METHOD AND MATERIALS The patients studied were selected from those cases which were recorded as having features of TLE in the hospital diagnostic index, the files of the neurological and neurosurgical departments and the index of the EEG Department. They had all attended The London Hospital between the years 1949 and 1967. This search made it unlikely that more than a few patients were missed. We thus obtained more than 800 possible cases. In each the clinical and EEG features were studied. Some patients who had only isolated episodes of automatism were rejected. Patients with long-standing epilepsy in whom a temporal lobe focus developed late and who had no clinical features of TLE were excluded, as were patients with psychiatric disorders in whom the attacks