UGIITXIXG CCRRKXTS IX BURIED CABLE 297 



for strokes to ground not arcing to the cable, at least during the time re- 

 quired for the tip of the channel to propagate from the earth towards the 

 cloud, which may be of the order of 50 to 100 microseconds, depending on 

 the height of the cloud. During this inver\-al ionization of the soil around 

 tlie base of the channel provides a path in the earth of low impedance com- 

 pared to that of the channel, as shown in the paper referred to previously. 

 It is possible of course that, durmg later stages of the discharge, the resis- 

 tivity of the earth to some extent may limit the current, as the impedance 

 of the completely ionized channel will then be lower and that in the earth 

 higher because of the lower current in the earth with resultant decrease in 

 ionization. This, however, would tend to reduce the current and thereby 

 decrease rather than increase the time to half-value, and at the same time 

 it would tend to cause a long duration tail current of low magnitude. 



3.2 hicidence of Cable Currents of Various Crest Values 



In Fig. 12 is shown the number of observed currents exceeding various 

 crest values, together with curves of the crest current distribution expected 

 on the basis of the theoretical cur\'es given in Fig. 2. The latter curves, 

 together with those in Fig. 3, are based on an incidence of strokes to ground 

 of 2.4 per square mile for 10 thunderstorm days, a value derived from the 

 rate of strokes to transmission line ground structures, as outlined in the 

 paper referred to previously. Although the observations appear to be in 

 fairly satisfactory agreement with theoretical expectations, a total of 15 

 currents is hardly sufficient as a check of the theoretical curves, particularly 

 since the latter presume a uniform earth structure. 



The intersections of the theoretical curves (Fig. 12) with the axis of ab- 

 scissae indicate that from ti\-e to seven of the currents were due to direct 

 strokes. Actually, visual evidence of direct strokes was found in but two 

 cases, in which the strokes occurred to and damaged test equipment. This 

 does not preclude the possibility of additional direct strokes, as evidence 

 thereof in the absence of cable damage may easily escape detection. 



?>.2> Incidence of Strokes to Ground 



In Fig. 13 is shown the incidence of strokes to ground observed during 

 1947 and 1948 from one point within the test section, by the method de- 

 scribed in Section 2.5. In the same ligure are shown the results of smiilar 

 observations by the same method, made at one location in New Jersey 

 during 1948 for purposes of comparison. Published data obtained from 

 direct visual and aural obser\-ations at one locality in Massachusetts^ 

 are also shown in the same figure. 



As shown by the curves in Fig. 13, the observed or apparent incidence of 

 strokes to ground dmiinishes as the radius of the obser\-ation area increases, 



