486 BELL SYSTEM TECHNICAL JOURNAL 



divide this number into the kinetic energy which it originally had, thus 

 obtaining the average energy spent per ion (or rather per pair of ions 

 generated, since each ionization produced two ions of opposite sign) — 

 a quantity amounting generally to several tens of equivalent volts (33 

 volts for air) .21 By performing such experiments with alpha-particles of 

 various initial speeds, it is possible to determine a function analogous 

 to the function /(F) defined for electrons in a previous section. This 

 function increases rapidly as the aplha-particle approaches the end of 

 its sharply-terminated trail, varying approximately as the reciprocal 

 of the cube root of the distance it has yet to go. 



Alpha-particles as they pass through a gas thus produce a countable 

 number of ions and suffer a measurable loss in kinetic energy. It is 

 interesting to enquire whether these two processes can be identified 

 with one another and explained by the nuclear atom-model — whether 

 the lost kinetic energy is altogether spent in detaching electrons from 

 the atoms of the gas and supplying them with extra kinetic energy. 



Before comparing any theory with the experimental data, one must 

 be aware of two complexities. In the first place, an alpha-particle may 

 transfer energy to an atom without ionizing it, so that the energy it 

 loses in passing through a gas may exceed that which it spends in 

 ionizing. In the second place, some of the ions produced by an alpha- 

 particle — notably, the detached electrons — may themselves be en- 

 dowed with energy enough to ionize, so that a measurement of the 

 total ionization in the gas may yield an excessive estimate of the num- 

 ber of ions actually and immediately produced by particles striking 

 atoms. Naturally the energy for producing all of these ions, "prirnary" 

 and "secondary" alike, comes from the alpha-particles, so that such 

 data as the aforesaid values for energy-spent-per-ion-generated have a 

 definite meaning. 



Discrimination between ions produced directly and indirectly is de- 

 sirable, indeed essential, for testing any theory; but thus far there is 

 no way for distinguishing the two, except in the case of very fast elec- 

 trons for which the trails have been photographed with great magnifi- 

 cation by C. T. R. Wilson- by his celebrated expansion-method, in 

 which each ion formed in the passage of such an electron through a gas 

 becomes the center of a visible droplet of water. In some of his pic- 

 tures, in Fig. 11, pairs of droplets and also groups of four, six and 

 more are seen. The paired droplets ha\e condensed upon the two ions, 

 positive and negative, produced by a single primary ionization (and 



-' R. W. Gurney, Proc. Roy. Soc. A107, pp. 332-340 (1925) and literature there 

 cited. 



'-■'Proc. Roy. Soc. A104, pp. 192-212 (1923). 



