Several years ago, Dr. Max B. Hardy told me that the inlay bark-graft had been used successfully with Chinese chestnuts at the U.S.D.A, laboratory in Albany, Ga., following Dr. B. G. Sitton's use of this method with pecans in Louisiana. (It is described in a bulletin from Michigan State College, East Lansing, Mich.) I tried it in a small way, and had some success using it on chestnuts in July and August. This spring I suggested it to Mr. Roark and Dr. Richards, both of whom tried it out, using Castanea mollissima stocks and various scion varieties. Mr. Roark used the inlay bark-graft in the spring, topworking a C. mollissima seedling with scions of the Colossal, a hybrid variety from California. About 50 per cent of these have grown this year. Dr. Richards tried it during July, on C. mollissima seedlings from a different source. None of the Colossal would grow on his trees, but he was partially successful with scions of the C. mollissima varieties, Hobson, Carr and Zimmerman. He then devised a variation in the method which was highly successful with C. mollissima varieties. This I shall call the Richards "Twin-T" bud. In "Twin-T" budding, a vertical slit is made in the bark of the stock. Then horizontal cuts are made through the bark at both top and bottom of the vertical cut. The bud piece is cut from the well matured part of a current season's twig, leaving a rather thick slice of wood beneath the bud. (It may be as thick as half the diameter of the twig.) The bud is inserted in the stock as in ordinary T-budding, then wrapped with a large sized rubber budding strip. (Westinghouse electrician's tape and Curity adhesive tape have also been used. Some other brands poisoned the buds.) The "take" of Chinese chestnut buds by this method has run from 60 to 90 per cent on Dr. Richards' trees of various sizes this year. In a short nursery row, buds were placed under first or second year bark, while larger trees were topworked by placing the buds mostly under the bark of second year limbs. The Colossal failed again on Dr. Richards' trees when budded by the "Twin-T" method, but Carr and other Chinese varieties were budded successfully. The graft-compatibility problem in chestnuts is one of considerable complexity. Thus Carr, which has presented incompatibility with certain stocks of C. mollissima at other places, grew on these trees, and Colossal, compatible on another C. mollissima tree, failed on trees which are apparently compatible with Carr. The Chinese chestnut species varies in its graft-compatibilities possibly as much as in other characteristics (growth, productivity, size and quality of nuts, etc.) so that nut nurserymen should begin to select their seed for chestnut understocks with a view toward getting strains with a greater degree of compatibility to the leading scion varieties.