292 BELL SYSTEM TECHNICAL JOURNAL 



recorded thunder, the distance to the lightning stroke could be determined 

 upon play-back of the wire record. In this manner the number of strokes 

 to ground within areas of various radii around the observation point could 

 be ascertained, and thus the incidence of strokes to ground. These ob- 

 servations were made during the 1947 and 1948 lightning seasons. 



III. Results of Observations 



3.0 General 



From the preceding discussion of theoretical expectations and of the ex- 

 perimental arrangement, it is evident that considerable attenuation would 

 take place between the stroke point and the nearest test points on either 

 side, for a stroke midway between the latter. Accurate evaluation of the 

 maximum current, voltage, and charge, and of the current wave-shape, 

 would thus be rather difficult for strokes nearly midway between test 

 points, since these quantities would have to be evaluated by extrapolation 

 from the obser\'ations at the points along the cable. Such extrapolation is 

 rendered more accurate by employing the theoretical attenuation curves 

 given in Fig. 5. This has been done for the currents, by trial and error, 

 until the current wave-shape derived at the stroke point approximately 

 coincided with that assumed for the attenuation curve used in the extrapo- 

 lation. 



These obser^-ations involving the cable structure extended over the 

 greater part of three lightning seasons, and included a total of 108 thunder- 

 storm days, 35 in 1946, 38 in 1947 and 35 in 1948. The average number of 

 thunderstorm days per year as recorded by the Atlanta Weather Bureau is 

 49, which compares with about 60 given on the U. S. Weather Bureau 

 map.^ The more significant data are tabulated in Table I. 



In the following, the observations made of currents, voltages and charges 

 along the cable are discussed for a number of the more important strokes 

 and compared with theoretical expectations. This is followed by a dis- 

 cussion of the observed incidence of cable currents of substantial magni- 

 tude and of the incidence of strokes to ground observed at one location 

 along the route and at a second point in the northern part of the country. 



3.1 Wave-Shapes and Attenuation of Currents 



In Fig. 8 is shown the distribution along the cable of the crest currents, 

 the crest voltages, and the charges, for the most severe direct stroke meas- 

 sured, which had a crest value of 70 ka and a total charge of 11 coulombs. 

 This stroke occurred to a 35-foot antenna connected to the cable and used 

 in oscillographic observations of induced voltages due to strokes to ground, 

 as another means of securing data on stroke currents. At this same point 



