300 BELL SYSTEM TECHNICAL JOURNAL 



apparent number of strokes to ground within an area of radius r would 

 then be, for an actual incidence n of strokes to ground and with P{I) = 



Jo 



\l-e-"' (1 + ar)] 



'0 



where a = kiuo/cY'^. 



The apparent incidence of strokes to ground Ua = Na/irr"^ is accordingly 



Ua = jAr^ [1 - e-"' (1 + ar)] 



By choice of a proper value of a in the latter expression a theoretical 

 curve, varying in substantially the same manner with r as a given observed 

 curve, may be obtained. The actual incidence is next obtained by taking a 

 value of n such that the two curves substantially coincide. This value of 

 n also corresponds to the incidence given by the theoretical curve for r 

 = 0, i.e. the value that would be expected if a sufficient number of ob- 

 servations were available for small values of r to permit extrapolation of 

 the observed curves to r = 0. The value of n obtained in the above manner 

 is about 2.8 for the New Jersey and Massachusetts and 1947 Georgia ob- 

 servations. The latter extended over the last half of the lightning season, 

 while the 1948 Georgia observations, which indicate a higher incidence, 

 extended over the entire season. The comparison, shown in the figure, 

 between the observations in Georgia during the first and second halves of 

 the 1948 season, indicates that the incidence during the first half is about 50 

 per cent greater than during the second half. A similar comparison of the 

 New Jersey observations, not shown in the figure, mdicates the opposite 

 trend, i.e. a somewhat smaller incidence during the first half. The diflfer- 

 ence, however, is less marked than in the Georgia case. 



There is reason to believe that this change in Georgia with the advance 

 of the season is due to a change in the character of the lightning storms. 

 During several years the more severe lightning damage on cable routes in 

 this territory has occurred during early-season thunderstorms, which ordi- 

 narily are of the "frontal" type extending over fairly wide areas where hot 

 and cold masses of air come together. These storms appear to be of greater 

 extent, duration, and severity than the "convection" type of storm, ordi- 

 narily experienced later in the season, which occur more frequently as the 

 result of local air convection currents but are of more limited extent and 

 duration than storms of the frontal typeJ 



