No one who has the honour to serve on the Council of this Society, and can therefore be tackled in person once in every month, can long avoid the additional but more alarming honour of being called upon to lecture. When I in my turn was brought to bay, I cast about for a suitable subject, but could not satisfy myself that any of the work in which I happen personally to be engaged had a sufficiently general appeal. I therefore decided to take advantage of my association with the Aerodynamics Department of the National Physical Laboratory, and to base my lecture on the work of other men. I want to make this clear at the outset, because I may omit in some instances to attribute credit where credit is due, and at all events it must be understood that the credit does not belong to me.

It has frequently been stated and written that in order to popularise light aircraft the first essential is the production of a reliable engine capable of being easily maintained and having a long life, at the same time selling at a low figure. In the first part of this lecture it is desired to point out the difficulties in the way of realising this ideal before remarking on the claims of the various types for adoption.

The study of aerodynamics is still largely empirical, and there seems little hope at present that the gap between theory and practice will be sensibly diminished by any means at present known to the applied mathematician. Yet the light that would, in all likelihood, be thrown on a number of practical problems by the complete solution of even a single case of turbulent motion would be very great. Of the problems that can be attempted, that of the flow of a fluid past a circular cylinder is likely to prove the least intractable.