February 8, iqooJ 



NA TUJ<E 



345 



Wlien the lamps are in use there is a very steep gradient of 

 electric potential al out them. I am informed that there is 

 a difference of about 50 volts be'- 

 tween the carbons of the Dovtr 

 arc lamps, and this entails steep 

 gradients in the air about them. 

 If, then, there were something to 

 cause a sudden change of potential- 

 gradient irrespective of the lamp, 

 this, when compounded with the 

 normal gradient due to the working 

 of the lamp, would give a gradient in 

 some places greater and in some less 

 than the normal. A path might thus 

 be opened for a discharge diflferent 

 from the normal one from pole to 

 pole, and this might pass away from 

 the lamp altogether, and even go to 

 the ground, provided there were a 

 sufficient gradient to continue it when 

 it had g(jt to a distance from the 

 lamp ; and such a gradient might 

 naturally exist in thundery weather. 

 Or, again, if there were something to 

 cause a sudden diminution in the 

 resistance of the air, a similar effect 

 might be produced. 



The striking of a flash would no 

 doubt be accompanied by a sudden 

 change in the atmospheric electric 

 potential. But I rather incline to 

 the other view, and to regard the 

 phenomenon as what I may call a 

 case of Nature's wireless telegraphy. 

 This view would make it depend on 

 electromagnetic waves propagated 

 from the flash. The flash would take 



the place of the sending instrument, the resisting air that ..f 

 the coherer, the gradient of potential, wl ether arli.i.ial (tha 



due to the electric works) or natural (that existing in thundery 

 weather), would take the place of the electromotive force of 

 the battery or cell which tends to send a current through the 

 coherer, while the electromagnetic waves would open a path for 

 the current in the air as in the coherer. 



The close similarity of the discharges from lamps three hun- 

 dred yards or more apart points to a distant cause, or at least 

 one which is much the same at places 100 or 200 yards apart in 

 a horizontal direction. This is not incompatible with the sup- 

 position that the path, when the discharge is fairly launched 

 from the lamp, depends on the atmospheric variation of the 

 atmospheric electric potential, which may very well be on an 

 extensive scale ; and such a similarity of path is what might have 

 been expected beforehand if the paths depend on electomagnetic 

 waves coming from the flash ; and it may well depend on a com- 

 bination of these two conditions. The difference from flash to 

 flash would seem to be in this way most easily accounted for. 

 For not only might different flashes, though close together in 

 time, come from different parts of the sky, but even if they 

 came from nearly the same quarter the mode of the transverse 

 ethereal vibrations would be pretty sure to be different. Now 

 the facility for the passage of a current aflforded by an electro- 

 magnetic disturbance would naturally depend jointly on the 



ilireciion 

 tendeil to 



rat inn and the direction in which the current 

 that is, on two independent vector quantities, 

 the pre-existing potential gradient and 

 the ethereal vibration ; and further, 

 the vibration might very well resemble 

 that in common rather than that in 

 plane polarised light ; so that there is 

 abundant room for complexity, and 

 for variation from flash to flash, as 

 regards the path of the local discharge. 

 The striation or beading of the dis- 

 charges remains to be considered. The 

 I>over lamps are worked on the alter- 

 nate current system, and it might at 

 first sight be supposed that perhaps 

 the beading may have something to 

 do with the alternation. But, as was 

 pointed out to me at the Dover electric 

 office, this could not be. For the 

 jieriod of the current is about the 

 liundredih of a seccmd, and if we 

 counted the beads we should arrive at 

 a duration of discharge much too great 



