474 Mr, Warren Be La Rue [Jan. 21, 



this is true only up to a certain limit ; after this has been reached it 

 rapidly increases, and ultimately the resistance becomes so great, as 

 the exhaustion is carried further, that it is easier for a spark to pass 

 between terminals placed at the same distance outside the tube in air 

 at atmospheric pressure. The diagram, Fig. 10, in which the abscissae 

 are the cube roots of the pressures in millionths of an atmosphere, 

 and the ordinates the number of cells necessary to produce a dis- 

 charge in a hydrogen tube 30 inches between the terminals, shows the 

 results of experiment. The pressure of minimum resistance varies 

 for different gases. We have determined it to be • 642 mm. = 845 

 millionths of an atmosphere for hydrogen, at which pressure the 

 potential necessary to produce a discharge through a tube with 

 terminals 30 inches apart was found to be only 430 cells. At a 

 pressure of 0-0065 mm., 8*6 millionths, it requires as high a 

 potential, 8937 cells, as at a pressure of 21 -7 mm., 28,553 millionths, 

 to cause the discharge to take place. At • 00137 mm., 1 • 8 millionth, 

 11,000 cells will not pass. The greatest exhaust we have obtained 

 in a hydrogen vacuum and an absorption by spongy palladium was 

 • 000055 mm., 07 millionth, which offered so great a resistance 

 that a 1-inch spark from an induction coil could not traverse the 

 tube. I will now exemplify what I have said by showing you a tube 

 with an absorption chamber (Fig. 11). I expect that the vacuum 

 will prove to be so good that the whole of the battery, 14,400 cells, 



Fig. 11. 



# 



2200 a-h Voluwes J ^zzr^^zzrC^^^^^ i^-S* 



will not cause a current to pass : you see there is no illumination of 

 the tube ; if I now heat the absorbing material so as to cause gas to 

 enter the tube, then the discharge of a much smaller number of cells, 

 namely 3600, illuminates the tube, and if I allow it to cool again the 

 discharge ceases. 



It has been suggested that there is a polarisation of the terminals 

 of a vacuum tube during the passage of electricity just like that 

 which occurs in a voltameter, and that this increases the obstacle to 

 the discharge ; but by an elaborate series of experiments we have 

 proved that such is not the case under the conditions of the experi- 

 ment. It is quite true that, after the connection between the battery 

 and the terminals of the tube has been broken, there is a deflection of 

 the needle when they are connected with a galvanometer, but we have 

 shown that this is entirely due to a minute static charge proportionate 

 to the capacity of the terminals.* 



* ♦ Roy. Soc. Proc' No. 205, 1880. 



