The Fortran program package GNOM carries out SAS data treatment and interpretation for the disperse non‐interacting dilute systems of different types. A general operator method based on Tikhonov's regularization technique is used. Here a finite real‐space distribution function [the correlation function p(r) for monodisperse systems and size distribution function D(R) for polydisperse ones] is searched for to fit the experimental data in accordance with the experimental conditions and the error bounds. It is possible to require non‐negativity of the solution. The package GNOM is applicable for arbitrary experimental conditions (arbitrary weighting functions and form factors) and provides solutions stable to statistical noise and termination effects. The program makes it possible to treat together experimental curves from the same object measured under different experimental conditions and to choose the best model from the given set. Interactive and batch versions are available. The program is adapted to NORD, VAX and IBM‐PC computers. It possesses convenient graphics possibilities.

The application of Tikhonov's regularization method [Tikhonov & Arsenin (1977). Solution of Ill-Posed Problems. New York: Wiley] for the solution of illposed problems in small-angle-scattering-data treatment is considered. Simple regularization algorithms are proposed for solving convolution equations in data desmearing (slit-width and polychromaticity problems) as well as for polydispersity problems. A general indirect approach of data processing based on the regularization method is described. Comparison with other data-treatment methods is made.

Different types of nuclear filters were investigated by means of small-angle X-ray scattering. The distribution functions of the cross-sectional squares were calculated. The obtained results are in good agreement with the data of other physical-chemical methods. Small-angle scattering is found to be effective for structural investigations of nuclear filters.