引用文献:Journal of Heredity, 54: 207-211, 1963
Yo Takenaka
 The Yoshino Cherry, [Prunus yedoensis Matsumura (Somei-yoshino)] the most popular flowering cherry tree grown in Japan, is also one of the spring attractions of Washington, D.C. However, its origin is obscure. It is said that it was first sold commercially about 100 years ago by the Edo (Tokyo) nursery of Somei under the name of Yoshino-Zakura (Yoshino is an historical spot in Nara Prefecture, famous for its cherry blossoms.). Mr. K. Fujino, who studied the cherry trees in Tokyo's Ueno Park from 1882 to 1884, identified three distinct species, one of which he named Somei-yoshino. His report was published in 1900(1), and in 1901, J. Matsumura gave it the species name of P. yedoensis(4). Since 1952, I have tried to determine the origin of P. yedoensis on the basis of observations and crossing experiments.
 Japanese botanists have made several studies of the taxonomy of those species of Prunus native to Japan, and have found them very difficult. Consequently, the species discussed in this paper have been given different names by the tax-onomists who have studied them. The names used in this paper are those recog-nized by Dr. J. Ohwi.
 Materials and Methods
 This species bears abundant and beautiful flowers, but sets only a few seeds and can only be propagated by grafting. However, it grows more rapidly than any other cherry species. Assuming that it might be a hybrid、 I collected seeds from a number of trees in 1952. The seeds were sown in 1953, and the seed-lings were observed from 1954 to 1963. My observations suggested that P. yedoensis might be a hybrid between P. lannesiana var. speciosa (Oshima-zakura) and P. subhirtella var. pendula form ascendens (Edo-higan), whose main characteristics are as follows: speciosa is distinguished from ascendens by larger underneath glabrous leaves and vigor-ously growing stems (Table I). The seedlings of yedoensis showed in these respects a series of intergrades ranging from speciosa to ascendens. Some of the seedlings bloomed in the spring of 1958, and since then, more than 100 trees have blossomed. In many flower characters, viz., size, color, hairiness of calyx, style and ovary, and calyx shape, the trees from seeds of P. yedoensis showed a wide range of variation from speciosa to ascendens. It was found further that a few of them showed typical characteristics either of speciosa or ascendens (SI-21 and SI-12 in Table I). The various characters and their occurrence are given in Table II.
 In further experiments started in 1957, reciprocal crosses between speciosa and ascendens were carried out. In 1961, eight hybrid plants began to bloom、 and since then 15 trees have blossomed. They were intermediate between the parents in the characters of stem, leaf and flower, and appeared as a whole to be rather uniform, though they showed some minor differences. No marked differences were found between reciprocal crosses. The variation in the distribution of their characters is given in Table III.
 Compared with P. yedoensis, six F1 plants had hairs on the style and ovary; four had scanty hairs, while the re-maining five were hairless. All F1 plants had about 15 percent larger leaves and flowers and a few more stamens than P. yedoensis, but otherwise closely re-sembled P. yedoensis (Table I). Since the parental trees used for this crossing experiment were selected at random, it is likely they were heterozygous and that F1 plants segregated for hairiness andimay he assumed that P. yedoensis is a hybrid between speciosa and ascendens and thenewly produced F1 represents a synthesized P. yedoensis.
 There are differences of opinion as to the origin of P. yedoensis. It is traditionally said that it was brought to Tokyo from Oshima, one of the Izu islands. But on that island the only flowering wild cherry is P. lannesiana var. speciosa (Oshima-zakura).
 Dr. E. Koehne gave a flowering cherry tree from the Korean Quelpart island (collected by Taquet, a missionary), the name of P. yedoensis var. nudiflora(2). Later, Dr. G. Koidzumi reported that P. yedoensis came from Quelpart island(3). On the other hand, Dr. E. H. Wilson stated, "To me, P. yedoensis Matsumura strongly suggests a hybrid between P. subhirtella var. ascendens Wilson and the wild form, P. lannesiana Wilson. It has many characters of the latter and in its venation, pubescence and shape of the cupula resembles the former"(6). Dr. Koidzumi visited Quelpart island in 1932 and found a tree which he identified as P. yedoensis. Since his identification was questionable, I visited the island in 1933 and observed that the tree, which was growing wild, showed differences from P. yedoensis; the hairs on calyx lobes and on the lower side of leaves were less numerous, and the peduncles were shorter. I concluded that it could not be P. yedoensis. I assumed that it might be a hybrid between P. subhirtella var. pendula form ascendens (Edo-higan) and P. quelpartensis (Tanna-Yamazakura; perhaps a form of P. verecunda) or some other cherry species(5).
 In Japan, the putative parental species of P. yedoensis, P. speciosa and P. ascendens are known to grow together in the Boso and Izu peninsulas. I surveyed wild cherry trees in these two districts. Though both species were found in Izu, only speciosa was found in Boso. Further, in Izu, I found trees which could presumably be the offspring of P. yedoensis, and also a tree which appeared to be a new hybrid between speciosa and ascendens. That tree was identical with one of my synthesized trees, and I named it "Funabara-yoshino" (Table I). These observations suggest that P. yedoensis may have origi-nated in the Izu peninsula.
Observations of the flowering cherry tree (Prunus yedoensis) suggest that it may be a hybrid between P. speciosa and P. ascendens and that P. yedoensis may have originated in the Izu peninsula.
 Literature Cited
1. Fujino, K. Species of flowering cherry in Ueno Park. Jap. Hort. Mag., 92:1, 1900.
2. Koehne, E. 95 Prunus yedoensis Matsum. var. nudiflora Koehne. Repretorium Specierum Novarum Regini Vegetavilis, 10:507, 1912.
3. Koidzumi, G. Prunus yedoensis Matsum. is a native of Quelpaert. Acta Phytotaxonomica et Geobotanica, 1:177, 1932.
4. Matsumura, J. Cerasi Japanicae duae Species novae. The Botanical Magazine. Tokyo, 15:100, 1901.
5. Takenaka, Y. On a native place of Prunus yedoensis. Bulletin of the Japan Society for Pre-serving Lands, Capes and Historic and Natural Monuments, 11:1, 1934.
6. Wilson, E. H. The Cherries of Japan, p. 16, 1916.

Figure 2 (THREE SPECIES OF CHERRY)- Flowering specimens of the cherry are (i1) Prunus lannesiana var. speciosa (Oshima-zakura); (B) Prunus yegensis; and (C) Prunus subhirtella var. pendula form ascendens (Edo-higan). (figures omitted)
Figure 3 (HYBRID CHERRY BLOSSOMS) Blossoms of Izu-yoshino which is one of the synthesized hybrids between P. subhirtella var. pendula form ascendens and P. lannesiana var. speciosa. (figures omitted)
Figure 4 (NATURAL AND INDUCED HYBRIDIZATION) The natural hybrid Funabara-yoshino (left), is compared with P. yedoensis (right). (figures omitted)