The exact mechanism by which the ribbon graft stimulates union is not known. In all cases, the fragments of the graft began to show coalescence before there was a real attachment of the graft to the tibia itself. During the preliminany stages, union began to appear between the fragments of the tibia; at three months the graft had coalesced and had begun to unite with the tibia. In Case 4 and Case 6, union had been sufficiently stimulated by this time, so that clinically there was no motion and the patient could walk without support. It would seem, therefore, that the graft initially stimulates callus formation in the fracture region locally and that the bone-forming elements there begin to assume their proper functions. After three months, coalescence of the graft occurs, at which time it may serve as a strut, allowing the rest of the fracture to unite

Regardless of the mechanism, there are several highly desirable features of this type of graft. Avoidanec of the fracture site, except for creating a suitable bed in which to place the graft, means but a minimum of disturbance of the blood supply to that region. Minor operative trauma and early coalescence of the graft make this also a helpful adjunct in the treatment of delayed union. Because of the relatively short operative time and the simplicity of the procedure, it can be used in patients of all age groups.

Union occurred in all of these cases except Case 5. It was necessary to repeat the procedure in this single failure, and successful union resulted. An analysis of this failure revealed that the graft did not span the fracture line sufficiently. This error in technique also ocurred in Case 4, but union occurred in that patient despite the error.

From a consideration of the cases reported, the following observations seem true:

1. Osteoperiosteal grafts of the ribbon type coalisce and unite to the shaft in approximately three months' time.

2. By their presence, union is stimutlated within the fracture. In six months' time, there is a solid bar of bone across the fracture and union of the fracture has progressed sufficiently far so that weight-bearing without support is begun.

3. Infection is no contra-indication, when the graft is placed in a non-involved area. Union will occur in the same manner as in non-infected cases.

4. The smaller the chips in the ribbon graft, the quicker the graft itself unites solidly into a bar of bone.

5. Non-fixation of the graft is important, but proper placing of the graft is essential. It must lie well above and below the fracture site. One graft placed so that it extended insufficiently below the fracture site resulted first in partial union, and later in non-union.