This note is presented as an idea to be criticised rather than as a description of a fully developed experimental technique. There are weaknesses in the method which will prevent it from giving absolutely correct answers, for example, the twist imparted to the boundary layer when a body of revolution is rolled cannot be represented. But some of the major effects can be represented, and the idea may possibly be developed to be of practical significance.

It is standard wind tunnel practice to measure the aerodynamic derivatives CYβ, C1β, Cnβ, CMβ, CLβ and CDβ by having a wind stream of constant direction, and measuring aerodynamic forces and moments on a model aircraft placed at angles of yaw and pitch to the stream.