It has long been recognised that landing and taking—off are the most hazardous periods of the average flight. In this respect an aeroplane is similar to a boat being launched or beached at high speed; the change from one medium of support to another is bound to involve possible damage especially if speeds are high.

So far as risk of mechanical damage, due to uneven aerodrome surface (or swell in the case of a seaplane), is concerned, the speed of landing should be the minimum practicable. Very low landing speeds have a possible disadvantage in that one may not have sufficient aerodynamic control to compete with gusts, etc., but there are few aeroplanes of to-day which have landing speeds so low that this objection can be raised.

The trend of design in recent years has been to increase aerodynamic efficiency by streamlining and to increase the wing loading.

Existing information on the aerodynamic characteristics of aircraft winter landing gear is very meagre and it was therefore desirable to test as many full size skis as possible in order to collect data concerning the various types of skis in actual use.

It was known that the air resistance of existing types is quite high and the experiments were intended to explore the possibilities of producing an improved type of ski having considerably lower air resistance.

The large pitching moments on present types of skis give rise to serious problems in trimming gear design, especially on high performance aircraft. The experiments aimed at producing a ski on which the pitching moments would be considerably reduced in order that the use of exterior trimming cables might be avoided.

The Author endeavours to indicate briefly the growth of flying—boats from their inception to the present day and to forecast their probable line of development in the future. It is shown that our present achievement has been arrived at after the comparatively short period of twenty—five years and an attempt has been made to indicate the type of flying—boat which might normally be expected to mature within the next decade.

The water characteristics of a present—day hull are discussed and typical curves are included giving the resistance, attitude and the effect of applied moment for a flying—boat of approximately 40,000lb., all—up weight.

Before proceeding to a detailed investigation into a particular case of a flying—boat of the future, the Author briefly describes the progress which has been made with regard to the engines and compares the relative advantages of the compression—ignition engine and the petrol engine.Reasons are given why the petrol engine still tends to retain its supremacy.

For the future development the Author has taken the particular case of a flying—boat of 300,000lb. all—up weight and has attempted to analyse the probable structure weight of such an aircraft and what could reasonably be expected in the way of pay load and performance.The general conclusions are that the construction of such a boat would be possible, its structure weight by means of an increased wing loading could be maintained at the present—day figure and that its performance would materially increase.

The standard method of strength calculation of a framework is based on the assumption that members are freely pinned together and external forces are applied only at nodes. Actually aircraft fuselage members are stiffly joined and transverse forces are often acting on them at considerable distances from nodes, e.g., weights of crew, instruments and oil, reactions of controls, stabiliser adjusting gear, etc.

The safest method of computing stress in such members is to assume both ends pin—jointed. As they actually are stiffly connected to adjacent members, they are not allowed to twist freely, i.e., the truss exerts moments on the loaded member. The effect of these end moments consists in diminishing deflection andfield moments. The decrease of maximum bending moment may attain nearly 50 per cent.

Since framework members are designed to withstand axial forces only, viz., tension or compression, they are comparatively weak against bending and have to be strengthened appreciably if transverse forces are present.

It has been known for many years that the rectangular wing stalls first in the central region, burbling spreading gradually tipwards with further increase of incidence. This phenomenon is useful in connection with lateral control, because this control is retained for a while (though reduced) past the stall. Increased use of the1 tapered wing to-day raises the question of whether this phenomenon holds good with tapered wings. There is already some doubt on this score.