Eight solar energetic proton events observed in the inner heliosphere on the spacecraft Helios 1 or 2 are studied. Particle pitch angle diffusion coefficients and mean free paths are derived from a power spectral analysis of the turbulence of the magnetic field during the time of the events. Quasi-linear theory of wave-particle interaction is applied strictly throughout in combination with the slab model of fluctuations. Several points receive special attention: (1) the large difference in scattering conditions from one event to the other, which requires an individual analysis of each event; (2) the selection of the exact stretch of magnetic field data to be analyzed and the influence of this choice on the result; (3) the exclusion of data periods that contain discontinuities and other nonoscillatory structures; and (4) short-scale variations of the overall field fluctuation level, which can be considerable.

As a consequence, the mean free paths of the particles are derived from time series of the varying momentary power density spectra as an appropriate average of the scattering conditions (as an average mean free path). These results from the magnetic field spectra are then compared with results of fits of propagation models to the particle time-intensity and time-anisotropy profiles observed at the same time. Of eight such comparisons, half of the cases exhibit for the first time an approximate agreement between the results of the two approaches, particle analysis and magnetic field analysis. The others, however, still show the discrepancy that is cited in the literature. Among the events with good agreement are the two extreme cases in the whole Helios data set, the event of April 11, 1978, Helios 2, with a mean free path from quasi-linear theory of 0.01 AU, and the events of June 7, 1980, Helios 1, with a mean free path from quasi-linear theory of about 0.8 AU at a rigidity of 0.1 GV. This shows not only that quasi-linear theory in a wave and slab model framework is indeed capable of producing correct results under certain conditions if an appropriate case-by-case analysis is conducted, but also that these cases can encompass a wide range of turbulent strengths. Several possibilities for the source of the discrepancy that remains for the other cases, which the theory and the assumptions applied in this work obviously do not explain, are discussed. However, at the present time no definite solution to the problem is obvious. ¿ American Geophysical Union 1993