Abstract: The present paper deals with the systematic position of Daphniphyllaceae. The genus Daphniphyllum was first described by Blume in 1826 as a member of Rhamnaceae. In 1858 Baillon removed it to the tribe Phyllantheae of Euphorbiaceae, while Müller (1869) raised this genus to the rank of family, Daphniphyllaceae. Although Müller’s treatment has been accepted by most botanists, including the present authors, its systematic position has been debated. The first aim in our studies on the cladistics of Hamamelidae is to answer the question which families should be included in this monophyletic group. By observing their pollen grains and stoma types of some representative species of Daphniphyllaceae, Hamamelidaceae and Buxaceae under light microscope (LM) and scanning electron microscope (SEM,) and analysing morphological, anatomical, palynological, embryological characters and chemical components in the three taxa and Euphorbiaceae, we find that Daphniphyllaceae is very similar to Hamamelidaceae, but greatly different from Euphorbiaceae, in inflorescence racemose or spicate, calyx nearly reduced, stamens numerous and sometimes synandry, connective usually exserted, disc absent, carpels 2; vessel with scalariform perforation plates and often not spiral-thickened, fiber bordered-pitted; stomata mostly paracytic; pollen 3-colpate; tapetum glandular, endosperm development cellular, obturator and caruncle absent; iridoid compounds present; sieve-element plastids S-type. The present authors have noticed the fact that Daphniphyllaceae is also similar to Magnoliaceae in the stamens numerous, anthers larger and filaments very short, connectives obviously exserted and with several bundles; anther wall thicker, endosperm development cellular, embryo small. It is considered that not only are Daphniphyllaceae and Hamamelidaceae phenetically close to each other but also much possibly derived from a common ancestor, the extinct group of Magnoliales. However, Daphniphyllaceae appears to be remote from Euphorbiaceae and Buxaceae in relationship and should be separated from Euphorbiales and Buxales. Meanwhile, since Daphniphyllaceae differs from the members of Hamamelidales in the incompletely septate ovary, drupaceous fruit, indistinct sexine sculpture of pollen grains, small embryo, and an unique alkaloid, daphniphylline, but lacking proanthacyanins, the establishment of an order, Daphniphyllales, for the family, is considered reasonable. According to our opinion, the order is related to Hamamelidales rather than to Euphorbiales as originally suggested by Huru-sawa (1954).

Key words: Daphniphyllaceae, Stoma, Pollen morphology, Systematic position