1. Decalcified lyophilised rat bone matrix prepared by Urist's method acts as an inductor of cartilage and bone when implanted into animals of other species, namely mice, rabbits and gerbils. Induction in rabbits and gerbils was very much weaker than in the mouse.

2. The site of implantation affected the outcome; intramuscular implants induced cartilage and bone more strongly and regularly than subcutaneous or intraperitoneal implants.

3. Rabbit transitional epithelium, growing in cortisone-treated gerbils, caused bone induction, but in general, results with this species suggest that it responds poorly to bone-inducing stimuli.

4. Cortisone, used as an immunosuppressant, did not inhibit bone and cartilage induction.

1. Cancellous bone cubes from calf and man were deproteinised with hydrogen peroxide and with ethylenediamine.

2. Long bones were removed aseptically from sheep, stored in the bone bank and used for cancellous homografts.

3. Holes were drilled in the upper part of the tibia or ulna or in the lower part of the femur of sheep. Some were left empty; others were filled with plugs of the deproteinised heterogenous bone, with autografts, or with homografts.

4. Histological appearances were studied after seventeen and thirty-six days.

5. At seventeen days repair was more advanced in the plugged holes; the biological result was better with the ethylenediamine-treated than with the peroxide-treated material. After thirty-six days repair was at an advanced stage. As much new bone had been deposited on the trabeculae of the deproteinised heterografts as on those of the homografts.

6. There was no evidence of metaplastic bone formation; new bone seemed to form from endosteal osteoblasts.

7. Certain clinical implications are briefly discussed.