The scattering of light by liquid boundaries and its relation to surface-tension. — Part I

The conception, introduced by Hardy and Langmuir, that a layer of orientated molecules forms the boundary of a liquid and determines its surface-tension, suggests some interesting questions for investigation regarding the optical properties of liquid surfaces. Should not the orientated molecular layer (assuming it to be real) be doubly refractive ? How is the configuration of the molecules at the boundary of a clean liquid surface influenced by the thermal agitation, and how is this related to the surface-energy of the fluid ? Is the surface-layer capable of producing (by reason of the thermal agitation or its optical anisotropy or other cause) the observed elliptic polarisation of the light reflected by liquid surfaces (clean or contaminated as the case may be) at the Brewsterian angle ? The present paper is the first of a series describing the results of work undertaken to find answers to the questions here raised. In various papers published previously, the internal scattering of light in transparent fluids has been discussed, and it has been shown that it is quantitatively connected with the spontaneous fluctuations in density arising from the thermal agitation of the molecules within the fluid. This internal scattering is a volume-effect and is quite distinct from another and very important type of light-scattering which may be expected, namely, that which occurs at the boundaries of reflecting and refracting media, and is a surface-effect , due to the agitation of the boundary. So far as we are aware, very little experimental work has been hitherto published on the subject of this surface-scattering by perfectly clean liquid surfaces. That a substance in the special circumstances of the critical state when it has a vanishingly small surface-energy may exhibit an observable surface opalescence in addition to the familiar body opalescence, was suggested by Smoluchowski in his paper of 1908, on the thermodynamics of the critical state. Nothing seems to have been done to follow up this suggestion till 1913, when Mandelstam published some observations on the special case of the light incident on the boundary between the two layers of a mixture of carbon di-sulphide and methyl alcohol near the critical solution temperature, at which the liquid develops a milky opalescence. Mandelstam noticed that in directions not greatly removed from that of regular reflection from the interface there was also some scattered light, and conjectured from his somewhat meagre and qualitative observations, that the effect was analogous to that predicted by Smoluchowski for the critical state of a single liquid. Early in 1923, the present authors took up the general problem of the light-scattering from optical boundaries and succeeded in observing the bluish opalescence of the clean surfaces of transparent and metallic liquids and discovered the special polarisation effects exhibited for large angles of scattering. A preliminary communication was published in ‘Nature,’ August 25, 1923. Since then, the subject has been extensively developed, no fewer than sixty liquids being studied, and quantitative observations made of the intensity and polarisation of the scattered light for the widest range of angles of incidence and observation and for different physical conditions of the fluid. The work has established a quantitative relationship between the surface-opalescence and the surface-tension of liquids. In Paper I of the series, the case of metallic liquids will be dealt with. In Paper II, transparent fluids will be considered. In Paper III, the effect of contamination on the surface-opalescence of water and the special phenomena of the critical state will be described. In Paper IV, the theory of the phenomena will be discussed. A recent (purely mathematical) paper, by Gans, which appeared about a year after our preliminary announcement was published, may be mentioned in this connection.