with Electron Currents in Different Gases. 431 



the discharge in the condition corresponding to A x inde- 

 finitely, but that a limiting minimum value of this potential 

 difference is gradually reached. The lowest potential dif- 

 ference under which we have been able to maintain the 

 discharge in the state A T A 2 is 11*5 volts. The potential 

 differences are m ensured from the negative end of the hot 

 filament, so that this is the greatest potential difference 

 between any part of the hot filament and the anode. As 

 about one-tenth of the filament at each end is too cold to emit 

 electrons in appreciable numbers, the greatest potential 

 difference due to the field which is effective in driving the 

 electrons across the gap would be some 0*3 volt less than 

 this. On the other hand, the electrons are emitted with 

 kinetic energy of thermal agitation in accordance with 

 Maxwell's law. The allowance for this is somewhat inde- 

 finite, but would probably wipe out the 0"3 volt under con- 

 sideration. To maintain the discharge in the state AjA 2 

 therefore it appears to be necessary that the electrons should 

 be able to acquire an amount of energy equal to that given 

 by falling through a potential difference very close to 

 ll'O volts. This is nearly equal fco the sum of the cathode 

 and anode potential drops in the ordinary luminous high- 

 current mercury arc. According to Aron's observations the 

 former is 5*4 and the latter 7*4 volts. 



The emission of light from the discharges under consi- 

 deration will be dealt with below. 



The Low Potential Discharge. 



The currents with small potential differences (0 A, fig. 2) 

 exhibit characteristics similar to those shown in fio-. 2 over a 

 wide range of pressure of mercury vapour and of the tempe- 

 rature of the filament. The behaviour at a pressure of 

 mercury vapour of the order of 1 mm. and with a heating 

 current of 1*06 amp. in the filament is shown in greater 

 detail by the following numbers: — 



Volts 12 3 4 5 6 7 8 9 10 11 12 13 14 



(.n?ci-oa e n»ps). } ° l 3 lS 4 ' 5 6 ' 6 8S 10 ' 8 13 ' S 15 ' 3 18 '° 2l 24 28 " 3 33<8 4 7'& 



calcuhTJd 8 I ° °' 8 21 3 '° 6 '° 8 ' 8 10 ' 9 13 ' 8 1G ' 9 2(H 23 ' 6 27 ' 2 31 34 ' 9 39 ' 1 



The currents are roughly proportional to the potential 

 difference raised to the power 1*5. This is shown by the 

 numbers in the last row, which are calculated by assuming 

 that the currents vary as V 3 ' 2 , and that the value at 5 volts 

 is correct. Below 7 volts the agreement with this formula 



& 



