The results of the Australian expedition show that during the day-time the absorption factor a is independent of the wave length, but that it appears to vary with the direction of transmis sion. For example in trans-Atlantic measurements the attenuation in a West to East direction is lower than that in the opposite direction. Examples of the simultaneous reception of signals both ways round the earth were noted. The reception of abnormally large signal intensities at the Antipodes, found by Lieut. Guierre on the S.S. Aldebaran in 1920, was also confirmed.
Some mean values of the attenuation coefficient for various types of transmission are given below European Stations received in the Atlantic a = o .ocn 8 American Stations received in the Atlantic a=o.00142 European Stations received in the Pacific a = o -00095 (2) Medium Waves.—Measurements across the Atlantic Ocean were made in 1923 by H. D. Arnold and L. Espenschied on a wave length of 30o metres. Agreement with Austin's empirical formula was obtained during the day, but during the night, values agree ing with the simple formula (up) obtained by neglecting the absorption term in Austin's formula were obtained.
(3) Short Waves.—The wave-length used by H. Hertz in his original experiments was about 3 metres. The engineers who developed Hertz's discovery, foremost among whom was G. Mar coni, found that longer wave-lengths gave greater ranges, and from 1895 onwards it was considered that the long waves were more suitable for long distance communication than short waves. Both during and after the World War of 1914-1918, however, more attention was paid to the short wave-lengths. The evolu tion of the thermionic valve had provided the radio-engineer with new tools for both transmission and reception. With it the gen eration of continuous waves down to wave-lengths of a few metres was a simple matter while, at the same time, its inclusion in wireless receiving sets as amplifier and detector had increased the sensitivity of such receivers many thousandfold. About the same time wireless amateurs both in England and America and also the engineers of the Marconi Company began to explore the use of short waves for long distance communication. Three main conclusions were drawn from the amateur and professional experiments. The first was that short waves travelled exceedingly long distances with very small attenuation, so that comparatively low-power stations were required to produce readable signals at the Antipodes. This characteristic of short wave3 could not have been predicted from the Austin formula and from our previous knowledge of the behaviour of long waves. Secondly it was found that, although the signal strength first fell off rapidly as the dis tance from a short-wave station was increased, the signals sud denly appeared in greater strength when a certain critical distance was reached, and only died out gradually as the distance was in creased further. The critical distance at which the strong signal suddenly appeared, the so-called "skipped distance" was found to vary with wave-length, being greater the shorter the wave-length. For example for a wave-length of 3o metres the skipped distance was found to be about 600 km. but for 20 metre waves it was about 1,400 km. The third characteristic of short-wave trans mission was that there often appeared to be two optimum wave lengths, one most suitable for day transmission and the other for transmission by night, so that, by the use of both, communi cation over the whole of the day could be maintained.
An explanation of the "skipped distance" observed in short wave was first given by A. H. Taylor and E. 0. Hulbert, who, in 1926, pointed out that according to (15) and for a constant value of the electron concentration in the upper atmosphere the reduction in the refractive index and thus the maximum angle through which waves may be deviated becomes smaller the smaller the wave-length. It was therefore suggested that the ray received at the edge of the skipped distance was critical in that it had been deviated at the level in the upper atmosphere where the electron concentration was greatest, and that waves meeting the layer with smaller angles of incidence than this critical ray actually penetrated the layer and escaped. This is tantamount to saying that there are no atmospheric waves of normal type re ceived within the skipped distance because there is insufficient electricity in the layer to bend them back. It has, however, been found that it is possible to receive a very weak signal within the skipped distance which, since its intensity is found to vary considerably, must be attributed to down-coming waves. Such signals appear to come from all directions in that they show no directional effects and this has led T. L. Eckersley to suggest that the radiation responsible for these signals is scattered from the waves which are passing overhead and which are bent back to the ground at greater distances.
The use of methods of concentrating radiation into a beam has been developed by the Marconi Company Engineers, and particularly by C. S. Franklin who has, after experimenting with parabolic reflectors of ter the manner of H. Hertz, designed an aerial consisting of a number of parallel vertical wires equally spaced ehind which is a "reflecting screen" formed similarly of vertical wires. Such aerials are used in the series of wireless links between Great Britain and the Dominions which have been erected by the Marconi Company for the British Post Office. The use of similar aerial systems with reflectors for the receiving stations has two advantages. In the first place the use of many aerials brings about the collection of electric wave energy from over a fairly wide area thus increasing the received signal. Sec ondly the reflector acts as a kind of shield and protects the aerial from undesired signals and atmospherics coming in the opposite direction.
Since the attenuation of short wave signals is so low there is not a very great difference in the intensity of signals received both ways round the earth. In the case of i6-metre signals sent from America to Germany, E. Quack has recorded oscillographic ally double signals with a spacing of 0.096 seconds. The first signal received was that travelling by the shorter journey across the Atlantic Ocean, while the second was that taking the long path via the Pacific Ocean. The same author has recorded in stances of the fourfold reception of the same short wave signal, the first signal arriving via the direct path and being followed by signals that have travelled once, twice and thrice round the globe.