“…Ahead of the tests, calculations should be made to seek an optimum test Reynolds number -low enough to give the required extent of laminar flow but high enough to minimise problems at the rear but even then, it is unlikely that complete success will be achieved without the use of suction/cooling. So, the case for either localised or distributed suction (or cooling) is even stronger than for a wing where the boundary layer, full-scale is largely turbulent; indeed, one can say that despite all the added complication, use of cooling/suction is almost mandatory -preferably* 28 ' cooling if practicable near the leading edge to prevent cross-flow instability and suction over the rear of the upper surface to prevent premature rear separation.…”