Recently, Moriyama (1977, 1978) and Moriyama and Asai (1980) suggested that the grain size distributions of lower alluvial plain sediments and of river bed materials were strongly affected by the size characteristics of rock-forming minerals and weathered materials in the upper drainage basin of the Yahagi River composed of granitic rocks. At the same time, it was also pointed out that hydraulic conditions in sediment-transportation and accumulation acted to vary the mixing proportions of normally distributed component populations of sediments. These suggestions are very interesting, but are still presumptive, and are not proved actually. In order to prove these suggestions, it is neccessary to make clear the manners of sediment transport occurred on the hill slope in source regions of drainage basin. Then, the writers analysed the grain size distributions of the weathered materials of granitic rocks which were supplied to river bed, and of the river sands which were considered as bed load, and of the suspended materials during floods in the source regions. In addition, they made clear how the grain size distributions of suspended materials changed in the proportions of component populations as hydraulic conditions. The writers selected two small basins different in size characteristics of rock minerals. One is the Mitachi River composed of the Inagawa Granite (coarser grained granite) and the other is the Shinpuku ji River of the Busetsu Granite (finer grained granite) . Both kinds of granitic rocks are the most widely distributed in the Yahagi River catchment basin. Samples of weathered materials were collected from B and C holizons of hill slope soil at each point. As for suspended materials, turbid surface water were scooped up every thirty minutes at the end of the small basins. Sediment grains coarser than 3.5 phi were analysed by the method of mechanical sieving (1/4 interval), and finer silt and clay particles were analysed by the photoextinction sedimentation method (Moriyama, 1976, 1980) . The results of analyses are shown in Figs. 3, 4, 6 and 7. First of all, both of the histograms of weathered materials of grainitic rocks and of river sands in the Mitachi River basin have three peaks around -2 phi, 0 phi, and 2 phi, respectively. Namely, peak size of each sub-population is coincided with each other between the weathered materials and the river sands. In the Shinpuku ji basin, both histograms of weathered materials and of river sands have two peaks around 0 phi and 2 phi, and the peak positions are also mostly the same. These facts are explained that the weathered materials of granitic rocks were carried and accumulated on river bed during the flood by being preserved their size characteristics. Next, the weathered materials of the Mitachi have several peaks around -2 phi, 0 phi, 2 phi and 5 to 6 phi (the peak of 3 phi also can be seen), and the suspended materials have the marked peaks around 2 phi, 3 phi and 5 to 6 phi. The peak positions recognized in both materials are similar to each other. On the other hand, the histograms of weathered materials in the Shinpuku ji have two peaks around 5 to 6 phi and 6 to 7 phi in clay population, and these peaks coincide considerably in position with those of clay populations of suspended materials. These facts suggest that the weathered materials outflow as suspension or wash load by being preserved their peak positions, and that the hydraulic condition has relation to the change of mixing proportions of sub-populations. The mixing proportions of 2 phi populations of river sands decrease gradually from the upper reach to the lower one in both rivers, while in the size distribution of suspended materials, the mixing proportions of 3 phi populations increase, then those of 2 phi populations increase in turn as discharge increases.