Si6 



NATURE 



[JUxNE 15, 191 1 



LETTERS TO THE EDfTOR. 



yrUe Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications.] 



Bieath Fi^uies 



On reading Lord Rayleigh's short article on breath 

 figures in Nature of May 25 and finding he was not 

 satisfied with the explanation 1 offered of them, I thought 

 it would be as well to see if it was possible to get some 

 further information on the subject. For this purpose I 

 repeated the experiments with variations in the conditions. 

 As the writer of the article seemed to think that the hot 

 gases act in some way in cleaning the surface of the plate, 

 the experiments were now made at lower temperatures. In 

 place of the blow-pipe flame a Bunsen burner was used, 

 and in order to protect part of the surface from the action 

 of the hot gases, the plate was supported on two thick iron 

 bars, so as to confine the action of the gases to the narrow 

 space between them, and further, the glass plate was kept 

 as cold as possible. It was intended to use ice for this 

 purpose, but it was found that it kept the under surface 

 always dewed, so that the gases could not come into con- 

 tact with the plate. Wet blotting-paper placed on the 

 upper surface was found to keep the plate cool enough, 

 and yet not so cold as to cause the deposition of dew. It 

 was found that it was not necessary to pass the plate 

 through the flame to get the breath figures ; it could be 

 held some distance above it with a similar result, but, as 

 might be expected, a longer time was necessary than when 

 in the flame, not only on account of the lower tempera- 

 ture of the gases, but also on account of the plate getting 

 heated by the longer exposure. 



In these experiments it was noticed that it is not enough 

 to allow the hot gases to travel along the space between 

 the two bars, as the hot gases only produce the effect 

 where they first strike the surface of the plate, and if it 

 is dust that is the cause of the action, it will be deposited 

 where the gases first touch the cold surface, the onward 

 flowing gases near the plate being nearly free from dust, 

 as well as being at a lower temperature. ' The plate should 

 therefore be moved over the stream of hot gases in such 

 a way that fresh gases come in contact with the different 

 parts of the exposed surface. The plate should also be 

 held at a considerable angle to allow of a free flow of the 

 gases over it. 



If the action of the hot gases is a cleansing one, then 

 heat alone ought to produce the same effect as the flame, 

 and cause the glass to take an even film of dew. To test 

 this a cleaned plate was placed— clean side up — on a metal 

 plate somewhat larger than itself, and the metal plate 

 Highly heated with a Bunsen burner. Some plates were 

 heated slightly, others to a temperature far higher than 

 those acted on directly by the flame, yet when cold these 

 plates were unchanged. Part of the plate was freshlv 

 cleaned, when it was found that the deposit of moisture 

 was the same all over the plate. Heat alone evidently has 

 no effect. I may here mention that those breath figures 

 are best developed by rubbing the back of the plate with 

 a piece of ice. By this process the image remains longer 

 in view, and gives time for inspection. 



An examination, by means of a short-focussed lens and 

 proper illumination, of the moisture condensed on the 

 ^"'■face of a glass plate, reveals some interesting points. 

 When the dew first begins to appear the water is seen to 

 be deposited in the form of very small lenses nearlv touch- 

 ing each other. On looking through the plate while the 

 dew is still condensing, nothing can be seen through it. 

 It acts very much like a piece of ordinary ground glass, a 

 landscape being invisible ; even the sky-line does not show. 

 If the cooling be now stopped and the deposit allowed to 

 evaporate it will be noticed that the spaces between the 

 lenses widen, the lenses drawing in their edges and leaving 

 clear glass. Objects can now be seen through the plate, 

 the transparency increasing as the dew evaporates. If, 

 however, in place of stopping the condensation at its early 

 stages we keep the plate cooled so as to cause more 

 moisture to be deposited, the lenses will be seen to grow 

 into each other and coalesce, until at last the regular 

 NO. 2172, VOL. 86] 



deposit of little lenses grows into irregularly sh.iii 

 blotches of water of considerable size. If the coml 

 be stopped at this stage and the plate dried m 

 cooled, it will be found that the appearance of i... . • 

 densation is not the same as it was the first time. 1 

 surface of the plate has been changed by the first conden^ 

 tion. On the dew making its appearance the second til 

 the uniformity in the lenses is gone. In addition to 

 general ground of small lenses there now appears 

 number of much larger ones, and as these are distribut 

 over the plate at about the same distance from each otY 

 as that occujiiid by the large evaporated blotches of wut^J 

 they would appear to be produced by something left 

 the water on the glass. By observing a particular grouf 

 of blotches near a mark on the plate, it was noted tha 

 each one of these reappeared on each successive dewingJ 

 The something which is left after evaporation has evidently^ 

 a considerable affinity for water, as these large lenset£i 

 condense more water than an equal area of small ones^^. 

 They stand high, and are the l»»t- to evaporate. 



But perhaps it will be asked, How can anything be !• ;• 

 after the evaporation of pure distilled water? At first s'r^'.r 

 one might feel inclined to say that nothing could be !• • 

 It is, however, evident that something has been kli 

 probably dust and gaseous impurities condensed along with 

 the water — and I may point out that something similar 

 happens when cloud particles of water are form«-d 

 on ions in the presence of certain gases. In this > 

 case there are no solid nuclei, yet the drops do not > 

 thoroughly evaporate, but leave behind them something » 

 large enough to be nuclei of condensation in air only * 

 very slightly above saturation. These obser^•ations on the ;j 

 behaviour of glass plates show how delicate their surfaces i 

 are and how easily the condensation on them may be 

 altered. 



The something which produces these breath figures, 

 formed by flames is of such a nature that it acts more] 

 powerfully the higher the temperature of the gases, that 

 is, the greater the difference in temperature is between 

 the gases and the glass. Of course, it is possible that th« 

 effect may be due to some gas or gases condensed on th 

 glass, and these gases having an affinity for water, 

 the other hand, it may be due to the gases depositir 

 their fine dust ; and while what we call clean glass repe£ 

 water, dust, on the contrary, attracts it. The fact th 

 this form of breath figure is easily washed off the plat 

 seems to point rather to dust than absorbed gases as 

 qause. 



The formation of these breath figures does not appea 

 to be so much a question of cleanliness as of the natii 

 of the foreign matter on the surface of the glass. If 

 impurity is of a water-repelling nature we get the lent 

 cular deposit ; if it has an affinity for water we 

 the uniform film. For instance, if the plate has 

 previously touched with paraffin oil, it will, evi 

 after a good deal of cleaning, give a very white obscu 

 deposit, due to the great convexity of the lenses. On 

 other hand, if the surface has been previously treated wit 

 caustic soda, the lenses are flat and the plate more trar 

 parent. 



As these experiments have generally been made wit 

 gas flames, and as the flame is an important factor, 

 was thought advisable to try other flames. Hydrogen whe 

 burned in filtered air, I have shown, gives rise to no fij 

 dust, and, so far as I know, this is the only form of cc 

 bustion which does not. It therefore ought to have 

 tested, but as the apparatus required would be somewl 

 complicated, it has not been done. Alcohol, however, 

 tried ; with it I could only succeed in getting very slij 

 indications of any action, though the plate was heated 

 more by it than what gave a marked effect with 

 Bunsen flame. This result is what we might expect if 

 figures are due to dust. A sulphur flame was also trikj 

 and, as might be expected, gave very marked results 

 a very small flame. This might be given as a typical 

 of the effect of condensed vapour. Something of this 

 may play a small part in the figures produced by 

 flames. 



It was thought it might be interesting to see how 

 fine dust on our windows acted towards condensed wate 

 The fine dust in the air is deposited on our windows in tli 



