572 



NATURE 



{Oct. 13, 1 88 1 



character as long as the coil alone was used ; but as soon as a 

 Leyden jar was introduce J, which was in the main equivalent 

 to an air-sparl; in a continuous current, the durational character 

 disappeared, aad nothing was viable but a bright line, the width 

 of which depended, not upon the duration of the discharge, for 

 no velocity of rotation in any way affected it, but only on the 

 width of the slit through which the discharge in the tube was 

 seen. But, notwithstanding the extreme rapidity with which the 

 discharge is effected, our experiments have already shown tliat 

 the sparlc or discharge is a complicated phenomenon, the various 

 parts of which take place in a certain order or sequence of time ; 

 and that in virtue of this sequence we have succeeded, at the 

 various pressures comprised within our range, in affecting ai.d 

 modifying it in transitu. This su/gested the idea that, although 

 the subject is surrounded with difficulties, it might still be 

 pos ible to form some relative estimate, at all events, of the 

 time occupied by the various parts of which the whole pheno- 

 menon is compj^ed. And in fulfilment of this the following are 

 some of the conclusions to which we have been led. 



The time occupied in the passage of electricity of either name 

 along the tube is greater than that occupied in its passage along 

 an equal length of w ire. 



This may be shown by connecting metallically a piece of tin- 

 foil near the air-spark terminal with another near the distant 

 terminal ; for it i^ then seen that the former derives as much 

 rehef from the latter as if the latter were not on the tulie. This 

 shows (i) that at the time when the electric disturbance reached 

 the nearer piece of tinfoil the more distant piece was unaffected, 

 and (2) that the di^tm'bance propagated along the wire reached 

 the second piece before the arrival of the same disturbance 

 propagated within the tube. 



The negative discharge occupies a period greater than that 

 required by the particles composini: the molecular streams to 

 traverse the length of the tube, but comparable with it. 



Proofs of thii proposition are to be found in the phenomena 

 of virtual shadow\s, and in other instances of the interference of 

 m:)lecular streams ; but, omitting detailed experiments, the gene- 

 ral argument on which the above conclusion is based is as follows : 

 If two molecular streams, one is-uing with positive relief from 

 the side of the tube, the other coming from the negative ter- 

 minal, show signs of interference, it is clear that the former of 

 ■.hese, which certainly started first, must have continued to flow, 

 at all ev nts, unlil the arrival of the latter. 



The time occupied by the passage of electricity of either name 

 along the tube is incomparably shorter than that occupied by the 

 emission of the molecular .streams, or (what is the same thing) 

 the time occupied by the negative discharge. 



In sup, ort of this conclusion we have time only to mention a 

 single experiment. If two pieces of tinfoil connected by a wire 

 be placed, one near the negative, the other near the positive end 

 of a tnbe through which a negative discharge with a rather long 

 nir-spark is passing, the former will show relief (positive) effects, 

 the latter special (negative) effects ; but no phosphorescence will 

 be caused at the latter, however long the air-spark used. When 

 the second patch is lifted off the tube and placed upon another 

 through which no current is passing-, phosphorescence is imme- 

 diately produced. The explanation of this appears to be as 

 follows : Tbe negative electricity, bursting into the tube, sum- 

 mons all the positive which it can draw from the tinfoil. Thi^ 

 IS answered so promptly, that the second patch gives up to the 

 first through the medium of the wire all the positive that it can 

 yield, or, which is the same thing, draws off from the first all 

 the negative that it can obtain; and ths is done before the 

 advancing negative reaches the distant patch. But so rapidly 

 does the negative .idvance, that it reaches the distant patch 

 bef jre the molecular streams have had time to flow from the 

 latter in a sufficient stream to produce phosphore.^cence ; and it 

 rea^^ches it in time to revoke the supply of positive to the nearer, 

 and to draw back the supply of negative w hich would have come 

 to, and with it the molecular streams which would otherwise 

 have flowed from, the further patch. When the second patch is 

 placed on an independent tube, where no such revocation is 

 possible, phosphorescence actually appears, showing that the 

 revocation k n > mere supposition, but a real phenomenon. 



From the last two laws it follows as a consequence that nega- 

 tive electricity, and therefore also electricity of either name, in 

 the tube outruns the molecular streams. 



We may now fairly ask whether the phenomena which we 

 have been studying have any counterpart in the larger operations 

 of nature which are going on around u-, and whether the con- 



clusions to which we have been led afford any explanation of 

 observed facts ? Many natural phenomena doubtless fundamen- 

 tally depend upon electricity ; how many we hardly yet know. 

