with Mercury contained in Tuhes. 531 



conducting electricity. It usually owes its existence to the 

 rapid movement of the mercury against the walls of the tube. 

 The phenomenon is the same as that occurring in an 

 ordinary mercury still ; it has been known since 1676 that 

 the motion of mercury in an exhausted glass vessel gave rise 

 to luminous effects, which since 1705 have been correctly 

 regarded as of electrical origin. (Cajori, History of Physics, 

 V . 96.) 



It is not, however, generally known that the effect can 

 readily be observed without the trouble of producing a 

 vacuous space and by employing the automatic pulsations 

 of mercury columns caused by heat. A quartz tube (10 cm., 

 4/3 mm., *46 mm.) was mounted between the glass bottles 

 as usual. This tube was then used in a variety of experiments, 

 and the mercury column always rapidly became continuous 

 when the current was switched off ; so that one may be 

 certain that it contained little of anything but mercury. 

 This tube was then placed *o that its central portion could 

 be conveniently heated by a Bunsen flame. The green glow 

 soon made its appearance. The portion of the tube imme- 

 diately above the burner appeared empty and gave out no 

 light ; but this region was followed on either side by about 

 2 cm. of space which glowed with the characteristic green 

 light. While the glow is present, the vapour is in a highly- 

 conducting state. A moving-coil milliammeter of four ohms 

 resistance was put in circuit with the tube, a rheostat, and a 

 large accumulator of 2 volts. When the column of mercury 

 as continuous, the reading of the ammeter was 50 milli- 



peres. The resistance in series with the lamp was 37 ohms 

 11. On applying the flame, the current varied irregular] v 

 from 4 to 5 milliamperes. Thus the effective resistance of the 

 space occupied chiefly by vapour was only about 400 ohms. 



It is to be noted that the conductivity is a concomitant 

 of the green glow. If the glow is absent and the columns of 

 mercury are separated, the vapour is non-conducting w hen 

 tested with low voltages. This matter was given particular 

 attention in a further confirmatory experiment performed 

 with a quartz tube (10*7 cm., 6 mm., *88 mm.). This was 

 put into an electrical circuit, as above. When the Bunsen 

 flame first caused the columns to separate, the current fell to 

 zero. When the glow commenced, the current rose from 

 5 to 6 milliamperes. On removing the flame the glow 

 continued for a short time, and so also did the current ; then 

 the glow and current ceased simultaneously. The ammeter 

 com in tied to record zero until the columns joined. 



It was found that the central portion of the tube was not 



w 



am 

 in a 



