No abstract

In a recent paper in these Proceedings (Series A, vol. 107, p. 587) Smith-Rose and Barfield have called attention to the two outstanding problems of the propagation of wireless waves over the earth s surface. A complete theory of wireless transmission must explain ( a ), why long-distance communication is possible, and ( b ), why large and rapid variations of signal intensity and apparent direction of propagation of the waves are observed at night, and, to some extent, during daylight, particularly in winter. Smith-Rose and Barfield further point out that both phenomena can be explained to some extent by the well-known Kennelly-Heaviside layer theory, but that it is generally admitted that further evidence of the existence of the layer is needed. They also describe accurate experiments designed to detect the existence of waves arriving at a wireless receiver in a downward direction ( i . e ., inclined to the horizontal), such as must be present if the Heaviside layer theory is correct. In these experiments Smith-Rose and Barfield sought, by directional methods, to detect a departure of the electric field of the waves from the vertical by means of a large Hertzian oscillator, and a departure of the magnetic field from the horizontal by means of a rotating frame aerial. It was, however, found that the conductivity of the ground was sufficiently high to make it act very nearly as a perfect reflector, and, because of the presence of the reflected wave from the ground, none of the effects sought for could be detected even in conditions such as are normally associated with signal strength and directional variations. These authors therefore concluded that the results of their experiments could not be considered as evidence for or against the Heaviside layer theory. In a later paper, Smith-Rose and Barfield describe further experiments of this type, again with negative results, and state that “ adequate experimental evidence on the existence of the Heaviside layer is still lacking.”

In previous communications we have outlined two experimental methods of examining the effects of the atmospheric ionized layer in short-distance wireless transmission. In the first type of experiment the existence of night-time interference phenomena between two sets of waves was demonstrated by changing the wave-length of the transmitter continuously through a small range and observing the resultant maxima and minima of signal intensity. It was suggested that such interference took place between ground waves and waves deviated through large angles by the upper atmosphere. In the second type of experiment the angle of incidence of such atmospheric waves at the earth’s surface was measured by comparing the magnitude of the electric and magnetic forces in the stationary wave system produced at the ground. The results of these experiments were interpreted as yielding a direct experimental proof of the existence of the Kennelly-Heaviside layer, and also as demonstrating that the “fading” of broadcasting signals at moderate distances from the transmitter was due mainly to interference phenomena between two sets of waves arriving at a receiver with an appreciable path difference. But there still remained the problem of the cause of the natural succession of interference effects which constitutes fading at moderate distances, and which takes place continuously throughout the night-time. These variations indicate either that the phase relation between the ground and atmospheric waves is continually changing at night, or that intensity or polarization changes of the atmospheric waves are taking place. In considering possible causes of phase variations, let us examine the relation between the path difference and the wave-length for a typical case of short-distance transmission. Let D represent the path-difference between the ground and atmospheric rays. Then the atmospheric ray arrives n wavelengths behind the ground ray at the receiver, where n = D/ λ , and λ is the wave-length. It has been mentioned above that a possible cause of the natural signal variations which occur at night is a continuous change of phase which would be produced by a change in n . Such a change might be brought about by changes in D, or in λ , or in both simultaneously, and it is necessary to decide between these possibilities. Changes in D might be brought about by a variation in the height of the layer, so that a Döppler effect at "reflection” is produced. In such a case the signal variation might be regarded as the beating between the ground-ray frequency and the reflected-ray frequency. On the other hand, if there is a slow variation of transmitter frequency, the frequency of the atmospheric ray would be different from that of the ground ray, because of the difference in times of emission from the transmitter, and, again, the natural changes might be regarded as beats. The suggestion has already been made by Breit that fading is due to the modulation of the carrier wave, and thus to change of wave-length. In the latter connection we have to consider the variation of both carrier wave and side-band frequencies. The results of our earlier experiments suggested that the change of side-band frequency necessary for the wireless transmission of music is sufficient to produce selective frequency fading, and thus a certain amount of distortion. But with the normal type of modulation the signal intensity is chiefly dependent on the intensity of the carrier wave, and the question whether a slow “swing” of the carrier wave is responsible for such fading (which is observed whether the carrier wave is modulated or unmodulated) seems still unanswered. The question, of course, is equally of interest in both continuous wave telegraphy and wireless telephony.

In the course of a series of signal intensity measurements made at Cambridge on the electric waves received from broadcasting stations a marked difference between day and night conditions has been observed. The day-time signals are sensibly constant but marked fluctuations of intensity become apparent about sunset and continue throughout the night. These variations are detectable at Cambridge on the signals from London, where they represent a change in strength of about 5 per cent.

No abstract