Most of the twenty-two patients who came for further treatment after analysis with a previous analyst (or analysts) re-created the problems of the preceding analysis and, in general, found the experience of transference unacceptable. In most cases the essential genetic facts were uncovered in the first analysis and there was no justification for considering it a failure. If any technical error existed in the first analyst it was in his insistence that the patient work 'analytically' and accept the reality of the transference. This was usually difficult for the patient to do. Where a change of analysts was necessary for external reasons, the