192 BELL SYSTEM TECHNICAL JOURNAL 



signal may be coming in on both loops while the vertical antenna 

 pick-up approaches zero. Several points of this kind are marked by 

 arrows below the midcile trace in Fig. 36. 



There are at least two simple possibilities which might account 

 for these relations. In case the wave approaches the receiving 

 point from directly overhead, the vertical antenna would receive a 

 "zero" signal while the loops would pick up an amount depending 

 upon the state of polarization. If this be true, the records indicate a 

 very rapid shift from the vertical direction of reception since the 

 antenna minima are short lived most of them lasting at best a small 

 fraction of a second. 



On the basis of wave interference it is apparent that two waves 

 approaching the receiving point in a 90-degree space phase relation 

 and 180 degrees out-of-time phase could give a maximum signal 

 on the two loops while that received on the vertical antenna was a 

 minimum. 



A compromise between these two viewpoints is probably a better 

 guess than either one of them taken alone. That is, the existence 

 of minima on the vertical antenna at the same moment that a strong 

 signal is coming in on the loops is perhaps due to the interfering 

 combination of waves having components in both the \ertical and 

 horizontal planes. 



Quality Distortion 



So far the data shown have been limited to the results of observa- 

 tions taken on special forms of transmission which are simplified 

 for the purpose of clearly exposing the basic facts. We wish now 

 to consider some of the more practical aspects of signal distortion. 

 The first test which we made at our field test station was to record on 

 slowly moving photogiaphic paper tape and on the high speed film, 

 the detected audio signal which resulted when the transmitter was 

 modulated by a pure 264-cycle tone. 



Fig. 37 is a sample of the general type of audio signal record ob- 

 tained and Fig. 38 shows copies of the wave shape of the received signal, 

 at particular times corresponding to the numbers of the oscillograms 

 on the records in Fig. 37. The abrupt displacement of the timing 

 trace indicates the point on the long record at which the snap-shot 

 oscillogram was made. A peculiar characteristic of these records is 

 the dark shadowy lines weaving back and forth through the band 

 recording the complete signal. These dark lines correspond to the 

 kinks in the wave shape shown in Fig. 38. As explained before, the 

 darkening of the record is caused by the greater quantity of light 



