Abstract
For pt.I, see abstr. A44813 of 1974. The generalization of the augmented plane wave (APW) method, made by Ferreira et al. is re-examined, and its consequences fully explored. The authors show that, among the possible values for lambda , the parameter of the generalized method lambda =0 which corresponds to the standard APW, is usually the best choice. Furthermore, it is shown that the results for lambda = infinity , when compared with those of lambda =0 (standard APW), can be used to estimate the quality of the calculation and the degree of convergence.