July 13, 1888.] 



SCIENCE. 



19 



which self-iiiduclion aids resistance proper in obstructing an alter- 

 nating current. But, in addition to these considerations, there is 

 that other indirect way which we have also mentioned ; viz., the 

 fact that conduction of alternating current may be confined to the 

 surface of a rod or wire if the alternations are rapid enough. This 

 cause must plainly increase total impedance; for the total channel 

 open to such a current is virtually throttled, as a water-pipe would 

 be throttled by a central solid core. 



But which part of the total impedance does it affect? Does it 

 increase the resistance part, or the inertia part ? In other words, 

 does this throttling of a conductor act by dissipation of energy, or 

 by mere massive sluggishness ? Plainly, it must act like any other 

 reduction of section : it must increase the resistance, the dissipat- 

 ing-power of a conductor, the heating-power of a current. Hence 

 the resistance of which we have spoken as entering into the total 

 impedance has by no means the same value as it has for steady 

 currents, and as measured by a Wheatstone bridge. It is a quan- 

 tity greater — possibly much greater — than this; and, in order to 

 calculate its value, we must know not only the sectional area and. 

 specific conductivity of the conductor, but also the shape of its sec- 

 tion, and the rate of alternation of the current to he conveyed. 



We may here note a vigorous controversy, or difference of opin- 

 ion, between Faraday on the one hand, and Sir W. Snow Harris on 

 the other. Faraday was often consulted about lightning-conductors 

 for lighthouses, and consistently maintained that sectional area was 

 the one thing necessary, weight per linear foot, and that shape was 

 wholly indifferent. Harris, on the contrary, maintained that tube- 

 conductors were just as good as solid rods, and that flattened rib- 

 bon was better still. Each is reported to have said that the other 

 knew nothing at all about the matter. Of course, we know that 

 Faraday was thinking of nothing but conduction, and conduction 

 for steady currents. Harris had probably no theoretical reason to 

 give, but was guided either by instinct or by the result of experience. 

 In this particular, Faraday was wrong, and Harris was right. 



But, it may be said, have not experiments often been made as to 

 the advantage of tape over rod forms of lightning-conductor, with 

 negative results? Yes, but the point usually attended to is the 

 deflagration of the conductor. Mr. Preece, for instance, with Dr. De 

 la Rue's battery, found ribbon and wire equally easy to deflagrate 

 by the discharge. But we are not examining which form of con- 

 ductor is least liable to be destroyed by a flash (probably there is 

 not much to choose between one form of section and another, for 

 there is no time for surface cooling) : we are examining which form 

 will carry off a charge most easily, and with least liability to side- 

 flash ; and here thin ribbon shows distinct advantage over round 

 rod. 



It is found that a rod of iron carries off a discharge more satis- 

 factorily than a rod of copper. It would seem as if the poorer con- 

 ducting qualities of iron enabled the discharge to penetrate deeper, 

 and so to make use of a greater thickness of skin. 



But, every one will say, surely iron has far more self-induction 

 than copper. A current going through iron has to magnetize it in 

 concentric cylinders, and this takes time. But experiment declares 

 aganist this view for the case of Leyden-jar discharges. Iron is ex- 

 perimentally better than copper. It would seem, then, that the 

 flash is too quick to magnetize the iron, or else the current confines 

 itself so entirely to the outer skin that there is nothing to magne- 

 tize. A tubular current would magnetize nothing inside it. Some- 

 how or other, the peculiar properties of iron, due to its great mag- 

 netic permeability, disappear. 



If it turns out to be true that an iron rod does not get magnetized 

 by the passage of a rapidly alternating current, it may be held a 

 natural consequence of the fact that such currents flow mainly in 

 its outer surface, and that such tubular currents have no magnetiz- 

 ing power on any thing inside them. 



The niagnetizabiliiy of iron is no objection to its employment in 

 lightning-conductors. Its inferior conductivity is an advantage in 

 rendering the flash slower, and therefore less explosive. Its high 

 melting-point and cheapness are obvious advantages. It is almost 

 as permanent as copper, at least when galvanized ; and it is not 

 likely to be stolen. Professor Lodge regards the use of copper for 

 lightning-conductors as doomed. 



