The study of 54 'isolated substorms' by Foster et al. (1971) is reexamined, and five different explanations of the results of this study are found to be consistent with the data. For each of the five explanations the decay time of the AE index is found not to be characteristic of internal magnetospheric processes but to be due to data selection. In four of the five explanations the time lag that maximizes the correlation coefficient between Bz and AE is found not to represent the time lag between an interplanetary cause and a consequent magnetospheric effect. It is shown that by simple extensions of the Foster et al. study, at least four of the five explanations can be eliminated. The implications of the Foster et al. study with regard to the 'growth phase controversy' are considered, and proposed extensions of the study appear likely to resolve the controversy.