Some preliminary experiments in 1927 showed that the maximum current produced by the incidence of short wireless waves on a tuned rectangular frame aerial was very critically dependent on the dimensions of the frame. An increase or decrease in the width or height of the frame by only a few centimetres might change the current many hundred fold, such current variations being quite independent of the tuning. Furthermore the reduction of current caused by, say, a decrease in the frame width could be compensated by an increase in the frame height and
vice versa
, but the changes in dimensions were not equal in magnitude, neither was their product a constant. In fact, the maximum current depended on the shape of the frame and also varied irregularly with the area. For a given wave-length there were certain critical areas for maximum current, and doubling the area of a frame did not quadruple the current (as when receiving long waves) but the current was reduced to one of negligible magnitude, although the frame was kept properly tuned. It was found that these anomalous effects could be explained by taking into consideration, not only the action of the passing wave, but also the mutual action between the currents flowing in adjacent parts of the frame. In order to do this, it is convenient to consider the current in any limb as the resultant of two component currents ; namely a “direct” component due to the incidence of the wave on the particular limb, and an “indirect” component due to the effects of the currents in adjacent limbs. These two components will, in general, differ in phase and amplitude, and the problem reduces to the determination of those conditions which tend to produce the maximum resultant current.