340 Dr. J. H. Vincent on 



4 metres square to get enough waves to illustrate optical 

 phenomena. The quantity of mercury necessary to use this 

 method is a great objection, but there is another. A method 

 of illuminating the surface by a beam of practically parallel 

 light would have to be devised. With waves of this length, 

 however, an ordinary short-exposure shutter could be used. 



Very short waves refuse to pass over a thin film of the 

 liquid lying on an amalgamated surface *. The use of 

 mercury in retraction experiments is then not very hopeful. 



There is, however, a medium course which might yielj good 

 results. This is to use an amalgamated surface for the 

 shallow regions, but to employ waves say 2 centim. long. To 

 carry this out it would be necessary to find out the smallest 

 depth of mercury over an amalgamated surface which would 

 propagate the waves ; and it is not improbable that this 

 depth would be small enough to allow of the experiments 

 being carried out in a trough of attainable size. 



The method actually used in obtaining the photographs 

 illustrating this paper is as follows. The liquid employed was 

 water, and the trough to hold it was made of a sheet of plate- 

 glass with a wooden rim. This trough was suspended by 

 being placed on a wooden board which formed the lower part 

 of a frame. The sides and top of the frame were of iron, and 

 the whole was suspended, as usual, by a rubber cord. A hole 

 cut in the board lets the light through from below upwards. 

 The light passes through the glass plate and the liquid. The 

 optical arrangements were as follows : — The light from the 

 spark-gap traverses a lens, and is thus rendered parallel, in 

 which condition it passes through the trough and its contents, 

 striking a second lens which is placed at a calculated distance 

 from the front lens of the camera-combination so as to bring 

 the light to a focus on the aperture in the lens-stop. The 

 camera is focussed on the surface of the water, and as in the 

 apparatus previously used, if the optical arrangements are 

 correct and the source and aperture small, no light can reach 

 the sensitive plate except that passing without deviation due 

 to the ripples on the liquid surface. 



Before every experiment attempted it is necessary to so 

 adjust matters that, in spite of any prismatic deviation due to 

 the glass plate not being horizontal, the image of the spark 

 appears at the aperture of the stop. This adjustment could 

 not be made once for all, because the balance of the hanging 

 system was disturbed by the introduction into the trough of 

 devices for producing shallow portions of water. 



The water had to be repeatedly changed, as the slightest 

 * See p. 196, second paper, Phil. Mag. Feb. 1898. 



