29>3 



NA TURE 



{July 26, 1883 



box if desired, or on the other hand their projecting ends can be 

 bent down and immersed in a freezing mixture when a cold dark 

 plane is wanted. 



The transition from the cold down-current to the warm up- 

 current is a thing I specially wished to observe, and it can be 

 readily seen by first letting the rid get thoroughly cold in the 

 dark and then turning on the light without removing the freezing 

 mixture. The down-current is now visible, and it persists for a 

 short time, varying from a second or two to a minute ; but as 

 the rod is warmed by the b^ani, it soon visibly slackens, turns 

 round, and establshes it-elf as an up-current; the transition 

 from a strong down- to a strong up-current only occupying a 

 few seconds altogether. If the light be now interrupted for a 

 short time, and then renewed, the down-current will be observed 

 as before, and in fact one may make the alternations with great 

 rapidity, permitting now the freezing mixture and now the hot 

 beam to gain the mastery. 



The turning b idily round of the dark plane is doubtless due 

 to a general convection current produced by the warming of the 

 glass of the box by the entering beam. The beam was, how- 

 ever, always filtered through wa'er in order to bring its heating 

 powers within manageable limits. 



To witness the effect of a diminution of pres ure on the pheno- 

 mena, a thick platinum rod with its end beaten into a narrow 

 spade, was sealed into an old lamp chimney closed by plane 

 g'ass ends, and connected with a water air-pump on one side, and 

 with a CaCl 2 tube, a tobacco pipe, and pinchcock on the other. 

 A little exhaustion and an intermittent opening of the pinchcock 

 was able to smoke the pipe in the orthodox mannner, and the 

 exhaustion was then proceeded w ith further ; an accumulation 

 of vacuum being often quickly turned on. At low pressure the 

 du t-free space surrounding the spade became large and ill- 

 defined, and the convection-currents were lazy and ineffective ; 

 the exhaustion was not pu-hed to extremes, because it was 

 difficult to keep any smoke still suspended, but the general fact 

 that the dark region broadened considerably under diminished 

 pressure was fairly well made out. The coat is enormously 

 thicker, however, than any Crojkesian or free-path layer. 



When examining the effects of electrification we sometimes 

 brought electrified rods near to the dark plane, and sometimes 

 we electrified the rod from which the plane was streaming. The 

 latter is by for the most effective, and the results are very striking 

 and interesting. It is not sensitive to minute differences of 

 potential however, and it required from fifty to one hundred 

 Leclanche's to exhibit distinct effects. We then found that posi- 

 tive electrification of the rod rendered the coat and the stream 

 broader, but made their outline hazy. Connecting the rod to 

 earth instantly sharpened it up again, making it beautifully clear 

 and distinct. Negative electrification sharpened the outline still 

 more, and narrowed it down still further, but the effect of 

 positive electrification was m ire marked than that of negative. 



When comparatively high potentials were used, such, for 

 instance, as would give millimetre s, arks if permitted, the effects 

 were violent. As the potential rose, the dark coat and stream 

 broadened, and ultimately disappeared ; reappearing again and 

 closing in from each side in a curious w ay, so as to reestablish the 

 clear dark plane depicted by Lord Rayleigh in his paper 

 above-mentioned, the instant an earth contact was made. 

 Violent negative electrification exhibits somewhat similar effects. 

 If any brush discharge took place, there was a violent black 

 chimney-like rush, and the whole box rapidly cleared of smoke. 



The electrical effects are not easy to describe, but they are 

 worth seeing. We sometimes used two solid brass cylinders 

 with rounded edges screwed on to brass rods and insulated from 

 each other, cylinders say three centims. by one ; sometimes we 

 used a cylinder and a point, sometimes two flat spades, and so 

 on. Connecting them with the poles of a Voss-Holtz machine 

 and turning very slowly, the change from two well defined and 

 sharp Lord Rayleigh planes through an interval of indistinctness 

 to vigorous and curiously-shaped black streamers is very striking. 

 Bu' in a few seconds all the smoke has gone ; it has not been 

 driven out of the box, it has been condensed on the box surfaces 

 and on the electrodes, which latter soon look as if they bad been 

 lacquered by an amateur, and make yellow greasy marks on 

 one s fingers. 



Moderate positive electrification of the rod, then, widens and 

 renders hazy the coat and the dark streaii ; earth connection, or 

 still better weak negative electrification, narrows it and renders 

 its outline beautifully sharp and distinct. The stream itself does 

 not show signs of electrification. Obstacles in its path deflect it, 



and it curls round them, forming rather a pleading stream-line 

 illustration. 



As soon as we had made out that the dark plane was con- 

 tinuous with a dark coat surrounding the bod)', we paid more 

 attention to the coat than to the plane. It s ems to me a some- 

 what important fact that solids have surrounding them a layer 

 into which dust particles do not enter, of a thickness which we 

 estimate as comparable with i/ioo:h of an inch, though it cer- 

 lainlv varies with tenperaiure, pressure, and electrical potential. 

 We first observed the dark C"at as a lining to a semi-cylindrical 

 scrap of copper sheet held in the binding screws formerly spoken 

 of, with the hollow turned towards the light. It can be seen 

 quite well, however, on a simple round rod or straight thick wire ; 

 and lor many reasons this is preferable. To avoid the shadow 

 of the rod and to see the coat all round it we return the light on 

 its path by a mirror, often also illuminating from above by means 

 of a 45 mirror. 



When the smoke is thick a feeble light is sufficient, but I prefer 

 a thin smoke and a powerful light. After tobacco has been in 

 the box some time, say half an hour, the smoke particles have 

 aggregated together and can be individually seen. It is then 

 very instructive to look al ng the end of the rod through a low- 

 power microscope. The diagram attempts to illustrate the 

 appearance. 



': / 



II 

 '/fl 



..'/ 



/ 



;,l 1 » ,\ 



, '/'mil ' 1 ' M 

 \\\\\w>/ 1 



iilki. ■ 



-J — uUICK 



MOVING 



The coat of dust-free air is perpetually being nibbed off and 

 renewed ; the attachment of it to the rod is not individual. I 

 believe all dark stria? seen in a smoky sunbeam are the wiped-ofl 

 coats of solid bodies, which, however, are now rapidly disap- 

 pearing by reason of the general diffu-ivene-sof the dust particles. 



The transparent coat on the in-ide of a glass tube full if 

 smoke can be seen, and when a point is heated the coat thickens 

 and rises, making a clear dark space, and then it proceeds to 

 roll itself up along with some of the dust into two distinct spirals 

 one on each side of the hot place. 



I have no h>pothesis whatever ready to account for the du t- 

 freeness of the film of air in contact with solids. But I believe 

 the existence of this film, and its electrical modification, to have 

 a close connection with various phenomena; for instance, the 

 easier discharge from negatively electrified bodies than from 

 positive — the dust-free coat is thinner : the coi.vective discharge 

 of electricity by hot bodies and its dissymmetry as observed by 

 Guthrie, the dissymmetry of the Lichtenberg tracings, the 

 abnormal dielectric strength of thin films of air as observed by 

 Sir Wm. Thomson ("Reprint," chapter xix.). For I imagine that 

 disruptive discharge would more easily commence in dusty air 

 than in clear air, and consequently that when the sparking bodies 

 are approached so that their dust free coats touch, the dielectric 

 strength is likely to be great. 



