The following paper does not pretend to be a complete and exhaustive survey of all the theoretical and experimental work which has been done up to the present on the subject of control beyond the stall.

It is confined to main questions and to a report of the practical research work upon which my colleagues and I have been engaged during the last two years.

In doing so I am well aware of the incompleteness of our research work and its methods, which aimed more at finding ad hoc than general solutions. The outlook of the engineer is different from that of the learned scientist. Not for him is the tranquil and contemplative atmosphere that surrounds purely scientific research work. If he breaks new ground he has neither time nor means to strive for a thorough, complete and general treatment of the particular problem he encounters. His work is limited by financial considerations; hampered by a rigid system of Works Orders; and by the lesser interest in so-called “unproductive” work which characterises industrial institutions. In short, from his point of view, the solution is infinitely more important than the theory or the methods by which it is obtained. A practical solution often presents itself automatically if the physical conditions of a particular problem are properly recognised, even if the degree of accuracy does not satisfy the rigid requirements of pure scientific research.