664 HENRY A. BOWL AND 



to be .taken into account, I believe it is scarcely worth while making 

 tubes, or flat strips, for such small sizes. 



It is almost impossible to draw proper conclusions from experiments 

 on this subject in the laboratory such as those of Prof. Oliver J. Lodge. 

 The time of oscillation of the current in most pieces of laboratory ap- 

 paratus is so very small, being often the T70 000777 ^ a second, that 

 entirely wrong inferences may be drawn from them. As the size of 

 the apparatus increases, the time of oscillation increases in the same pro- 

 portion, and changes the whole aspect of the case. I have given 

 TT7Vr7 of a second as the shortest time a. lightning flash could proba- 

 bly occupy. I strongly suspect it is often much greater, and thus de- 

 parts even further from the laboratory experiments of Professor Lodge, 

 who has, however, done very much toward drawing attention to this 

 matter and showing the importance of surface in this case. All shapes 

 of the rod with equal surface are not, however, equally efficient. Thus, 

 the inside surface of a tube does not count at all. Neither do the corru- 

 gations on a rod count for the full value of the surface they expose, for 

 the current is not distributed uniformly over the surface; but I have 

 recently proved that rapidly alternating currents are distributed over the 

 surface of very good conductors in the same manner as electricity at 

 rest would be distributed over them, so that the exterior angles and cor- 

 ners possess much more than their share of the current, and corruga- 

 tions on the wire concentrate the current on the outer angles and dimin- 

 ish, it in the hollows. Even a flat strip has more current on the edges 

 than in the centre. 



For these reasons, shape, as well as extent of surface, must be taken 

 into account, and strips have not always an advantage over wires for 

 quick discharges. 



The fact that the lightning rod is not melted on being struck by 

 lightning is not now considered as any proof that it has done its work 

 properly. It must, as it were, seize upon the discharge and offer it an 

 easier passage to the earth than any other. Such sudden currents of 

 electricity we have seen to obey very different laws from continuous ones, 

 and their tendency to stick to a conductor and not fly off to other ob- 

 jects depends not only on having them of small resistance, but also on 

 having what we call the self-induction as small as possible. This latter 

 can be diminished by having the lightning rod spread sideways as much 

 as possible, either by rolling it into strips, or better, by making a network 

 of rods over the roof, with several connections to the earth at the corners, 

 as I have before described. 



