160 



BELL SYSTEM TECHNICAL JOURNAL 



shown in Fig. 12, strip 4, which gives the date and time the record 

 was started, July 23, 1925, 2:06 a.m. (Eastern daylight saving time). 

 The record in Fig. 13 is of the carrier and side-band signals with 

 500-cycle modulation made at Riverhead, L. I., May 25, 1925, 1 :25 a.m. 

 More gain was used in the side-band amplifiers for this record in 

 order that the effects of fading could be brought out more prominently. 

 In this record only half of the side-band signals were recorded, the 



Time 



Fig. 13 — Fading record of carrier and side-band signals, made at Riverhead, L. I. 

 Timing interruptions in side-band signals, 5 seconds apart 



zero reference line being at the edge of the strip. The upper trace 

 is the upper side band, the center the carrier and lower trace the 

 lower side band. Where the traces of the signals overlap a darker 

 record is obtained. This record may be confusing at first but if 

 strip 5 is examined where the amplitudes of the signals are not so 

 large a better picture of the form of the record will be obtained. 



It is obvious from these records that the carrier and side-band 

 signals do not fade together as a unit. The carrier may pass through 

 a zero value with still considerable amplitude in the side-band signals 

 as in strips 1 and 3. In the first case, strip 1, the three frequencies 

 successively fade through points of minimum signal in the order lower 

 side-band, carrier and upper side-band; and in the second case, 

 strip 3, the three frequencies fade through points of minimum signal 

 in the reverse order. This is a definite indication of selective fading; 

 that IS, fading is a function of frequency as well as time. 



