The rotary blade consists of two portions which influence a tillage resistance, i. e, the straight portion and the bending tip portion. First, the total tillage torque curve induced in the standard blade was measured by the tillage tests. Secondly, a tillage torque curve induced in the straight portion was measured by the tillage test using the rotary blade, which was cut off the bending tip portion. Then, the tillage torque curve induced in the bending tip portion was calculated by subtracting the one induced in the straight portion from the standard one. The following results are obtained in this study. 1) The total tillage torque curve can be depicted approximately by a quadrilateral with a concavity, and the one in the straight portion by a triangle. 2) The average of the maximum total torque was 1.8 times as large as one in the straight portion. The occurrence of the maximum torque in the standard blade and the straight portion appeared at 35% and 24% (average) of the total curve rising period, respectively. 3) The average of total torque in the deepest point of the concavity was 39% of the maximum one, and it appeared at 63% (average) of the period. 4) The average of the tillage energy per revolution was 23.758 and 9.166kgf. m in the standard blade and in the straight portion, respectively. Then, the one in the bending tip portion was predicted to be 14.592 kgf. m. Therefore, the tillage energy dissipated in the straight portion and in the bending tip portion were predicted to be 39 and 61% (average), respectively, of the total tillage energy. 5) The area of tillage surface and the length of cutting arc were plotted with the rotation angle. Consequently, it appeared that the rotation angles of the standard blade which the area and the length became the maximum agreed well with that of bending tip portion. 6) This curve was slightly different from the experimental curve, however, we may predict approximately the curve of tillage resistance by this method.