In introducing the subject of valve steels to this Society I do so, not because of any special knowledge which I possess, but on account of the general interest which is being taken in the subject at the present time. It is quite evident that the valve steels which have been regularly used are not satisfactorily meeting all the present conditions, and therefore advances will have to be made, especially as much greater engine efficiencies are being demanded. Presumably the valve has to withstand much severer conditions than formerly. In many cases the result, as far as the valve is concerned, means a higher working temperature, which in turn affects many other properties of the steel such as scaling, the tendency to air-harden, etc.

It is with considerable diffidence that I find myself about to address you to-night upon the subject of the Paper, which is “The Training of Aircraft Apprentices.” I say with considerable diffidence, firstly; because I realise how very important to the future of aeronautical practice is this question of the preliminary training of mechanics, and secondly, because my own practical experience of this subject is limited to a few years.

I hope that some of those who will take part in the discussion, after the reading of the Paper, will correct any mis-statements I may make in connection with civilian apprentices and will be able to put before you such problems as, are peculiar to that side.

In Mr. Fearn's article on this subject in the February number of the Journal, in which a solution is given in the form of a Fourier series, a reference is made to the unsatisfactory results of an attempt to use double differentiation of the displacement curve. It is generally recognised that such a process is fundamentally inaccurate, and analytical methods, either exact or approximate, are usually given in books dealing with mechanisms. For mechanisms of the piston-connecting rod type, in which the higher harmonics of the motion of the driven member diminish very rapidly, two harmonics in all generally give all the necessary accuracy, both for displacement and acceleration. Mr. Fearn's solution is of this type, though he includes a fourth harmonic (see e.g., the expression for ρ at the top of page 138).