A description is given of a series of experiments to determine the attenuation of wireless waves over land. The intensity of the signals from the London broadcasting station (2 LO; λ = 364 m) was measured along a number of radial lines in seven different directions up to a maximum distance of 100 miles. From these measurements curves were drawn of the overall attenuation, and derived curves showing only the surface absorption effect. The experimental results were then compared with curves drawn in accordance with Sommerfeld's theory of attenuation. The results show that the attenuation is not the same in all directions, the difference being quite important, and the comparison with theory shows for all directions a considerably greater attenuation than would be expected from the previously determined value of the earth's conductivity.It is suggested that both these features of the results could be explained as an effect due largely to the well-wooded nature of the English country-side. To show this, an experimental method was developed by which the energy-absorbing property of individual trees could be directly measured. The results obtained enabled the total energy-absorbing properties of a given tree-covered area to be calculated, thus providing a correction for trees which could be introduced into the theoretical formula.The conclusion arrived at was that the greater part of the discrepancy between Sommerfeld's theory and experiment could be ascribed to the effect of trees, and also that the same agency offered a satisfactory explanation of the difference of attenuation observed in different directions—the greatest attenuation corresponding to the most-wooded regions, and vice versa.The experimental results show no evidence of any screening or other effect due to hill systems encountered in the region investigated, which included the North and South Downs and the Chiltern and Cotswold ranges.Finally, check-measurements of attenuation on two other wave-lengths gave results satisfactorily confirming the conclusion derived from the main experiments.