In recent years, time honored statistics purporting to show this or that in tuberculosis have been subjected to criticism because of the alleged inclusion of case material that was not tuberculosis. Thus, with the recognition of Legge-Perthes disease, the reports of the results of the treatment of tuberculosis of the hip joint in children became the object of acrimonious dissension. Particularly has this been the case since the operative school, headed by Hibbs,1Albee and Hass, has challenged the rationale of the conservative treatment of joint tuberculosis. The older European school, as exemplified by Rollier,2has stoutly maintained that a considerable proportion of tuberculous hips (40.6 per cent), knees (58.4 per cent) and feet (66.6 per cent) heal with retention of normal motion and do not recur on prolonged observation. The operative school suggests that these best cases never were tuberculosis and asks whether it is not likely