94 THE BELL SYSTEM TECHXICAL JOURNAL, JANUARY 1952 



to that at the end of exposure.) A second microscope made it possible 

 to view and to photograph the screen of the tube simultaneously. The 

 general procedure was to obser^'e continuously during periods of dis- 

 turbed transmission, taking pictures at regular intervals of 5 to 10 min- 

 utes. When conditions were seen to be changing rapidly, pictures were 

 taken much more frecjuently. The large oscilloscope with its vertical 

 amplifier had a bandwidth of about 150 mc and hence caused some 

 deterioration of the pulse. It, however, was less tiring than the small 

 scope, especially for long periods of observation and was watched to 

 follow the general trend of events. It was capable of resoh'ing the pulses 

 resulting from two-path transmission when the path cUfferences were 

 large. 



The sweep circuits for the two indicator oscilloscopes were practically 

 identical. The horizontal sweep \^oltage for each consisted of the linear 

 portion of a sine wave which was generated by a c-w oscillator operating 

 at one third of the pulse repetition frequency of 10 mc. Each oscillator 

 was synchronized with the incoming pulses by means of a 10-mc voltage 

 deri\'ed by amplifying the pulse energy through a narrow band amplifier. 

 This circuit provided very satisfactory s\aichronization e^'en during the 

 times when signal amplitude was so low as to produce a xevy poor signal- 

 to-noise ratio. Timing markers were provided on each roll of film by 

 periodically photographing a series of pulses spaced by an inter\'al of 9 

 milhmicroseconds. 



RESULTS OF THE EXPERIMENT 



The picture at the left of Fig. 3 shows the transmitted pulse. The 

 right-hand picture shows the recei^•ed pulse under what were considered 

 to be normal transmission conditions. It is seen that, except for the addi- 

 tion of noise and widening of the pulse due to passage through the ampU- 

 fiers and other equipment, the pulse shape is unaffected. The time cali- 

 bration on this and the following photographs are in millimicroseconds, 

 each mark representing one millimicrosecond (0.001 yns). 



During the summer of 1950, when this experiment was in progress, 



A 



Fig.3 — (Left) transmitted pulse (right) received pulse — normal transmission. 



