Yields of herbs and essential oils of two interspecific hybrids of mint were determined. One is F1's that involve M. arvensis L. var. piperascens MAL. (Japanese mint) and M. spicata L. (pilose form), the other is F1's that involve M. arvensis L. var. piperascens MAL. and M. spicata L. var. crispa BENTH. Physicochemical properties of essential oils as well as the components that make up the essential oils were also studied. Results acquired were as follows: - (1) F1's grew vigorously and gave better yields of herbs than the parents, though the yields of the different strains (clones) consisting F1's varied one another (Table 1, 2). It was recognized too that drought resistance and rust immunity which F1's possessed. (the parents of non-Japanese mints also had these characters) were also the cause of high productivity of herb, besides the vigorousness of growing. (2) Some F1 strains (clones) wese by no means inferior in yield (%) of essential oil from herb, to the Japanese mint which was generally high in essential oil content. Consequently, their production of essential oil per unit acreage was higher than that of the Japanese mint. (3) F1's that involved Japanese mint and M. spicata var. crispa were superior to F1's that involved Japanese mint and M. spicata (pilose form) in yields of herb and essential oil. Moreover, in the former F1's, there were strains which gave essential oils of good qualities (Table 1, 2;. The authors recognized consequently M. spicata var. crispa was superior as the mother species of interspecific hybrid. (4) Essential oils of two varieties of M. spicata (non-Japanese parents) contained about 25% of free menthol and 50% of unsaturated ketones. Most of F1 strains also contained 25~45% of free menthol and 25~50% of unsaturated ketones in their essential oils. Their herb had a smell peculiar to their unsaturated ketones. However, two F1 strains (F1 89), (F1 92) that involved Japanese mint and crisp mint had about 75% of free menthol and only a few percentage of unsaturated ketones in their essential oils, the composition of which was alike that of Japanese mint (Table 3, 4). They were good perfumers like Japanese mint. (5) Unsaturated ketones which were contained in the essential oils of the non-Japanese mints and the Majority of F1's that involved Japanese mint and non-Japanese mint, or at least most part of the ketones, were supposed to be l-carvone from the physicochemical properties of the essential oils (Table 5). The F1's of interspecific hybrid, especially the last mentioned two F1 strains (F1 89) and (F1 92) had many desirable characters as described above. They closely resembled, though not superior to, Japanese mint in the quality of essential oils and had many good characters (as drought resistance, rust resistance etc.) which originated from the parent of non-Japanese mint. They will answer our purpose as the stocks for the improvement of Japanese mint. From the above mentioned results, the authors are interested in the crossing experiment of Japanese mint with many other species of mint as the means of improvement of Japnese mint.