Note on the explanation of a so-called "intertraction" phenomenon

Sir Almroth Wright observed that a disc of filter paper soaked in coloured albumen solution, fixed to the under side of a cover-glass which was floated upon a salt solution, showed coloured streamers radiating out along the surface. In seeking an explanation, I concluded that the streamers were due to the upper layer of the salt becoming lighter, owing to loss of salt to the filter paper by diffusion, and that this alteration of specific gravity caused currents outwards towards the surface of the solution beyond the cover-glass, which is slightly higher than the under side of the coverglass, and the only region to which upward streaming currents could go. In confirmation of this view, I showed that if the cover-glass was tilted the streamers (which I assumed to indicate the direction of the currents) only went to the upper side of the cover-glass, and if the positions of salt and albumen were reversed, so that the upper layers of solution would become more dense by diffusion, the streamers indicated downward currents. Sir A. Wright has criticised my explanation, and the criticism which seems to have most weight is that the upward currents would be expected to produce a uniform spreading cloud of colour, not a system of radial streamers. I have therefore tested directly whether a slow upward current of water, impinging on a disc of filter paper soaked in dye solution, drives out the colour in a uniform cloud or in streamers. The photograph shows that fine streamers are actually produced. The disc was supported in the surface of water, about 3 mm. above a tube 1 cm. in diameter delivering water at a rate of about 36 c. c. per minute. Variations of the rate of flow slightly altered the appearance of the streamers, but a uniform cloud was never produced.