1. We investigated the hypothesis that mono- and bi-articular muscles perform different functions: the former are chiefly dependent on their mechanical advantage, while the latter are considered to be mainly concerned with controlling the direction of an external force. 2. Seven subjects were asked to exert a constant external force in various directions from three different positions. Feedback was given on the amplitude (300 and 600 N) and direction of the force vector. 3. During each trial the position of the subject was registered. Ground reaction force and muscle activity (EMG) from the main mono- and bi-articular upper leg muscles were recorded. Link segment modelling was used to obtain net moments about the knee and hip joints. For each muscle the mechanical advantage was calculated in each force direction. 4. The task of controlling the ground reaction force was performed with little interindividual variation as reflected by the variability of the different force and EMG variables. 5. A linear relationship between the difference in activity of rectus femoris and hamstrings and the difference in net moment around the knee and hip was found. This relationship showed very high correlation coefficients of 0.96 (300 N) and 0.97 (600 N) and was independent of position. Mean correlations between this activity difference and the angle of the force vector were also high:-0.95 (300 N) and -0.94 (600 N). 6. The mono- as well as the bi-articular muscles increased in activity when a larger mechanical advantage could be obtained from them, except for the biceps femoris (short head). 7. The results support the hypothesis that bi-articular muscles have a unique role in controlling the distribution of net moments about the joints, and as a consequence, in controlling the direction of the external force exerted on the environment.

We studied six patients to determine the effects of unilateral marginal resection of the proximal part of the fibula on stability of the knee and on gait. At the time of the operation, the fibular collateral ligament and the tendon of the biceps femoris were reattached, but no attempt was made to stabilize the fibula otherwise. The patients were tested an average of sixty-one months after operation. Stability of the knee was measured with an instrumented system. Gait was evaluated with an optical electronic three-dimensional digitizing system and a multicomponent force-platform. The gait of six healthy control subjects of similar age was also studied, and the reproducibility of measurements of stability of the knee was investigated in four healthy adults. There were significant differences between the side on which an operation had been done and the contralateral side with regard to the extent of anterior translation and of total anterior-posterior translation of the tibia at both 20 and 90 degrees of flexion of the knee, and in total varus and valgus rotation of the knee (the number of degrees from a position of maximum varus to one of maximum valgus angulation) at 20 degrees of flexion. The measurements of gait and of motion of the knee were found to be normal when compared with those in the control subjects. In the ground-reaction measurements, there were some significant differences from normal in the medial-lateral plane, but they were clinically unimportant. Resection of the proximal part of the fibula can lead to instability of the knee.