In our previous paper [1], we reported on some x-ray scattering experiments from black foam films. In this paper the scattering curves obtained were interpreted by simply comparing Fourier transforms of one-dimensional electron density models with the measured data. Total reflection was not taken into account. The film thicknesses obtained by this method for the common black films (CBF) were somewhat higher than the values obtained by optical methods [2, 3]. The thicknesses obtained for the Newton black films (NBF) were unreasonably high. Meanwhile, similar experiments with the same substances have been reported [4]. These authors have used the optical matrix formalism, valid at all angles [5, 6], for computing and fitting the reflectivity profiles. The film thicknesses obtained by this method for CBF and NBF, however, are rather small, smaller than the values derived from optical measurements [2, 3]. Stimulated by this work, we tried a new interpretation of our previous experimental data. We applied a program, which had originally been developed at the Hahn-Meitner Institute for investigating thin films deposited on solids and which is also based on the optical matrix formalism [7]. Presently, only three-layer-models can be treated by the program. Varying the thickness and the electron density, a series of models for both, CBF and NBF, were calculated and compared with the experimental data from [1]. The best correspondence between the experimental data for CBF, taken from Fig. 2 in [1], has been obtained with the following model: two hydrocarbon layers of thickness 12.7 angstrom and electron density 0.27 e/angstrom3 with one water layer of thickness 39.6 angstrom and electron density 0.36 e/angstrom3 in between. It was assumed that all three layers have a roughness due to thermal motion of 2.5 angstrom (defined as standard deviation of a Gaussian distribution). Thus, the total film thickness for this model is 65 angstrom. The results of the calculation based on this model are presented by solid lines in Fig. 1. The experimental data from [1] are shown as open circles. Since the intensity values are known on a relative scale only, the data were multiplied by an arbitrary factor.
