THE question of a rational system of comparing the efficiency of aeroplanes, even when they are of similar type, is a thorny one, and many methods based on formulæ of varying reliability have been suggested from time to time. It is, in our view, rather doubtful whether it will ever be possible, or even desirable, to agree upon any formula which will cover all types of aeroplanes. The qualities of an aeroplane inevitably differ so much according to the purpose for which it is designed that any formula applicable to all cases is almost impossible to conceive. However that may be, there is little doubt that where it is possible to select a class of aeroplanes of approximately similar characteristics, the best method of arriving at a comparison is to go to fundamentals and consider the machines from a purely aerodynamic point of view. Here, work that has been done all over the world, and is still proceeding, in wind tunnels of all descriptions is reaching the stage when a rational system of comparison can be visualized. There are, of course, still discrepancies in the detailed figures arrived at. But these are becoming smaller, and the results from different wind tunnels are approaching each other more closely. Perhaps more important still, when the discrepancies still appear they are becoming comprehensible. The mist of ignorance is slowly dispersing, and when the discrepancies cannot be avoided, they can now, in a preponderating number of instances at any rate, be explained. The importance of this position cannot be overestimated. It is mainly due to a beginning of an understanding of the true nature of turbulence; an understanding not only of its nature and cause, but of its significance. This has led to a change in experimental methods by which a smooth flow in the wind-tunnel—which was only a short time ago considered the chief end to be aimed at by the tunnel designer—is no longer the one desideratum, or even necessarily a desirable feature. We still, of course, must be able to obtain a completely regular and smooth flow when we need it. But it has become almost more necessary to be able to reproduce turbulent conditions in the tunnel. So that we now have the phenomenon, as it would have been considered in the “old” days, of the research worker deliberately introducing turbulent flow into a tunnel.