268 



NA TURE 



[Jan. 22, 1885 



Now this sudden conversion is just what might happen 

 in the case of the atmosphere, only the change of state 

 would be from vapour to liquid. Picture to yourselves 

 aqueous vapour accumulating and increasing in quantity 

 in dust-free air, saturated, over-saturated, nothing to start 

 the condensation ; it goes on accumulating ; the atmo- 

 sphere becomes unbearably damp, soaking into and 

 through everything. At length at some point something 

 causes it to give way, and condensation takes place. 

 Instantly it spreads from this point as from a centre, 

 volumes of liquid are produced, and fall not as a shower 

 but as a splash, deadly and destructive by the mere 

 weight and impetus of its fall. 



Instead of this, what really happens ? The moisture, 

 on becoming saturated, finds myriads of minute dust 

 particles or nuclei, round which it condenses ; the 

 more numerous the nuclei, the more minute may be the 

 globules of mist formed ; it never becomes supersaturated 

 at all. The instant it is saturated it begins to condense, 

 and we have the mist or visible cloud, and in this form it 

 may last for any length of time. Under certain influences, 

 however, not yet fully understood, but which I wish in 

 part to illustrate to-day, these minute globules may congre- 

 gate into larger ones. Too large to remain slowly falling 

 through the air, they begin to fall more quickly as their 

 size increases, and we get the fine shower ; or, if the 

 aggregation goes on further, and the drops do not 

 evaporate much as they fall, we have the heavy down- 

 pour, the thunderstorm, or the tropical deluge —all varieties 

 of rainfall caused by the different size of the aggregated 

 water globules. 



Were there no nuclei, condensation would not begin, 

 and were there but few nuclei, condensation could only 

 begin at a few points, and a quite different kind of mist 

 might present itself ; one which would consist of compara- 

 tively large and rapidly sinking globules — small for rain- 

 drops, but large for mist globules, a kind intermediate 

 between mist and rain, such a mist as is met with in 

 clear moist climates, and known in England as a Scotch 

 mist. 



Note this, that to get a fine permanent fog, you must 

 have an enormous number of centres of condensation. 

 Mr. Aitken {Trans. Roy. ioc. Edin., about 1879) estab- 

 lished this fact, that every spherule of mist must have 

 condensed itself round a minute solid dust particle, a 

 nucleus, and that without such nuclei condensation could 

 not go on. The minuteness of the nuclei able to act in 

 this way is extreme, an almost immeasurably small 

 quantity of matter being sufficient to precipitate a copious 

 cloud. Their size is quite beyond a microscope. 



[Mr. Aitken's experiment was here shown with appa- 

 ratus from the Royal Institution. A long glass tube is 

 filled with moist air, carefully filtered through cotton wool 

 and glycerine, after Tyndall, and is then suddenly ex- 

 hausted by an air-pump. It is thus cooled far below the 

 dew point, but no precipitation occurs ; and the tube, 

 well illuminated, is seen to remain clear. Now ignite 

 a platinum wire inside it with a few Grove cells, and 

 let more filtered air enter. As soon as this is done 

 exhaust again ; instantly a thick cloud is precipitated, 

 condensation occurring round myriads of nuclei given off 

 from the platinum wire -which, however, has not appre- 

 ciably lost weight. I wonder if this experiment could not 

 give Sir Will. Thomson a fifth limit to the size of atoms 

 by estimating the loss of weight of the platinum spiral 

 and the number of globules in the resulting mist.] 



A familar illustration of the effect of nuclei on vapour 

 is given by the simple experiment of writing on a pane of 

 glass \\1tl1 a stick, and then breathing on it. Where the 

 writing has wiped away the dust, the moisture condenses 

 less easily and in much fewer and larger globules than 

 where nuclei are abundant ; consequently the writing 

 becomes visible. 



In studying the properties of any physical agent, it is 



essential to be able to employ it or exclude it at pleasure. 

 One must have insulators to investigate electricity ; one 

 must perform optical experiments in a dark room ; and 

 to study the properties and functions of dust it is im- 

 portant to be able to remove it, and to obtain dust-free 

 spaces. 



