430 EVENING DISCOTJESE. 



satisfactorily explained by assuming that the shorter the wave-length the more likely 

 it is that a ray will penetrate the layer, so that large angle deviation no longer occurs. 



I mentioned in an early part of this discourse that it is nowadays quite a common 

 thing for signals to be obtained both ways round the earth. This result was first 

 recognised by a German radio engineer, Herr Quack, who published in 1926 what 

 we might call documentary evidence of signals received in Germany from America, 

 both by the Atlantic and Pacific paths. In the following year Quack supplemented 

 his first announcement by another just as startling, to the effect that he had succeeded 

 in recording on an oscillograph signals which had travelled once, twice, thrice and 

 even four times round the earth. These multiple signals were found to follow one 

 another at intervals of approximately one-seventh of a second, which is the time 

 required for electric waves to travel a distance equal to the circumference of the earth. 



In 1926 it was, of course, known that short waves appeared able to cover enormous 

 distance with but little loss of strength, but I do not think that anyone would have 

 predicted from the facts then known that a short wave signal would be reduced only 

 to about one half of its original strength in travelling round the earth. 



Although the fii'st round-the-earth echoes were obtained using the American trans- 

 mission received in Germany, it was soon found that the same type of effect could be 

 obtained using local sending stations. A special study was made of the signals sent 

 out by Nauen at a place Geltow, situated about 50 kilometres away. These experi- 

 ments yielded results such as are shown in Pig. 3, where we see examples of echo 

 signals spaced one-seventh of a second apart. 



These observations are specially noteworthy in that they yielded an accurate value 

 of the time required for a wireless signal to travel round the earth. This value was 

 found to be 0.1385 second. Now if we assume that the short-wave signal travels 

 along the ground we can calculate what the time should be, since we know the circum- 

 ference of the earth and the velocity of light. This calculated time is found to be 

 0.1338 second. Now the difference is clearly outside the limit of experimental error, 

 and indicates that the waves do not travel along the surface of the earth. We picture 

 the waves as travelling to and fro between the ground and the layer, thus taking a 

 3 per cent, longer path than that which hugs the surface of the earth. 



Now although the phenomenon of multiple signals is of great interest to us from 

 -a theoretical point of view, we must not forget that to the practical man they are 

 spurious signals which are liable to mutilate the Morse signals which are to be received. 

 Whenever the amplitude of the echo signal exceeds about 30 per cent, of the main 

 signal it is found that trouble in practical operation is experienced. 



In his investigation of round-the-earth echoes Quack came across another type 

 of echo, the origin of which is not quite clear. This was an echo which was observed 

 to be associated with the direct main signal and came at intervals of about one- 

 hundredth of a second after it. These echoes were, therefore, satellites of the main 

 signal and, therefore, quite distinct from the round-the-earth echo. Sometimes as 

 many as seven of these echoes were observed immediately after the main signal. 

 This interval of one-hundredth of a second indicates that the waves had been reflected 

 at some point 1,500 km. away, and this led Quack to suggest that these were the 

 ordinary type of Heaviside layer echoes, the waves having travelled vertically up 

 and down and the height of that layer being 1,500 kilometres. This hardly seems to 

 me to be possible, for the kind of antenna used in Quack's experiments appears to be 

 incapable of sending energy vertically upwards. Here we might note that in America 

 observers have also observed these medium retardation echoes, but their times 

 of retardation show a very wide range of values. 



Mr. Eckersley, of the Marconi Company, has brought forward evidence which 

 strongly suggests that some of these echoes are due to scattered radiation from the 

 point on the ionised layer where the short-wave beam meets it. In this case we are 

 to look upon the beam as being somewhat like a searchlight, an echo being received 

 from the luminous patch where the searchlight meets the layer. This certainly 

 gives echo times which fit in with some of the observed values, but there are many 

 others which are as yet unaccounted for. 



One very interesting practical point arises when these sateUite echoes round the 

 main signal are received. It is usually found that at the same time the round-the- 

 «arth signal is present and is free from such subsidiary echoes. Now the sateUite 

 echoes tend to distort and blur the main signal and sometimes make it quite un- 

 recognisable. We therefore meet with instances in which the signals received 



i 



