1226 THE BELL SYSTEM TECHNICAL JOURNAL, NOVEMBER 1954 



tainly can conclude from this that circular electric wave transmission 

 over great distances is possible. 



The long waveguide line also provided a very convenient way of dem- 

 onstrating additional multi-mode transmission effects. For example, in a 

 multimode medium we maj^ use mode filters. One such filter may have a 

 very low loss for the circular electric waves but very high loss to other 

 waves. Such mode filters have been built, and Fig. l-l shows the trans- 

 mission changes which result when introducing one. The upper half of 

 Fig. 14 shows the photograph of the oscilloscope in the time interval to 

 llAi with no mode filter in the line. (This is the same as Fig. 12.) On 

 introducing the mode filter into the waveguide, the received signal is 

 altered as shown in the lower half of Fig. 14. We observed that energy 

 propagating in the undesired modes has largely disappeared; it was 

 absorbed by the mode filter. 



There are still a few small pulses in the lower half of Fig. 14 which 

 cannot be in the TEoi mode because of their time position. Starting at 

 time l.loA^, there is a series of regularly spaced pulses in Fig. 14 labeled 

 TEo2 , and a single small pulse labeled TEos . The geometric placement 

 of resistive material in the mode filter leads us to anticipate low filter 

 losses for the entire circular electric (TEon) family of modes, and there- 

 fore the extra pulses were suspected of being in higher-order circular 

 electric modes. Only two such modes, TE02 and TE03 , were above cut- 

 off. The TE03 pulse was tentatively identified by noting that its group 

 velocity was 55 to 60 per cent of that of the TEoi pulses. High attenua- 

 tion in the TE03 mode prevented additional TE03 pulses from being 

 observed. 



In the case of the TE02 series of pulses, it was possible to get a fairly 

 accurate measure of relative group velocity, confirming the identifica- 

 tion as TE02 . Note that the seventh TEoi pulse coincides with the sixth 

 TE02 pulse, and that the pulse at 7 At shows on the TEoi train as being 

 too large in amplitude. 



The smooth decay of the TEoi train in the lower half of Fig. 14 in the 

 interval At to 6A^ is in marked contrast to the corresponding pulse train 

 in the upper half of the figure, and is graphic illustration of the im- 

 portant effects that mode filtering can produce. 



Another very important transmission observation appeared during 

 Mr. Beck's experiments with the 5" diameter line. He observed that the 

 attenuation, as measured by the amplitude decay of the shuttling pulse, 

 was a function of the position of the piston at the far end of the line. 

 Translations of the far end piston on the order of 10 to 40 centimeters 

 changed the overall transmission from a condition in which the original 