Methods of removing dust from air are : — 



(1) Filtration through cotton-wool, or cotton-wool and 

 glycerine, packed tightly. Tyndall has shown how effec- 

 tive this can be made with proper management. 



(2) Allowing it to settle. In a few days or a week most 

 of the dust has settled out of stagnant air. Prof. Noel 

 Hartley employed atmospheres of hydrogen in his old 

 and careful experiments on " spontaneous generation," 

 because it was too rare for germs to float in. 



(3) Condensing vapour in the air several times. Mr. 

 Aitken has shown that successive condensations of vapour 

 gradually purify air by removal of nuclei, until it is quite 

 clear. He shows that the ability of vapour to condense 

 is an extremely delicate test of the presence of such nuclei, 

 and that when the dust particles are very few, conden- 

 sation takes place not as cloud but as fine rain or 

 Scotch mist. Doubtless, the cause of actual Scotch mist 

 is the clearness and purity of the Highland air induced 

 by frequent and continued rains. 



(4) Keeping a hot body in air for some time. This, 

 Tyndall calls " calcining " the air. 



(5) Discharging electricity into it from a point. 



I must say a few words about the two last methods. 

 When a hot body is held under a sunbeam, a dark stream 

 of dust-free air is seen rising above it. This was dis- 

 covered by Dr. Tyndall, and investigated by Lord 

 Rayleigh, as well as by Mr. Clark and myself. 1 A hot 

 spiral of platinum wire in a bell-jar produces this dust-free 

 stream, and so gradually clarifies the air in the jar. That 

 this is not due to combustion or evaporation we proved by 

 using the smoke of burnt magnesium, which answers per- 

 fectly. Lord Rayleigh has shown that a cold body is 

 similarly effective, and causes a desce}jdmg dust-free 

 stream. 



We have found that the dust-free streamer is only a pro- 

 longation of a dust-free coat which surrounds all warm 

 bodies. The dust is kept away from them by molecular 

 bombardment. It has been shown by Tait and Dewar, and 

 by Osborne Reynolds, that a Crookes bombardment is 

 effective at even ordinary pressures provided the bodies 

 bombarded are small. Dust particles are very small, and 

 so they get driven by molecular impact away from hot 

 surfaces and towards cold ones : the distance through 

 which they are so driven away being easily measured by 

 observing the thickness of the dust-free coat round an 

 illuminated body at known temperature." Two black tin 

 vessels or glass flasks can be put under a bell-jar, one of 

 the flasks full of warm water, the other of cold. On now 

 burning magnesium, or otherwise filling the jar with smoke, 

 the cold one will presently be found thickly covered with 

 a deposit, the warm one will be nearly free. What Tyndall 

 calls " calcining " the air, then, is really bombarding the 

 dust out of it on to the cool wall surfaces. The deposition 

 of lamp-black on a cold body held in a flame is thus ex- 

 plained. Whenever the air is warmer than bodies it 

 deposits its dust and smoke upon them ; whenever bodies 

 are warmer than the air they keep the dust off, except 

 when the weight of some of the larger particles is sufficient 

 to overcome the bombardment ; a thing which is very 

 likely to happen on a horizontal and slightly warm surface. 



1 Mr. Aitken commenced the same investigation after reading my pre- 

 liminary note of July 18S3 in Nature, and has followed it up in much the 

 same way as we have, obtaining very similar results. 1 have just seen Mr. 

 Aitken's paper in the Traus. Roy. Soc. Edin., vol. xxxii. Part II. He 

 therein criticises one or two of the views I somewhat hastily expressed in 

 the preliminary note referred to. But our views were naturally modified by 

 further experience, and in the complete paper in the Pkit. Mag., -March 18S4, 

 they are more carefully expressed. It would have been better if I had not 

 written to Nature until the investigation was complete. 



- See Lodge and Clark, Phil. Mag., March 1884 ; also Nature, July 26, 

 1883, vol. xxviii. p. 297, and April 24, 1884, vol. xxix. p. 612. 



