Studies on Viola kitamiana Nakai. Comparison with Viola crassa Makino

Abstract

Viola kitamiana Nakai (1928) is hitherto known to be endemic to the Shiretoko Peninsula, northeast Hokkaido, Japan. Nakai classified this species in the sect. Nomimium-subsect. Campylostylae, based on the shape of its stigma. He also noticed of its general habit being so much alike to Viola crassa Makino of the sect. Dischidium. Later V. kitamiana was regarded to belong to the sect. Dischidium and by some authors as a variety of V. crassa of the same section.
From our investigation, we found that there are some marked differences between V. kitamiana and V. crassa. The similarity of the habit of these species can be ascribed to their similar habitats. Though they are so much alike to each other in the character of rhizomes and especially of leaves in their texture, anatomical structure, colour, and serrations etc., the shape of leaves and the stipules, the shape and size of stomata are different. On the whole the leaves of V. kitamiana are triangular and shallowly cordate at the base, but in V. crassa they are almost depressed orbicular and more deeply cordate (Figs. 1, 2, 13-14, 3; Table 2). The stipules of V. kitamiana are smaller and crenate-dentate on margin, but in V. crassa the stipules are larger and almost entire. In the former, the stomata are often grouped in a pair or a trio measuring in average 35.7×20.6μ; those in the latter are larger and scattered, measuring 42.8×25.7μ (Table 3). V. kitamiana has always a single scaly leaf at the base of stem, while in V. crassa it is rarely found (Figs. 1a; 2A 18-19). The size of pollen grains and seeds show also certain quantitative differences. In the former pollen grains are mostly 24-27×23-24μ in size, but in the latter they are 36-37×32-33μ (Figs. 2A, B, 7-8). Seeds are on an average 2.2×1.5mm in the former: 2.5×1.9mm in the latter (Figs. 2A, B, 11-12).
It is especially noteworthy that the shape of stigma of V. kitamiana differs entirely from that of V. crassa (Figs. 2A, B, 9). On this basis we can not regard this species to belong to the sect. Dischidium. As to the true systematic position of Viola kitamiana we would like to report in our forthcoming paper.