All of the five cases in Group 1 were in reality congenital, or birth fractures. So far as we know, the fractures are all ununited. The eight cases in Group 2 were fractures of infancy or early childhood. Probably thus far none of these are united. Therefore, so far as treatment is concerned, this group is just as unsatisfactory as Group 1. In some of the patients in Group 2, particularly in Cases 8 and 9, brother and sister, there may be hereditary factors. Seven of the thirteen patients in Groups 1 and 2 were operated on at the Clinic without success. However, two patients have not been traced.

The six patients in Group 3 all sustained their fractures in middle or late childhood. In Case 15, in which there was failure, the patient was operated on under the age of puberty, the remainder were operated on after puberty, and all obtained union. In Case 6 (Group 2) a second operation was performed after puberty, with failure; a low grade infection may have been the cause.

In a paper I published last year in which forty-four cases of non-union in the tibia, with eight failures, were reported, six of the failures were in fractures sustained in childhood. The percentage of success for the fractures sustained in childhood was 25. The percentage of success for the entire group was 81.8. If the six cases in children were omitted, the percentage of successes in the adult would be raised to 94.

Syphilis was ruled out in all cases; however, in one instance the mother gave a positive Wassermann reaction, but the father and the patient gave negative reaction. This patient cannot be traced. In some of the recent cases analyses of the blood were carried out and the calcium, magnesium, and phosphorus contents were found to be normal.