As a preliminary step in the study of a system consisting of a flying insect and the air surrounding it, a comparison is made between the system under natural conditions of free flight and under experimental conditions (when the insect is held stationary), in order to ascertain under what conditions conclusions reached with the insect held stationary would hold for free flight. To avoid the difficulty of making this comparison directly in respect of each of the many factors involved, the resultant force which, acting continuously on the body of the insect, would most nearly replace those cyclicly changing forces that normally maintain or modify the state of motion in flight, is taken as an index of what is occurring in the system. When the insect is held stationary in 'still air' it is found in most cases that this resultant, though of sufficient magnitude to support the insect in flight, does not act through the centre of gravity: the line of action of the resultant intersects the body axis at some point behind the centre of gravity. The position of this point depends on the amplitude of wing beat. When the insect is exposed in a wind tunnel to a stream of air of appropriate speed and direction, however, the resultant acts through the centre of gravity, as in free flight. The forward displacement of the point of intersection between the line of action of the resultant and the body axis, when the insect is exposed to a stream of air, is analysed further, and it is found: (1) that the effect of the stream of air on the body of the insect is negligible in this connexion as compared with the effect on the wings; (2) that changes in amplitude of wing beat do not account for the forward displacement of the resultant; and (3) that when the insect is exposed to a stream of air the path travelled by the wing tip on its downward beat is displaced forwards along the body axis in a direction which would tend to produce the observed displacement of the resultant force. This forward displacement of the path travelled by the wing on its downward beat converts the elliptical course, characteristic of wing movements when the insect is held stationary in 'still air', into the figure of '8' course commonly associated with insect flight, and is dependent on the movement or position of the third antennal joint relative to the second, which in turn is determined by the action of the stream of air on the third joint with its arista. The characteristic attitude of the legs in flight, and the continued vibration of the wings when air is blown at the insect from in front also depend on the sensory inflow from the antennae. The maintenance of the figure of '8' path involves the interaction of the sensory inflow both from the antennae and from the halteres.