The length of the step is found to increase linearly with the average speed of progression both in walking and in running: only when running at top speed the step length increases at a slower rate. This function is substantially the same in walking as in running and it does not seem to be affected by the incline of the ground or by increasing the body mass with added weights. Two components of the step length are individuated: the first is a constant, independent of speed, and it amounts to about 0.3-0.4 m, the other is speed dependent and it is covered in a constant time of about 0.3 secs. The first component has probably the significance of a movement of translation of the body, sustained by the lower limbs, and it is therefore related to the dimensions of the body. The second component has been tentatively interpreted as due to the free fall period of the body, which is subjected to an upward push in the early part of the step: the constant time employed to cover this second phase of the step has then the significance of the parabola time. On this assumption, the height of the parabola being known, the gravitational force acting on the body is calculated and found to amount to a fraction only (ab. 0.45) of the acceleration of gravity. An increase of speed frequency can be obtained by increasing the gravitational force up to its maximum (9.81 m/sec²) or by decreasing the height of the parabola.