experimental animals serving some dark purpose of the Foundation. Eventually, however, a surprisingly strong concurrence emerged. We found we were impressed by the extent to which in our discussions we placed heavy reliance on the products of two types of research: first, those few longitudinal studies related to juvenile delinquency and crime that had been pursued in this country and, second, a few experimental studies that had sought to measure the consequences of different official interventions in criminal careers. These two research strategies had taught us much about crime and its control. Other strategies-case studies, cross-sectional surveys, participant observations, and similar techniques-had indeed been productive, but it was the longitudinal and experimental designs that firmed up the knowledge that the others helped to discover.