512 BELL SYSTEM TECHNICAL JOURNAL 



mediate impulses, persisting for a fraction of a second, accompanied 

 by an unusually intense frying sound, indicating a predominance of 

 high frequencies. At no time did this type of disturbance appear to 

 possess marked tonal quality. Each impulse was followed by a quiet 

 period after which a swish occurred. During several periods when 

 the static was sufficiently intermittent, the interval between the 

 beginning of the static impulse and the beginning of the swish was 

 timed. Approximately 70 observations were made, the shortest 

 period recorded being 1.2 seconds and the longest, accurately de- 

 termined, 3.0 seconds; Many ranged between 2.5 and 2.8 seconds. 

 No consistent progression of the length of this swish lag was observed 

 although at certain times a predominance of either long or short 

 periods existed. Later work indicated the long and short periods to 

 be about equally divided between night and day. 



During one night of the New Hampshire work an auroral arc 

 appeared extending from northwest to northeast. Near the north- 

 west end of the arc frequent flashes occurred, but these were too 

 obscure for any details to be made out. A similar but much weaker 

 flashing was observed to the southwest. At times the flashes appeared 

 to extend along the horizon from northwest to southwest. By visual 

 observation while listening to the atmospherics, it was found that 

 nearly every flash coincided with a static crash possessing the promi- 

 nent frying sound. These crashes were in most cases followed by 

 swishes, usually of the descending variety, although occasionally a 

 short ascending whistle occurred simultaneously with the start of the 

 descending swish. 



According to information supplied by the United States Weather 

 Bureau, no lightning storms occurring during this period lay in the 

 direction where flashes were observed to be concentrated and no storms 

 were reported as near as 100 miles to our observation point. The 

 Weather Bureau supplies the information that, under favorable re- 

 flecting conditions, lightning flashes might be seen 40 miles, but could 

 not be seen 100 miles. It therefore appears reasonable to suppose 

 that the flashes observed were of auroral origin. A report supplied 

 by the United States Magnetic Observatory at Tucson, Arizona shows 

 a magnetic storm beginning August 27. Through the following days 

 the disturbance gradually reduced, reaching a low level on September 1. 

 Our observations show the swish intensity to be high from the evening 

 of August 30, when observations began, to September 1. Through 

 September 1 and up to the termination of the test on the morning of 

 September 2, the swish intensity appeared to be reducing although 

 occasional high intensity periods occurred. These and earlier data of 



