ivith Electron Currents in Different Gases. 437 



Emission of Radiation. 

 At low pressures of the order of 0*001 mm. when the kick 

 from A to A X A 2 (fig. 2) does not set in until potential 

 differences exceeding 30 volts have been applied, we have 

 not been able to detect any evidence of the presence of a 

 visible discharge either on the low range A or on the high 

 range AxA 2 . The tests were carried out by visual observations, 

 both directly and through blue glass, with appropriate dia- 

 phragms to eliminate the glare from the filament, and by 

 examination of the tube with a large direct-vision spectro- 

 scope by Hofmann. It may be that there is a faint luminous 

 discharge which is swamped by the diffused light coming 

 from the hot filament ; but, if so, its intensity must be 

 very low. Under these conditions there is, however, an 

 emission of ultra-violet radiation, whose presence can be 

 detected by its effect on the electrode E (fig. 2). This is 

 shown by the following examples : — 



(a) At p = about O'OOl mm. and a heating current of 

 1*05 amps, through the filament the thermionic current at 

 40 volts was 429 microamps. At this potential (corresponding 

 to A, fig. 2) the current suddenly jumped to 1056 microamps. 

 at 39*5 volts on the higher range. The point corresponding 

 to Ai was located at 36 volts. No effect on the electrometer 

 connected to E could be detected on the low-current range 

 corresponding to O A below 30 volts, but at 36 volts 5 divi- 

 sions in 30 seconds were obtained. On the range corre- 

 sponding to AjAg a deflexion of 50 divs. in 30 sees, at 40 volts 

 was observed. Evidently a large increase in the ultra-violet 

 emission sets in with the kick on to the higher range. 



(b) At a pressure similar to that in the experiment just 

 described, and with a heating current of 1*07 amps., the 

 current was found to jump up at 39 volts and down at 

 31 volts. At 39 volts the current was about 280 microamps., 

 and it jumped to 1848 microamps. at 36 volts. The following 

 observations up to 30 volts were taken on the low-current 

 range, and that at 40 volts on the high-current range : — 



Volts across gap A B — ► O (fig. 1)... 15 20 25 30 40 

 Electrometer deflexions in 30 sees.... 5 13 200 



In this case also there is a big increase in the ultra-violet 

 emission when the high-current discharge sets in, but there 

 was a distinct effect which set in between 20 and 25 volts 

 on the low-current range. This effect may have been present 

 in experiment (a) : but under the conditions of that expe- 

 riment it would have been too small to detect with any 

 certaiuty. It is difficult to be sure that these small effects 



