1. The effect of implanting heterogenous anorganic bone, homogenous organic bone, autogenous compact bone from the iliac crest, and autogenous bony callus into circumscribed defects in the femur of albino rats of the Wistar strain is described.

2. Neither heterogenous anorganic bone nor homogenous organic bone appeared to induce new bone formation in a healing defect.

3. Some of the osteogenic cells of autogenous callus implants survived transplantation to a bone defect and gave rise to new bone formation. This did not occur when compact bone from autogenous iliac crest was implanted.

4. Implants of autogenous callus, autogenous compact bone, homogenous organic bone and heterogenous anorganic bone all impeded the normal development of host bone trabeculae in a healing bone defect, seemingly because they acted as physical barriers to the proliferating host callus. None of the implant materials appeared to suppress the healing reaction ofthe host.

5. Implanted homogenous organic bone was removed and replaced by host bone more quickly than was implanted heterogenous anorganic bone, and it appears to be the better material for grafting into bone defects.

6. Autogenous callus or autogenous cancellous bone is a superior implant material to autogenous compact bone and is the bone graft material of choice.

7. The absorption of all the implant materials used in this investigation was associated with the presence of multinucleated giant cells.

8. The activity of multinucleated giant cells may be influenced by the organic matrix of the material which is to be absorbed.

9. Except when fresh autogenous callus was implanted into the defects, the rate of healing in the grafted defects was slower than that in the control defects. In the defects grafted with fresh autogenous callus the healing rates of the control and grafted defects were the same.

1. An experimental study of the healing mechanism in circumscribed defects in femora of albino rats of the Wistar strain is described.

2. Only the outer one-fifth of the defect is repaired by subperiosteal bony callus, the rest of the defect being repaired by endosteal callus.

3. Subperiosteal callus does not bridge the defect until endosteal callus is developed fully.

4. As peripheral callus matures the greater part of the endosteal callus is resorbed, with the exception of trabeculae attached to the margin of the defect.

5. The resorbed area in the medullary part of the defect is gradually obliterated by deposition of inner circumferential lamellae.

6. There appear to be differences between the mechanism responsible for repair of fractures of a long bone and that which heals circumscribed bone defects.

1. Penetrating defects were cut in the femora of twenty-five albino rats. In fifteen of the animals the defects in the right legs were protected with cellulose-acetate shields while those in the left legs were unprotected and allowed to heal as controls. In the remaining ten animals the defects in both legs were protected with shields made of homogenous organic bone.

2. New bone was found to proliferate into the concavity of the shields in most of the animals and this protruded beyond the contour of the femur. The development of the protuberance appeared to depend upon the degree to which the shield was adapted to the femoral surface.

3. The cellulose-acetate shield was not removed by the host, but the homogenous organic bone was actively resorbed; multinucleated giant cells were associated with this process.

4. There are indications that the maintenance of the protuberance is dependent upon the continued presence of the shield. Exostoses protected by intact cellulose-acetate shields have been recognised up to eighteen months after operation.

5. The function of the shield in the formation of the bony protuberance is thought to be two-fold, in that it protects the haematoma from invasion by non-osteogenic extra-skeletal connective tissue, and that it governs the size of the haematoma and prevents its distortion by the pressure of the overlying soft tissue.