1. The histology of the healing of apple grafts during the first growing season is described, with special emphasis on the formation of overgrowths known as callus knots. Piece-root tongue grafts of Wealthy scion on French crab root were used. 2. Callus is produced exclusively by tissues located outside of the xylem cylinder. Any living tissue of the bark, excluding the periderm, may proliferate. The cambium may contribute very little of the callus. 3. Proliferation is inhibited locally by crushing of the surface cells, by desiccation, or by destruction of the tissues by fungi. 4. The gap between the respective xylem cylinders of the stock and scion does not become filled with callus during the first season. Indications are that the gap does not become filled during subsequent years. The limited amount of parenchymatous filler callus formed does not become differentiated into vascular tissues during the first year. 5. Unobstructed contact between the respective calli of the stock and scion favors a mingling of the parenchymatous cells, and the continued proliferation of a mixed callus. No method has been found for distinguishing microscopically between stock and scion cells in the mixed callus. 6. In well-matched grafts, an arc of cambium, continuous with the respective cambia of the stock and scion, is differentiated from cells of the intervening callus. The complete cambial layer sheathing the union subsequently lays down a vascular cylinder, which comprises the second annual ring of the tree. 7. The size and vigor of the tree at the end of the first growing season are conditioned by the degree of effective continuity in the new vascular sheath around the union. 8. Excess callus is proliferated wherever there is a local obstruction to effective vascular union between the stock and scion. Large masses of callus, known as callus knots, develop on the scion lip as the result of obstructed union between the scion lip and the stock. 9. Localized callusing may be independent of the general vigor of the tree. Histological interpretations are suggested for the presence of large lip knots on good trees, and the absence of pronounced knots on poor trees. 10. The size and permanence of lip knots or other masses of excess callus are conditioned by the extent and permanence of obstructions to vascular union. 11. Certain phenomena that have been attributed to incompatibility may be interpreted as the result of poor grafting. 12. The internal differentiation of masses of non-pathogenic excess callus resembles somewhat the previously described histology of crowngall.