Wind tunnels became major data-gathering devices for aeronautical engineering. They were indispensible for the design of airships and airplanes (and later also for optimizing the aerodynamic behavior of cars, trains, buildings and bridges). It is not astonishing, therefore, that theoretical concepts of fluid dynamics, too, emerged in close relation to wind tunnel testing. As the scope of fluid dynamics broadened after the Second World War, its roots in aeronautical engineering became obscured. I focus on the relation of the empirical and the formal in boundary layer theory, turbulence research, and gas dynamics. It is rarely remembered that it was in the wind tunnel where the empirical roots of these fields of 20th century fluid dynamics grew in close context with problems of aeronautical engineering.