68 



A'A TURE 



[May i6, 1895 



vlectrification of air by water falling in drops through it, even 

 though we now know that, if there is any such electrification, it 

 is not the main cause of the great negative electrification of air 

 which has Ijeen found in the ncighlx)urhix>d of waterfalls. For 

 this purjiosc an ex[>eriment has l)een very recently made by Mr. 

 -Maclean, Mr. (ialt, and myself, in the course of an investigation 

 regarding electrification and diselectrification of air with which 

 we have Ijeen occupieil for more than a year. The api>aratus 

 « hich we used is before you. It consists of a quadrant electro- 

 meter connected with an insulate<l electric filter * applie<l to test 

 the electrification of air drawn from different |xirts of a tinned 

 iron funnel, 187 centimetres long and 15 centimetres diameter, 

 fixed in a vertical jiosition with its lower end open and its upper 

 end closed, except a gl.ass nozzle, of f6 mm. aperture, admitting 

 a jet of Glasgow supply water (from Loch Katrine) shot vertically 

 down along its axis. The electric filter (R in the drawing), a 

 simplified and improved form of that described in the Proieedings 

 of the Royal Society for March 21, consists of twelve circles of 

 fine wire gauze rammed as close as po-ssible together in the middle 

 of a piece of block tin pipe of l cm. bore and 2 cm. length. 



One end of it is stuck into one end of a |x:rforalion through a block 

 of paraffin, K, which supiwrts it. The other end ((;') of this |)er- 

 foration is oinnected by block tin pii>e (which in the ap|>aratus 

 .ictually employed was 4 metres long, but might have licen 

 shftrtcr), an'i india-rulilK.T tubing through Ijellow.-, to one or otlier 

 of two short tiutlel pi|Ks (m and I') projecting from the large 

 funnel. 



§ 7. \Vc first applied the india-rublicr pi|)e to <lraw air from 

 the funnel .at the upper outlet, r, and m.ade many exiK'riments 

 to test the electricity given by it to the receiving filter, R, under 

 various cunrlilinns as to the water-jet : the liellows being worked 

 as unif'triidy as the rMK-'rator could. When the water fell fairly 

 ihrimgh the funnel with no droj>s striking it, and through 90 cm. 

 '<{ free air l)eiow its mouth, a small negative electrification of K 

 wai in every case otwervcd (which we th<«ight might pissibly Ik- 

 attributable lo electrification of the air where the water was 

 caught in a liasin aUiut 90 cm. l>elow the mouth of the funnel). 



• Kelvin, M.iclean, Gall, " On the Illselcctrification of Air." I'rpc. Roy. 

 .Soc, MArch tf, 189s. 



But when the funnel was slanted so that the whole shower of 

 drops from the jet, or even a small part of it, struck the inside of 

 the funnel, the negative electrification of R «as largely increasctl. 

 -So it w.as also when the shower, after falling freely down the 

 middle of the funnel, impinged on a metal jilate in metallic com- 

 munication with the funnel, held close under its mouth, or 10 or 

 20 cm. below it. For example, in a scries of experiments made 

 last .Monday (March 25), we found "28 of a volt in 15 minutes 

 with no obstruction to the shower : and 4"i8 volts in five minutes, 

 with a metal plate held three or four centimetres below the 

 mouth of the funnel ; the air being dra« n from the upper tmtlet 

 (!'). Immediately after, with r closed, ami air drawn from 

 the lower outlet (M), but all other circumstances the same, we 

 found '20 of a volt in five minutes with no obstruction ; ,^nd 

 67S volts in five minutes with the metal plate held below the 

 mouth as before. 



§ 8. These results, and others which we have found, with 

 many variations of detail, confirm, by direct test of air drawn 

 away from the neighbourhooti of the waterfall through a narrow 

 pipe to a distant electrometer, Lenard's conclusion that a pre- 

 ponderatingly strong negative electrification is given to the air 

 at every place of violent imixict of a drop against a water-surface, 

 or .against a wet st)lid. But they do not prove that there is 

 no electrification of air by drops of water falling through it. We 

 alw,ays found, in every trial, decisive proof of negative electrifica- 

 tion : though of ct)miMratively small amount when there was no 

 obstruction to the shower between the moulh of the funnel and 

 the catching basin 90 cm. belt>w it. We iiilenil to continue the 

 investigation, with the shower falling freely far enough down 

 from the mouth of the funnel to make quite sure that the air 

 which we draw off from any |Mrt of the fimncl is not sensibly 

 affected by im^vact of the drops on anything below. 



§9. The other discovery * of Lenard, of which I told you, is 

 that the negative electrification of air, in his experiments with 

 pure water, is <liminishecl greatly by very small quantities ot 

 conunon .s;ilt dissolved in it, that it is brought to nothing by "OI I 

 per cent. ; that |X)sitive electrification is produced in the air 

 when there is more than "Olt per cent, of salt in the water, 

 reaching a maximinn with about 5 per cent, of salt, when the 

 l>ositive electrical effect is about equal to the negative effect 

 observed with pure water, and falling to 14 per cent, of this 



amoimt when there is 25 per 

 cent, of salt in the solution. 

 Hence sea-water, containing 

 as it does about 3 per cent, 

 of common salt, may be ex- 

 pected to give almost as strong 

 positive electrificatiim lo air 

 as ]>ure water would give ot 

 negative in .similar circum- 

 stances as to comntt>tion, 

 l.enard infers that breaking 

 wa\es of the sea must give 

 positive electricity to the air 

 over them ; he finds, in fact, a 

 recorded observation by F.xner, 

 on the coast of Ceylon, slew- 

 ing the normal jiosilive electric 

 potential of the air to be not- 

 ably increased by a storm at 

 sea. I believe I,enard's dis- 

 covery fully expliiins also some very interesting observ,ations 

 of atmospheric electricity of my own. which I described 

 in a letter lo Dr. Joule, which he published in the 

 I'roiffdin,!^ of the Literary and Philosophical .Society ot 

 Manchester for October iK, 1859. + " The atmospheric effect 

 ranged from 30" to about 420" (of a heleroslalic torsion electro- 

 meter of ' the divided-ring ' species] during the four days which 

 I h.td lo test it ; that is to s;»y, the elect rometric fi)rce |x;r foot 

 of air, measured horizontally from the side 1 if the house, wnsfrom 

 9 to alMive 126 zinc-copper water cells. The weather was almost 

 |)erfectly settled, either calm, or with slight east wind, an<l in 

 general an ea.sterly liaze in the air. The eleclronieler twice 

 within half an hour went above 420', there being al the lime a 

 fresh lemi)orary breeze from the ea.st. What I h.td previously 

 observed regarding the effect of east wind was amply confirmed. 



• " Uelicr die Kleclricit.'il <lcr W.-Livjffiillc." Tabic xvii. p. 618. AnnaltH 

 lier Phytik uitd Chemie^ 1893, vol. xlvi. 



♦ Kcpuhllshed in " Eicclrmlalio and Magnetism." " Almosphcric Elec- 

 Iricily," xvi. \ ifyi. 



NO- ^lll' VOL. 52] 



