144 BEI.L SYSTEM lECIIMCAL JOL'RXAL 



negative glow does not overarch the whole cathode, and the thickness 

 and current-density of the dark space keep their fixed mininunn 

 values, the voltage across it remains constant likewise. This is the 

 normal ralliode-fall of the glow. It is an even more thoroughgoing con- 

 stant than the thickness or the current-density of the dark space, for 

 these \ary with the pressure of the gas (the dark space shrinks both 

 in depth and in sidewisc extension, if the current is kept constant 

 while the gas is made denser) while the normal cathode-fall is iniiiuine 

 to changes in pressure. It depends both on the gas and on the ma- 

 terial of the cathode; the recorded values extend from about (iO \-olts 

 (alkali-metal cathodes) to about 400 volts. Attempts have been made 

 to correlate it with the thermionic work-function of the cathode metal, 

 and there is no doubt that high values of the one tend to go with 

 high values of the other, and low with low. When the cathode is 

 entirely overspread by the negati\c glow and the dark space begins 

 to thicken, the \'oltage across it rises rapidly; the cathode-fall is said 

 to become anomalous, and ma\' ascend to thousands of volts. 



Almost the whole of the voltage-rise from cathode to anode, as I 

 have stated, is generalh' comprised in the cathode-fall; the remainder, 

 although spread o\er all of the brilliant divisions of the discharge, is 

 inconsiderable unless the tube is unusually long. The field strength 

 in the Crookcs dark space is also much greater than anyw'here else in 

 the glow. This is illustrated b\' the two curves in Fig. 9, representing 

 the field strength in the discharges sketched above them. (For the 

 region of the (Vookes dark space, however, the curves are defective.) 

 In the luminous clouds the electric force is feeble, and the>- in fact 

 are not essential to the current-flow; if the anode is pushed inwards 

 towards the cathode, it simply swallows them up in succession with- 

 <iut interfering with the current; but the moment it invades the 

 Crookes dark space, the discharge ceases unless the electromoti\'e 

 force in the circuit is hastily pushed up. The mechanism wliiili keeps 

 the glow ali%'e lies concealed in the dark space. 



One naturally tries to in\enl a mechanism re.seml)ling llie one 

 suggested for the arc: the cathoile-fall serves to give cnerg\- to tJU' 

 electrons emerging from the cathode, so that they ioni/e molecules at 

 the e<lge of the neg.iti\e glow; and the ions fall .igainsl the cathode 

 with energy enough to dri\e out new electrons. But the details are 

 more difficult to ex|)lain. The cathode-fall gives much more energy 

 to the electrons than they need to ionize any known molecule, so that 

 apparently its high \-.ilue is what the ions require to give them enough 

 energy to extract electrons from the cathode. We can hardly argue 

 that the electrons are thermionic electrons; the cathode does not 



