Oct. 27, 1 881] 



NATURE 



615 



inconsiderable electromotive force to send an electric discharge 

 through even a moderate length of rarefied air, though it passes 

 far more freely through rarefied air than through air at the 

 ordinary pressure. I will endeavour to show you that experi- 

 mentally. 



[An experiment was here exhibited in which the coatings of a 

 Leyden jar were connected with the terminals of a Iloltz ma- 

 chine, and also, by two branches, with each other, each branch 

 involving an intenuption by air. One branch led through a 

 universal discharger, the brass knobs of which were separated 

 half or three-quarters of an inch, the other through a long tube 

 filled with rarefied air — a so-called aurora tube. The second 

 branch being at first broken, the knobs were adjusted to a 

 distance not too great to allow the spark of the jar to pass with- 

 out fail. The connection with the terminals of the aurora tube 

 being now restored, the discharge, which was at liberty to pass 

 by either branch, chcse the aurora tube.] 



It appears then that the resistance to the passage of the elec- 

 tric d ischarge is greater across about thi-ee-quarters of an inch 

 of air at ordinai7 pressure than across the whole length of the 

 tube, which I suppose is somewhere about five feet, so that, 

 although there is considerable re-istance to the passage of the 

 electric discharge through rarefied air, it is very much less than 

 through air at ordinary pressure ; lut although it is very much 

 less, it is very far indeed from being inconsiderable. Mr. De 

 La Rue has a splendid battery of about 1 1,000 cells of chloride 

 of silver. It required about 2000 of these to send electric dis- 

 charges through tubes perhaps two or three-quarters the length 

 of that, but not quite so broad, exhausted to such a degree as 

 to oppose lea-t re>istance to the passage. We see then that, 

 if one may judge l-y laboratory experiments, it requires a 

 very considerable electromotive force to send an electric discharge 

 through even a moderate distance in rarefied air. 



Now attempts have been made to measure the height of the 

 aurora, and very large figures have been brought out. It is said 

 to be fifty or sixty, or even eighty miles high ; I think some have 

 made.it even higher than that. It is a difficult matter of course 

 to measure with much certainty, because you want a base to 

 measure from ; and the two stations must be distant a {*iw miles, 

 in order that you may get a sufficient angle of parallax. Then 

 with observers situated a few miles apart it is a dilticult matter, 

 with such a variable nnd ind finite phenomenon as aurora, to fix 

 on what they should observe. Possibly in the future, when 

 such observers may be put in connection by telephone and be 

 able to speak to one another and tell each other n hat sort of 

 aurora they see, and settle by conversation at that distance what 

 particular part they shall observe, we shall get more certain 

 results. However, there can be no doubt that, although there 

 may be some uncertainty as to the precise height of the aurora, 

 it is very high indeed. 



Now, even in spite of this great height, the auroral streamer 

 subtends a very considerable angle at the eye of the observer. 

 If this be a discharge, the length of that discharge must be very 

 considerable, probably many miles. Where shall we get the 

 electromotive force sufficient to stnd a discharge through so 

 great an interval of air, rarefied though it be, and that not too 

 highly? I say "and that not too highly," because experiments 

 with exhausted tubes have shown that the resistance to the 

 passage of the spark through the tube goes on dimini-.hirg as 

 you make the exhaustion higher, until you reach a certain point, 

 after which it goes on increasing ag.nin, and this exhaustion, at 

 which the resistance to the pas-age of the discharge is least, is 

 by no means very considerable as exhaustions go nowadays. 

 Tubes have been so exhausted that rather than strike across a 

 millimeter within the tube from terminal to terminal, the dis- 

 charge would pass some inches cutside in air. Well, then, it 

 would appear from that that %Ye do not gain so very much as 

 regards facility of passage for an electric discharge by going up 

 to a tremendous height in the air. Where then can we get an 

 electromotive force sufficient to send an electric di-charge through 

 such a length ? Sir William Thomson, in the case of the Atlantic 

 cable vhich failed the first time, as I said before, obtained 

 earth-currents indicating an electromotive force of a few Daniell 

 cells ; but a few Daniell cells, or a few scores of Daniell cell«, 

 or a few hundreds of Daniell cells wculd be quite insufficient for 

 sending a discharge through such a space as I have spoken of, if 

 laboratory experiments are to be any guide. There is, however, 

 one instance of electric phenomena where we have tremendous 

 tensions to deal with — I mean atmospheric electricity. In the 

 case of the atujosphere we may have the electric spark striking all 



the distance from a cloud to the earth, perhaps half a mile or a 

 mile. I need not say that the electric spark I refer to is a flash 

 of lightning. Here I found some difficulty in getting the dis- 

 charge to strike through air of ordinary density, across more than 

 about three-quarters of an inch, lut in lightning it strikes all that 

 distance that I mentioned. Atmospheric electricity of tension 

 sufficient to strike across a mile of air at ordinary density (or at 

 least slightly reduced as ycu go up) might have an opportunity of 

 striking across many miles of rarefied air, and if the experiment 

 \\ hich I have showed you just now with that tube is to be any 

 guide, then it would be competent to do so. In atmospheric 

 electricity it is conceivable that we may have a sufficient tension 

 to cause the electric discharge to strike across that great distance 

 which the length of an auroral sti-eamer must be. 



It has long since been remarked that displays of aurorce seem 

 in some way or other in high latitudes to take the place of thun- 

 derstorms in low latitudes. Well, then, I will endeavour to 

 explain what I imagine takes place. I do not enter into any 

 speculation as to the cause of atmospheric electricity. We know 

 as a fact from its manifestation that it exists, and that is sufficient 

 for my purpose. Suppose now that the air, especially the higher 

 portions of the air, over a large tract of country, say to the north 

 of us, were more or less highly electrified — positively or nega- 

 tively, as the case may be — we will suppose positively — if the elec- 

 tric tension w ere sufficient, although, considering that the air is a 

 non-conductor, we might not have a flash of lightning, which 

 gathers into itself in one moment the electricity from an entire 

 cl ud and >ends it down into the earth (we might not have tension 

 enough to produce such a discharge, the resistance to the passage 

 of electricity from one portion of the air to another, wrhich at 

 any rate would be comparatively dry compared with what we 

 have in warm latitudes, would prevent it by itself alone), we 

 might nevertheless have a discbarge taking place in the higher 

 regions 6f the atmosphere where the air is rarefied, and accord- 

 ingly opposes less resistance to the discharge. 



Now let me refer to this figure. Thisgre.it circle, V^pe, I 

 suppose to represent a section of the earth by a plane passing 



through its centre. This blue [faint in cut] outside represents 

 the atmosphere, the height, of course, being enormously exagger- 

 ated, in order to make it visible at a distance. Suppose in some 

 way or other a portion of this upper atmosphere, as c, got con- 

 siderably charged positively or negatively, say positively; it 

 would act by induction on the earth below. The opposite elec- 

 tricity, negative in that case, would be accumulated underneath, 

 as at c, and this portion of the earth would form, as it were, a 

 portion of a Leyden jar, the lo«er atmosphere being the di- 

 electric or glass of the jar, the upper atmosphere being partly, 

 you may say, the di-electric and partly also the charged coating. 

 It would be represented more precisely by an imaginary coating 

 outside composed not of tinfoil, but of some badly-conducting 

 substance. The positive electricity about C would be bound 



