96 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1952 



Fig. 5 — Received pulses — disturbed transmission. 



normal one. It is also possible that more than thi'ee paths were involved. 

 On a number of other occasions the pulse coming by way of the second 

 path appeared to be of greater amplitude than the one coming by the 

 main path. This same effect has been obser^'ed by ]\Ir. Crawford and his 

 colleagues on the angle of arrival equipment. 



Information obtained from the above set of pictures shows that for 

 a time-division multiplex system using the length of pulse used here 

 (7 millimicroseconds at the base) and operating over this path, pulses 

 would have to be spaced a minimum of about 14 millimicroseconds apart 

 if it were desired to avoid cUstortion at all times. If verj^ much shorter 

 pulses were used the spacing might be reduced to 9 or 10 millimicro- 

 seconds. However, the 7-foot path difference indicated by these pictures 

 is about the maximimi ever observed and occurs rather infrequently so 

 that if somewhat closer spacings were employed troubles would result 

 only a small percentage of the time. 



The next series of pictures, Fig. 5, taken July 8, show an example 

 of a more common type of multi-path transmission. Here the path dif- 

 ference is apparently less than for the last series. At 11 : 15 there are two 

 distinct pulses with an apparent path difference of about six feet (6 milli- 

 microseconds) if judged from the spacing between the peaks of the pulses. 

 However, from the length of the disturbance at the base line, which we 

 consider a better criterion, the path difference was more nearly four feet 

 At 11:22 distortion of the trailing edge of the pulse was the only indica- 

 tion of a second path. For the pictures taken at 11:29 and 11:44 the 

 path difference is sufficiently small that there is almost complete can- 

 cellation of pulses, only the leading portion of each pulse being present. 



