a total potential of 4,000 volts would be required, and 

 such large electro-motive forces are out of the question. If 

 the pressure of the gas were lowered, less electric force would 

 be sufficient ; but the paths of the a particles would be longer, 

 the chamber would need to be higher, and the total potential 

 as great as ever. 



It is absolutely necessary to use a narrow ionisation cham- 

 ber if sufficient electric force is to be obtained without the use 

 of enormous battery power. Clearly it would be no gain to 

 use such a chamber if the radio-active material were spread 

 on one of its walls. For in this case some of the particles 

 would complete their full ranges within it, others only part, 

 and an estimate of the ionisation to be expected would ren- 

 der it necessary to take into account the amount of the range 

 completed by each particle as determined by the nature and 

 physical conditions of the gas and the dimensions of the cham- 

 ber, the reckoning beingf further complicated by the fact that 

 the ionisation produced by the particle is not constant along 

 its path. It is possible that an experiment might be arranged 

 in which a thin sheet of a radiation entered the chamber 

 through a slit at the side, and spent itself within the cham- 

 ber without touching the walls. It would be necessary to 

 make sure that the same portion of the range was completed 

 within the chamber by the particle, no matter with what ga« 

 the chamber was filled. I have not yet tried this plan. 



It will now be clear, I think, that the method actually 

 used is not without its advantages. It avoids the use of very 

 high potentials, and does not require lengthy and uncertain 

 calculations. It has also this in its favour, that it gives the 

 range of the particle in the gas, so that it i? possible to 

 make a sufficiently accurate estimate of the amount of any 

 air that may be present. The ])re'sence of this air can then be 

 allowed for. 



Let us, therefore, proceed to consider the assumptions 

 and approximations which the metlind requires. 



