Abstract
The author's variational method of calculating g-factor shifts for atomic hydrogen trapped in a rare-gas solid is applied to the case of the alkali metals as impurities. An alternate derivation of the results is given and a simple physical model is discussed. The calculated shifts are compared with the experimentally observed ones for H, Li, Na, and K trapped in Ne, Ar, and Kr. Reasonable agreement is obtained if it is assumed that where two resonances have been reported the impurities are trapped at substitutional and octahedral interstitial lattice sites. The connection between the variational formalism and a previous perturbation treatment of Adrian is discussed and it is shown that a reinterpretation of the latter yields improved agreement with experiment. Crude estimates of the relaxation about substitutional and interstitial impurities are also given.