150 BELL SYSTEM TECHNICAL JOURNAL 



article. "Cosmic rays" are by definition the cause of that residue of 

 the ionization of the atmosphere, which is left over after deduction of 

 all of the ionization which can be traced to the action of the rays of 

 radioactive substances, or any other known cause whatever. 



I write "ionization" instead of "conductivity": this is because it is 

 well enough known that when a volume of gas is and remains con- 

 ductive, it is constantly being ionized — which is to say, electrons are 

 constantly being detached from its molecules. Molecule and electron 

 become ions, the members of an "ion-pair"; if there is an electrified 

 body nearby, those of the sign opposite to its charge are drawn to it 

 (those of the other sign going to the earth) and step by step their 

 charges neutralize its own. The question set above may therefore be 

 rephrased: Are ion-pairs appearing in the atmosphere oftener than 

 rays of known varieties are making them? 



To answer this question, one must of course exclude all of the rays 

 from radioactive substances (and, naturally. X-rays and other artificial 

 kinds) from the volume of air which one is studying; or if they cannot 

 be excluded altogether, one must estimate the ionizing-power of those 

 which remain. They are important in cosmic-ray research as causes 

 of error; they are important also in another way, as tentative models 

 for theories of the cause of the residual ionization. In attempting to 

 explain this mysterious residue, the procedure both natural and wise is 

 to begin by supposing it similar to alpha-rays, or else to beta-rays, or 

 else to gamma-rays. In the end it may be found needful to postulate 

 some ionizing-agent entirely different from all of these; but in the 

 beginning, they should be the guides. 



Of the three named classes of rays, one (the alpha-particles or posi- 

 tive corpuscles) has a sharply-limited range in air so short, that such as 

 come from the ground are stopped within a few inches ; and a range in 

 metal so extremely short, that even if one wished it would not be 

 possible to build a durable air-chamber with solid walls so thin that 

 the alpha-rays could get through. 



The second (the beta-particles or fast electrons) have a range which 

 varies from one radioactive element to another, and even from one 

 particle to another among those emitted by a single element. This 

 range is often much longer than that of an alpha-particle, and less 

 precise. Air at a distance of many metres above the ground may be 

 sensibly ionized by beta-corpuscles proceeding from the soil. All 

 along their paths, beta-particles engender ion-pairs profusely — in air 

 at normal (sea-level) density, never fewer than forty per cm. Never- 

 theless they do not enter an air-chamber walled by a very few centi- 

 meters of lead, and need not be a serious cause of confusion. 



