Jan. 22. 1 885 J 



NA TURE 



2 7Z 



a lens, f, which may be an ordinary eye- 

 glass laid across the hole, and so adjusted that its focus 

 shall come about the middle of the liquid. A plane 

 mirror, G— a hand-glass will do — is then either held or 

 fixed, so as to reflect sunlight perpendicularly upon the 

 lens. It will readily be seen that the light, concentrated 

 by the lens, is plane-polarised by the Nicol prism, then 

 modified by the selenite, and finally analysed by reflection 

 from the extremely minute particles of lac. Accordingly, 

 to a person walking round the table with his eye on a 

 level with the tumbler, the vertical beam of light in the 

 liquid appears to change colour four times. Thus, if the 

 selenite and Nicol are so adjusted that viewed from the 

 west it appears blue, then from the south it will be yellow, 

 from the east blue, and from the north yellow again, if 

 then the selenite be removed from under the Nicol, from 

 both west and east it will be seen as a bluish-white beam 

 of light, while from the north and south it will be invisible 

 altogether, as though a screen had been placed over the 

 lens. By arranging or holding a small mirror, H, at an 

 angle of 45 , by the side of the tumbler, the observer may 

 see the blue colour of the beam from the west side, on 

 which he stands, while at the same time the mirror shows 

 him that its colour, when viewed from the north or south, 

 is yellow. Or three mirrors may be arranged so that all 

 four aspects of the beam may be observed at once. I do 

 not know a more beautiful and striking way of demon- 

 strating the properties of the polarised ray. 



Experiment 4. — I now come to the most interesting of 

 my experiments. This polarisation of all light reflected 

 at right angles to the line of incidence is, I believe, 

 accepted as the special characteristic of very finely-divided 

 solid matter. I applied the test to the light upon the 

 candle-flame. I held the Nicol in the plane at right 

 angles to the mean path of the rays, looked through it at 

 the soft spot of reflected sunlight, and rotated it. When 

 the crystal crossed the line of incidence at right angles, 

 the spot vanished ; when it coincided with it, the spot was 

 brightest. With a selenite film in addition to the Nicol 

 prism the usual change of colour could be seen, the red 

 and green film showing more distinctly than the blue and 

 yellow. By using the Nicol over the eye-piece of the 

 spectroscope I found that every part of the spectrum of 

 the reflected sunlight is polarised alike, showing that the 

 flame behaves with respect to light exactly as a solution 

 containing extremely fine solid particles. I made a large 

 number of experiments with a view to ascertain how far 

 this similarity would hold, and I now proceed to give 

 some of the most important. 



Experiment 5.— I arranged the heliostat with the 

 candle-flame in the focus and the spectroscope at right 

 angles to the line of incidence, with the Nicol prism over 

 the eye-piece, and the condenser arranged to focus the 

 "white spot" of sunlight on the slit. I then gradually 

 lowered the candle so as to bring the apex of the flame 

 into the light. There was no break in the appearance of 

 the spectrum on passing from the hot flame to the non- 

 luminous smoke. Low down, the flame reflected only 

 the more refrangible rays, as far as the middle of the 

 green ; towards the apex it reflected also the red. All 

 the reflected light was polarised. 



Experiment 6. — With the same arrangement as before, 

 I turned the spectroscope so as to have the slit horizontal. 

 I burnt some soda in the Bunsen burner at a little dis- 

 tance, so that the vapour from it came to the candle. 

 The result is depicted in Fig. 2. The continuous spec- 

 trum of the inner flame is crossed by the bright sodium 

 chprojeel a little distance beyond it on either 

 side to the limits of the outer flame. In the centre is a 

 iand, the spectrum of the sunlight on the flame, 

 and on this all the Frauenhofer lines, includin 

 lines perfectly black, as in my drawing. It was very 

 curious to see the two ends of the sodium lines standing 

 out bright against the dark background on either side, 



visible still as bright lines, though faintly, upon the flame 

 itself, up to the band of sunlight, and then strongly 

 reversed by contrast with its greater brilliancy. I believe 

 I am the first who has succeeded in reversing the sodium 

 lines by reflection. It requires a bright sun to do this ; 

 otherwise the red end of the spectrum is not strong 

 enough, but I have succeeded in showing it to several 

 friends. 



Expo intent 7. — With the same arrangement, substi- 

 tuting a spirit lamp charged with soda for the candle, 

 nothing was visible to the naked eye ; the flame seemed 



1 of =andle-flame in [he focus of the h=!i< 

 D lines reversed by reflection. 



to vanish in the glare ; only in the spectroscope the 

 bright lines were seen unaltered. With the Bunsen a 

 brightly illuminated column of dust was seen rushing out 

 of the tube, each particle vanishing as it reached the 

 perfectly invisible flame, and was burnt. Several sub- 

 stances, e.g. copper oxide, and ammonium molybdate, 

 give in the outer flame a spectrum which in my small in- 

 strument appears continuous, though lacking the " hard " 

 look of the spectrum of an incandescent solid. But they 

 give no reflection with the strongest sunlight, behaving as 

 true vapours. It will be observed that, though I have 