It is found that a conductor is more eflicicnt in carrying off a dis- 



charge and preventing side-flash, in proportion as its self-induction 

 is lessened ; say, by spreading it out into a thin sheet, or cutting it 

 up into a number of wires, or otherwise. But no conductor is able 

 to prevent side-flash altogether, unless it is zigzagged to and fro so 

 as to have practically no self-induction : in that case the side-spark 

 is nearly stopped. But so long as a conductor is straight fand a 

 lightning-conductor must, of course, be straight), so long will there 

 be some tendency to side-flash, however thick it be made. It may 

 be a foot or a yard thick, and yet not stop it. A man touching a 

 lightning-conductor, however well earthed, might perhaps receive a 

 shock suflicicnt to kill him. 



How can this tendency to side-flash be further diminished ? To 

 stop a pipe full of water from being burst by a blow given lo the 

 water, you will make the pipe elastic. An elastic cushion will ease 

 off the violence of the shock of a water-ram. 



Electric inertia was known by the other name of ' self-induction ; ' 

 electric ' elasticity ' is known by the other name of ' capacity.' In- 

 crease the capacity — not the thickness or conducting-power, but 

 the electrostatic capacity — of your conductor, and it will be able 

 to carry off more. 



The only practicable plan is to expand it over as much surface as 

 possible. A lead roof, for instance, affords an e.xpansion of fair 

 capacity which may be easily utilized ; and there should be as little 

 mere rod-projection as possible before some extent of surface be- 

 gins. Flat sheet for chimneys is better than round rod : it has at 

 least some more capacity, and much less self-induction. 



For tall isolated chimneys Professor Lodge would suggest a col- 

 lar of sheet metal round the top and at intervals all the way down ; 

 or a warp of several thin wires instead of a single rod, joined to- 

 gether round the chimney by an occasional woof ; or any other 

 plan for increasing capacity and area of surface as much as possible. 

 As to the liability of things to be struck, several questions suggest 

 themselves : Is a small knob at a low elevation as liable to be 

 struck as a large surface at a higher elevation ? Is a badly con- 

 ducting body as liable to be struck as a well-conducting one ? In 

 other popular words, does a good conductor ' attract lightning ' ? 



In answering this question experimentally, one must draw a 

 careful distinction between the case of a flash occurring from an 

 already charged surface, which has strained the air close to burst- 

 ing-point before any flash occurs, and the case of a flash produced 

 by a rush of electricity into a previously uncharged conductor too 

 hastily for it to prepare any carefully chosen path by induction. 

 The two cases are (i) steady strain, (2) impulsive rush. 



Experiment on the liability of things to be struck when the air 

 above them is in a state of steady strain, gradually increased, shows 

 that the flash actually prefers to jump three times as much air to a 

 sharp point, and encounter a megohm resistance, rather than take 

 the short direct path offered by a bigger knob. 



By modifying the experiment so as to get an impulsive rush, all 

 bodies are equally liable to be struck if at the same height, and no 

 one is more liable than another: simply the highest is struck if they 

 are at all equally conducting. But by making one bad-conducting, 

 its protective virtue is gone. This is the real objection to a bad 

 earth : it cannot protect well against these sudden rushes. 



Sudden rushes are liable to occur : the clouds spark first into one 

 another, and then, as a sort of secondary effort or back kick, into 

 the earth. In these cases the best conducting and highest objects 

 are struck, quite irrespective of any question of points and knobs. 

 Points are no safeguard against these flashes. The point gets 

 struck by a vivid flash. It has no time to give brushes or glows : 

 its special efficacy in preventing discharge e.xists only in the case of 

 steady action, where the path is pre-arranged by induction. In the 

 case of these sudden rushes, the conditions determining the path of 

 discharge are entirely different. No doubt they have to do with 

 what is called the ' time-constant ' of the various conductors. 



Electrical oscillations are of considerable interest, and have sun- 

 dry practical bearings. When a flash strikes a system, the elec- 

 tricity goes rushing and swinging about everywhere for no appar- 

 ent reason, just as water might surge about in a bath or system of 

 canals into which a mass of rock had just dropped, splashing and 

 overflowing its banks. Just so with electricity. Bell-wires, gas- 

 pipes, roof-gutters, conduct side-flashes in a way most puzzling to 

 the older electricians ; and thus gas may get ignited in the most 



