Abstract
The proton-size correction to the hyperfine structure in the ground state of atomic hydrogen is re-examined. It is shown by means of dispersion relations that this correction can be expressed as an integral over experimentally measurable cross sections for electron-proton scattering. This clarifies the physical nature of the correction, puts it on a rigorous basis and lends support to previous analyses. In the absence of experimental data, we give some theoretical estimates for the correction. They agree with previous estimates, and therefore we cannot explain the present experimental value for the hyperfine splitting. We discuss some possible implications of this disagreement and suggest some experiments which would clarify the situation.