SOME roA-/7-.\//'( )/»•.(/>■)• .;/M-.f.V(7:\ IX I'liYsics n iis 



electrons enter the vacuum freely; electricity has no liorror of a 

 vacuum, any more than nature generally. Still there is sometliin^; 

 which suggests the horror vaciii of the scientists before (lalileo; for 

 the electrons which arc alreaily partway across the vacuum tend, by 

 their electrostatic repulsion, to [lush back their followers which are just 

 emerging from the metal. This is the space-charge effect, which bns 

 liecome famous since the audion became almost as common an ob- 

 ject as the incandescent lamp in the American home. I shall presently 

 have to write down the equations describing this effect; for the time 

 In-ing we may ignore it, so long as the electron-stream is not more 

 profuse than a photoelectric current gencralh- is. The electrons of 

 these scanty discharges enter into the vacuum and pass over without 

 hindrance. 



At this point it is advisable to say what is meant by a "vacuum." 

 Scientists are growing more exigent year by year in their use of this 

 term; thirty or forty years ago people spoke of "vacuum tulx;s" mean- 

 ing tubes so full of gas that they would transmit a big current with 

 a resplendent luminous displa\', but this self-contradicting usage has 

 become quite intolerable. At the present day the least density of 

 gas, or the highest vacuum, commonly attained corresponds to a gas- 

 pressure about 10" " as great as the pressure ancfdensity of the atmos- 

 phere. This means that there are about 10~* molecules in a cubic 

 centimetre of the "vacuum," which may make the name sound absurd. 

 But the practical criterion for a \acuum is not whether the remaining 

 atoms seem many or few, but whether they are numerous enough \.o 

 alTcct the passage of a discharge; and as an electron shooting across a 

 tulx? 10 cm. wide and evacuated to this degree has 999999 chances 

 out of a million of getting clear across without encountering a molecule, 

 the tulx." is vacuJHis enough for any sensible definition. 



Next we will imagine that a gas is introduced into the tube, in 

 quantity sufficient so that each electron going from cathode toward 

 anotle will collide on the average with one or possibly two atoms on its 

 way. It is best to Ijegin by thinking of one of the noble gases, of which 

 helium, argon and neon are the ones in common use; or of the vapour 

 of a metal, mercury vapour being much the easiest of these to work 

 with; for their atoms liehave in a simpler and clearer manner toward 

 the electrons than do the molecules of the commonest gases, particu- 

 larly the oxygen molecules which are so numerous in air. In fact the 

 practice of using the noble gases and the metal vapours — that is to 

 say, the monatomic gases — whcre\er possible in these researches ought 

 really to be regarded as one of the great advances of the last few years; 

 our predecessors would certainly have learned more about the dis- 