 But there are two in particular, namely, lightning and the 

 aurora, which are unquestionably electrical, and whose corre- 

 spondence with the spark proper, and with the discharge in 

 rarefied gases, respectively has often been noticed. On these I 

 venture to offer a few remarks. 



To say that both of these phenomena are dependent on the 

 electi-ical state of the atmosphere is not saying much ; both for 

 other reasons, and especially because we do not know upon what 

 atmospheric electricity itself depends. But it is clear that it is 

 to a knjw ledge of the distribution of such electricity that we 

 mu t look for a proximate, as well as an approximate, explanation 

 of the facts. 



In a thunder-cloud we have an aggregation of aqueous particles 

 small enough to remain, temporarily at least, suspended in the 

 air. AH of these, it would appear, are similarly electrified, and 

 by their mutual repuUion are restrained from fun her coalescence. 

 By their presence the ground below the cloud becomes induc- 

 tively electrified in the opposite sense ; and as soon as the cloud 

 by its motion comes within sparking distance, or by an increase 

 of its charge attains sufficient tension, a spark discharge takes 

 place, which, as we have seen above, is a flash of lightning. A 

 similar action may naturally take place between two clouds 

 oppositely electrified. The electrical tension required for a 

 flash of lightning is of course enormous. It has been calculated 

 that in order to produce directly from a battery of the mo.>t 

 favourable construction a spark of 42 inches, equal to that given 

 from my great induction-coil, from 60,000 to 100,000 cells would 

 be necessary ; while for a flash of lightning a mile long not less 

 than 3,500,000 cells would be required. 



In some interesting experiments on water flowing from a small 

 orifice in a cistern Lord Rayleigh has found that the breaking of 

 the continuous column into drops is checked by communicatmg 

 to it a small charge of electricity ; but that it is promoted by a 

 large charge. We may imagine with him that something of the 

 same kind takes place in tbe cloud ; and th t before the flash the 

 aqueous particles are kept apart by mutual repulsion due to their 

 being all highly charged with electricity of the .same name ; but 

 that after the flash they are left either whhout charge or with so 

 slight a charge as to promote their coalescence and their 

 consequent fall in the form of rain. This would be an explana- 

 tion of the well known downpour which frequently occurs after 

 a flash of lightning. 



There is morefver another form of lightning to which the dis- 

 charge in our vacuum-tubes offers, to say no more of it, consider- 

 able analogy, namely, that commonly known as ball lightninj^. 

 The appearance of ball lightning is described as that of a 

 luminosity or ball of fire moving generally towards the earth, in 

 a direction more or less oblique, and disappearing in most cases 

 before reaching the gr.und. In some tubes, the exhaustion of 

 which is very moderate, say, having a pressure of several milli- 

 metres of mercury, it happens not only that the blocks of light 

 termed entities by Mr. De La Rue are formed, but also thst 

 these entities travel along the tube from the immediate neigh- 

 bourhood of the positive terminal to a finite distance in the 

 direction of the ni-galive, and then disappear. It would seem 

 not unreasonable to sup, ose that ball lightning is due to condi- 

 tions not dissimiLir to those of such tubes, mamely, to a discharge 

 occurring in the upper regions of the air, at an elevation of per- 

 haps twenty miles, more or less, where the pressure is moderate, 

 that is to .^ay, greater than that under which an auroral-like dis- 

 play could take place, and yet less than that which would give 

 rise to a true spark or ordinary flash of lightning. And if 

 further we effect the discharge in the tube by the gradual out- 

 pouring of electricity from a charged Leyden battery, or other 

 conden-er, through a suitable resistance, or if we use an induc- 

 tion-coil, then the condenser, or coil, will represent the charged 

 cloud, or portion of the atmosphere, from which the ))henomenon 

 proceeds ; and the analogy will perhaps be considered sufticien'ly 

 close to render farther observations in proof or disproof of the 

 theory desirable. 



Let us now turn to the aurora. Sufficient experiments have 

 been made this evening to show that the dischai'ge in rarefied 

 gases differs from that in gases at higher 1 ressures ; and that the 

 difference corresponds exactly to that observed between tbe 

 diffused, gen'le, and flickering play of the aurora and the sudden 

 crashing spark of a flash of lightning. It is also abundantly 

 clear that at an elevation of twenty or thirty miles above the 



