LIGHTNING PEEK 



193 



On the other hand, spheres are very little affected because corona 

 does not precede spark over, and the discharge is small and confined 

 to a short path. The field is very nearly uniform, and ruptures 

 r-tart everywhere along the path between the electrodes at approxi- 

 mately the same time. 



THE LIGHTNING ROD 



A study was made, using models built to scale, to determine 

 the protective value of lightning rods. In making these tests, volt- 

 ages varying from 60 cycles to oscillations and impulses were used. 

 The "clouds" were also varied from a point to a large flat plane. 



Fig. 34. — Target made by lightning, i^-inch needle 



It seems that all possible conditions were covered by this range, 

 and since the general behavior of the discharges was the same in 

 all cases, it is believed the results apply to actual lightning discharges. 

 In making the tests the rod was placed in the center of a large 

 piece of drawing paper on a flat plane. Unless otherAvise stated, the 

 plane was of metal. The cloud was placed above the rod. The tests 

 show that lightning from a cloud overhead does not always strike 

 the highest object. The discharges either took place to the rod or 

 to the plane. When the discharge hit the paper a small hole was 

 made. A record was thus obtained of the distributimi of the dis- 

 charges. When the cloud was a plane, a small point was placed 

 at its center to represent the storm center. The apparatus is shown 

 in Figure 32, while sample charts are shown in Figures 33 and 34. 



