When small fragments of the bones of young dogs are used as autogenous grafts, there is a considerable tendency to form new bone. New bone formation is very much less marked when the experiment is carried out on adult dogs.

The tendency to proliferate and form new bone in these small grafts appears to be dependent upon the presence of living osteogenic cells in the grafts.

Osteogenic cells in small numbers are distributed throughout the cortex of the graft, as well as in the cambium layer of periosteum and endosteum, and they get sufficient nutrition to survive. Boiling for ten minutes kills these cells and prevents osteogenesis.

The general principle is laid down that it is fallacious to fragment a bone graft more than is necessary for the technique of the operation, since minute fragmentation lowers both its supporting and its osteogenic functions.

It is emphasized that the intact blood supply of the bed of the graft and the cambium layer of the periosteum are as important for regeneration as is the density of the graft for supportive purposes. There is no evidence in these experiments that metaplasia of other connective-tissue cells to bone-forming cells plays any part in the new bone formation associated with bone grafting.