272 



NA TURE 



\Jan. 22, \\ 



SOME EXPERIMENTS ON FLAME 



IN December 1SS1 my attention was casually called to 

 the popular superstition that sunlight puts the fire out. 

 Returning from a walk I had found the blinds of my 

 sitting-room closely drawn, for the benefit, as I was told, 

 of the fire, which was low. On my appearing somewhat 

 sceptical about the use of this proceeding, my landlady 

 cited the above-mentioned superstition as a well-known 

 fact. For her benefit and instruction I made the poker 

 red hot, and focused the sun's rays on it with a bull's-eye, 

 showing her that, though the bright light prevented the 

 red heat from being seen, it had not extinguished it, and 

 was, moreover, capable of making a smaller piece of 

 metal red hot. But 1 was myself so struck with the 

 power of even the December sun in overcoming the light of 

 the most highly incandescent body, that I determined to 

 make further experiments. Even the intense glow pro- 

 duced by heating in the blowpipe flame a small piece of 

 chalk, though it was sufficient to light up the whole 

 room, entirely disappeared in the sun's rays. This led 

 me to ask what would be the result of testing the sun's 

 light in the same way against that of a flame. If, accord- 

 ing to the older theory, luminous flame consists of incan- 

 descent solid particles, then I should expect that these 

 would behave under the strong light exactly as the white- 

 hot iron did, while, on the other hand, if as some have 

 maintained the white light of a flame proceeds from 

 gases of great vapour-density, then I might expect 

 results which, if not different, would be at least 

 interesting. 



Experiment I. — Accordingly, on December 7, 18S1, I 

 arranged my large condenser — a lens 5 inches in dia- 

 meter, and 20 inches focus — so as to throw the image 

 of the sun upon the flame of a paraffin candle. To my 

 delight a round spot of light of a bluish-white colour 

 and peculiar soft appearance was visible on the flame 

 itself. That the flame, whether gaseous or consisting of 

 incandescent particles, could reflect light, was certain. 

 It remained for me to determine the characteristics of 

 this reflection. From its colour and peculiarly "soft" 

 appearance it reminded me of fluorescence. I therefore 

 proceeded to test the question with the spectroscope. 



Experiment 2. — I examined first the spectra given 

 when a beaker of petroleum or one of solution of quinine 

 sulphate was placed in the focus. I should mention that 

 my spectroscope, which I designed and made myself, 

 slides up and down the supporting pillar, so that it can 

 be adjusted to any height. The table carrying the slit, 

 and telescope, and prism (dense flint of 6o°), can be fixed 

 in three positions to the stand, so that the slit may be 

 vertical, horizontal, or directed vertically downwards for 

 examining solutions with the light thrown up from be- 

 neath. It is also provided with a doublet, equivalent to 

 the B eye-piece of a microscope, used as a condenser to 

 throw the image, which maybe an enlarged or diminished 

 one at pleasure, of any object upon the slit. The whole 

 arrangement is very simple, and far more convenient than 

 that of the ordinary laboratory spectroscope. Bringing 

 the instrument thus armed to bear upon the strongly 

 illuminated solution, I found the field of view to be filled 

 with a soft and even light, that seemed to obscure the 

 Frauenhofer lines as if some thickened luminous solution 

 had been poured over them. Fvery moment some par- 

 ticle of dust floating into the focus would cause a tiny 

 flash as its image crossed the slit, of hard clear light, like 

 that of the candle-flame, only that it showed the Frauen- 

 hofer lines. But after filtering the solution, carefully 

 cleaning the beaker, and excluding all extraneous light, 

 the Frauenhofer lines vanished, ami nothing was visible 

 either with quinine or petroleum but the soft continuous 

 spectrum of fluorescence. I have described these well- 

 known phenomena thus minutely that I may emphasise 

 the very different results obtained in the following experi- 



ment. To the naked eye the spot of sunlight upon the 

 candle-flame was of exactly the same soft quality, and 

 nearly the same colour as that upon the fluorescent solu- 

 tion. I replaced the candle in the focus, arranged the 

 condenser of the spectroscope so that the white spot 

 should come upon the centre of the slit, and occupy one- 

 third of it. The field of view was filled by the spectrum 

 of the flame, but across the centre was a bright band of 

 light extending far into the violet, brightest in the blue, 

 and showing nil t lie Frauenhofer lines distinctly, especially 

 in the blue and violet. Unmistakably I was dealing with 

 reflected light, and not with fluorescence. My thoughts at 

 once reverted to Prof. Tyndall's " blue cloud." I knew 

 nf two ways of producing an extremely fine precipitate 

 showing the same characteristic phenomena. I added 

 dilute hydrochloric acid to a weak solution of sodium 

 hyposulphite, but this preparation I found to be trouble- 

 some from the rapidity with which it loses its optical 

 properties, so I discarded it in favour of the following. I 

 diluted some French polish with about fifty times its bulk 

 of methylated spirit, and added a few drops of the solution 



Fig. i.— a, tumbler con 



polarising apparatus ; c . 

 cardboard to screen off 

 light ; g, mirror ; H, side 

 different azimuth may be 



3 support 

 , sheet of 



uperfluous light ;" F, lens to concentrate^ the 

 lirror in which the colour of the beam in a 



to a glass of water. The precipitate of lac resulting is 

 sufficiently fine for every purpose, and will remain in 

 suspension for days. The light from the heliostat passing 

 through this solution gives the same soft opalescent reflec- 

 tion, with the same spectrum strongest in the blue and 

 violet, showing all the Frauenhofer lines distinctly, as it 

 does upon the candle-flame. 



Experiment 3. — There is another special characteristic 

 of matter in extremely fine division common to Prof. 

 Tyndall's " blue cloud " and the above-mentioned solu- 

 tions. Light reflected from it is completely polarised in 

 the plane at right angles to the line of incidence. I am 

 in the habit of showing this by the following arrangement, 

 which I believe to be new, and which is so simple that 

 any one can exhibit it. It is shown in Fig. I. A is 

 an ordinary plain tumbler, half filled with "lac preci- 

 pitate," and covered with a piece of window-glass, B. On 

 b is laid amounted selenite film, C, and upon this again the 

 polarising prism l>, used with the microscope. A retort- 

 stand supports a sheet of cardboard, E, with a hole in the 

 centre, which shades the liquid from superfluous light. 



