著者

長戸 一雄 河野 恭広

出版者

日本作物学会

雑誌

日本作物學會紀事 (ISSN:00111848)

巻号頁・発行日

vol.32, no.2, pp.181-189, 1963

被引用文献数

6

---

The studies reported here were undertaken to explore the relations among hardness distribution, the three dimensions of kernel and structure of endosperm tissue with reference to varietal differences of grain texture. 1. Hardness ratio. Hardness distribution of rice kernel is represented by hardness distribution along the dorsiventral line and the lateral line crossing at the central point on the cross section of kernel as reported previously. However, it is more concise and convenient to be indicated by the ratio of hardness of the middle point to that of the central point (Hardness ratio). On the cross section of the kernel of which hardness ratio is less than 1.0, the central core is hardest and hardness becomes smaller toward the peripheral region, and distinct difference can not be found between the hardness of dorsiventral line and that of lateral line. (fig. 1 Century Patna and Zenith) On the section of which hardness ratio is more 1.0, hardness is largest on the middle region and becomes smaller toward the central core and the peripheral region, moreover the dorsiventral region is softer than other region (fig. 1 Asahi, Cody and Yamadanishiki). It is assumed that the former is the characteristic of Indica and the latter characterizes Japonica. 2. Relation between hardness ratio and length-breadth ratio of rice variety. Negative correlation is found between hardness ratio and length-breadth ratio. Regression lines are Y=-0.036X+1.042 and Y=-0.303X+1.603, in Indica and Japonica varieties respectively, neverthless, some Japonica varieties of which hardness ratios are more than 1.18 distribute fairly apart from Japonica line and their length-breadth ratios are 1.7 or thereabout (fig. 2). 3. Relation between length-breadth ratio and thickness-breadth ratio. Generally speaking, positive correlation is recognized between length-breadth ratio and thickness-breadth ratio, but this correlation is scarcely applicable to Japonica varieties (Fig. 3). 4. Relation between hardness ratio and thickness-breadth ratio. There is negative correlation between hardness ratio and thickness-breadth ratio in Indica varieties, yet this correlation is ambiguous in Japonica varieties as well as the relation between length-breadth ratio and thickness-breadth ratio (fig. 4). 5. Structure of endosperm tissue on the cross section. Shapes of the cross sections of kernels vary from round to spindle-shaped according to the thickness-breadth ratios and correspond roughly to hardness ratios as above mentioned (Fig. 5). (1) Cells of the central core. Cells of the central core of A-group (hardness ratio approximately 0.93) are somewhat isodiametric and arranged radially, while those of E-group (hardness ratio approximately 1.20) are uneven and markedly flattened and arrangement of them is disordered. Shapes and arrangement of cells of other groups show intermediate figures between A-and B-groups according to the hardness of central core of each variety. Shapes and arrangement of cells of central core may be affected by the density of strarch in cells, therefore they are correlated with the hardness of central core (fig. 1). (2) Cells along the dorsiventral line. Cells along dorsiventral line are not much different from thme of other region in A-group, but those of E-group are extremely flattened along the dorsiventral direction and arrangement of them is disordered, and those of C-and D-groups are flattened to the extent according to the hardness of dorsiventral line. In Japonica varieties (C, D-and E-groups) starch accumulation in cells of several layers along the dorsiventral line is slightly or markedly insufficient, for this reason, these cells are nattened and arrangement of them is disordered by the oppression of surrounding cells. This characterisic of endosperm structure may be the making of the facts that the kernels of E-group become often white-cored during development and dry kernels of Japonica especially of E-group make frequently dorsiventral