Liglit hy ^rerruiy and its ]\tj}Oitr. 77 



sqiieezino- them between iron flanoes. Near one o£ the 

 windows, a short side tube of iron was screwed in, at right 

 singles to the main tube. This side tube wis prolonged in a 

 thin glass one. Under the main tube were a row of bunsen- 

 burners, and bricks were arranged to keep in the heat from 

 them. A small quantity of mercury was placed in the tube, 

 which latter was kept slightly sloped^ so that the mercury 

 ran down to the end furthest from the side tube. When the 

 tube was heated sufficiently the mercury vapour drove out the 

 nir, and rose into the vertically placed side tube. It was 

 there condensed, and fell back into the hot tube. Thus the 

 space between the mica windows was filled with mercury 

 vapour. The windows were kept hot to prevent condensation 

 of mercury upon them. In order to detect the absorption of 

 light by the column of vapour, the light from a paraffin lamp, 

 diffused bv means of oround oiass, could be observed throuorh 



■I " O ~ / ^ o 



the tube. A comparison lio'ht was arranged by the side of 

 the tube, the illumination being derived from the same lamp, 

 by a reflector. The light which had come through the 

 mercury vapour could be observed simultaneously with the 

 light which had come from the same lamp through the air. 



When the tube was cold the two lights were adjusted to 

 equal brightness, by suitably inclining the reflectors. The 

 tube was next heated so as to fill it with mercury vapour. 

 It was then found that the two lights were, as before, equal 

 in brightness, so that no absorption by 3 metres of saturated 

 mercury vapour, at the boiling-point of mercury, could be de- 

 tected. We may consider the experiment accurate to, perhaps, 

 10 per cent. It may be concluded that the coefficient of 

 absorption for the light of a paraffin lamp cannot exceed '0003. 



To determine the absorption by mercury in the solid 

 form, a very thin film must of course be used. In order to 

 get this, I have made use of Prof. Dewar's observation that 

 mercury can be deposited on glass as a fihii by condensation 

 from the vapour, using liquid air as a cooling agent *. A 

 glass bulb of 630 c. c. capacity contained mercury vapour, 

 saturated at 20° C. Inside this, and concentric with it, was 

 a very small bulb, into which liquid air could be poured. 

 When this was done the mercury vapour condensed as a 

 transparent film on the outer surface of the internal bulb. 

 This film had an area of 11 sq. cms. 



The vapour-tension of mercury at 20° C. is '0013 mm. If 

 we compute the density by assuming the simple law of gases 

 to hold good for the saturated vapour (and probably this is 



* I must express my best thanks to Prof. Dewar for giving me the 

 opportimity of caiTying out this experiment. 



