4 Lord Rayleigh on Reflexion from Liquid Surfaces 



depth of about an inch, and at the bottom of the dish was 

 placed a piece of darkly-coloured glass, not quite horizontal, 

 with the view of annulling the diffuse reflexion. The reflected 

 image was observed with a nicol, from which the glass covering 

 disks had been removed, simply held in the hand. The 

 appearance of the dark spot, brought to the centre of the 

 sun's disk, was at once recognized to be dependent upon the 

 condition of the water- surface. When the surface was clean, 

 the spot was dark and its width (in the plane of incidence) 

 was about ^ or ^ of the solar diameter. Moreover there was 

 a strong appearance of colour, brown above and blue below, 

 evidently due to the dependence of the polarizing angle upon 

 the wave-length of the light. But when the surface was 

 greasy, even although camphor fragments still rotated briskly, 

 the band lost its darkness, became diffuse, and showed but 

 little colour. When the greasy film was about sufficient to 

 stop the camphor movements, the contrast with the effect of 

 clean water was very marked*. 



The surface was cleansed by the aid of the expansible hoop 

 employed in my former experiments. This is made of very 

 thin sheet brass, about 2 inches wide. It is placed upon the 

 water, already reasonably clean, in its contracted condition, so 

 that the area enclosed is but small. When it is opened out, 

 say to a circle of about 10 inches diameter, the internal sur- 

 face of the water is rendered more clean, and the external less 

 clean, than before. To get the best result it is desirable to go 

 through the operation of expanding two or three times, pro- 

 bably because the cleaned water-surface acquires grease from 

 the internal surface of the brass hoop. It will be evident that 

 the action depends upon the hoop not being completely 

 wetted f. Otherwise the grease could repass from the out- 

 side back into the interior. For this reason the hoop cannot 

 be expected to succeed with liquids like alcohol. 



By taking advantage of the apparent motion of the sun in 

 altitude, these observations may be repeated in summer with- 

 out any heliostat, or reflexion, other than that of the water 

 itself. Thus on June 26 the dish was placed on a table below 

 the window of an un darkened room, and the passage of the 

 spot across the sun's disk was watched. The spot was 

 central at about 4 h ra , and the instant of centrality could be 

 determined to within 10 s , and probably to within 5 s . On 

 August 15, when the sun's motion in altitude was slower, 



* Phil. Mag. November 1890, p. 400. 



t Since imperfect wetting must be attributed to residual grease, it 

 would appear that the operation of the hoop is incomplete at best. Never- 

 theless, it is a very useful and convenient appliance. 



