At the very beginning of these investigations in growing Asiatic chestnuts as timber trees, it was believed that greater success in establishment could be obtained by planting seedlings, rather than by direct seeding. In direct seeding trials during the early thirties the planted nuts were promptly devoured by rodents. Sixteen years of field experience has proven the soundness of this belief. The imported nuts were planted in the Division's nursery at Glenn Dale, Md., and the resulting seedlings distributed as 1- and 2-year-old trees to cooperators throughout the eastern United States. In order to thoroughly test the possibilities of direct seeding as an economical method of establishment, this Division during seven years (1939 to 1942, and 1944 to 1946) planted over 7,000 nuts by direct seeding in 200 locations in 18 eastern States. It was suspected that the greatest hazard to direct seeding in or near forests would be rodents. Accordingly, in the spring of 1939 and 1940, 400 nuts and 600 nuts, respectively, were coated with a strychnine-alkaloid rodent repellent, and a comparable number of seeds, for both years, were left untreated to serve as checks. The checks were held in sphagnum moss at Beltsville, Md., and the nuts to be treated were packed in sphagnum moss and expressed to Denver for treatment by the Division of Wildlife Research, the Fish and Wildlife Service, Department of the Interior. Only 5.9 and 2.5 per cent of the treated seeds developed into seedlings, whereas 22.6 and 13.5 per cent of the untreated seeds produced seedlings. Not only did more of the treated seeds fail to germinate than of the untreated seeds, but the seedlings from the treated nuts were less vigorous. Because of the results obtained, the rodent-repellent study was discontinued at the end of the second year. In 1941 and 1942, over 4,000 untreated chestnut seeds, representing 22 strains, were planted in 12 locations in eight eastern States. The seed source was entirely from American-grown, Asiatic chestnut trees growing in 28 locations in 16 eastern States. They represented Chinese, Japanese, hybrids, and also a limited quantity of American chestnut seed. Seed of the American species was included primarily to determine whether or not it differs from the Asiatic species with reference to establishment by direct seeding. The results for the two years confirmed our earlier beliefs: Only 2.2 per cent in 1941, and 4.0 per cent in 1942, developed into seedlings, of which only a remnant have survived. No species or strain differences were apparent. "Tin Can" Method