12 AERONAUTICS IN THEORY AND EXPERIMENT [CH I § 13. Experience has shown that the instruments at our disposal for wind channel tests enable the forces and moments to be measured to an accuracy of approximately 1 %. This is extremely fortunate in view of the fact that frequently the effects of very delicate changes in relative positions of parts and modifications in shape require to be investigated. In consequence great care must be taken to ensure that the models themselves be made with extreme accuracy. It has been remarked, and the real significance will be brought out in the next chapter, that the nature of the air flow in the vicinity of any body determines the forces and pressures exerted by the wind upon it. If a strut, whose shape is such, as to allow the air to stream easily past it, be not quite truly formed, the discrepancy from the perfect shape will continually act as a source of disturbance to the air by altering the direction of the flow in its neighbourhood. Energy will thus be uselessly expended and the resistance of the strut consequently increased. Certain portions of the strut—and in fact of most bodies entering into aeronautical work—may be more sensitive than others to such discrepancies in shape. It is the object of the next chapter to study the laws of air flow and the principles that must be used in the determination of the shapes of bodies most suitable for the functions demanded of them. The importance of this aspect of the subject cannot be too strongly emphasised.