SCIENCE. 



[Vol. XII. No. 2i 



violent thunder-storms occur after a spell of fine weather, and the 

 soil is likely to be dry. It is best, therefore, to run your conductor 

 pretty deep, and there make earth. 



It is all very well to connect the conductors to water-mains if 

 near; but, if they are far off or non-existent, it is no use ; and in 

 no case, in Professor Lodge's opinion, should they be used as sole 

 earths, certainly not gas-mains. In dry weather they are not earthed 

 at all well, and a strong charge may then surge up and down them, 

 and light somebody else's gas in the most surprising way. It does 

 not often happen, but it may happen in sandy soil after dry weather. 



It is a superstition to place much reliance on the testing of con- 

 ductors with a galvanometer and Wheatstone bridge. A galva- 

 nometer and Wheatstone bridge are powerless to answer many im- 

 portant questions. A Leclanche cell can no more point out what 

 path lightning will take, than a trickle down a hillside will fix j'ou 

 the path of an avalanche. The one is turned aside by every trivial 

 obstacle, and really chooses the line of least resistance ; the other 

 crashes through all obstacles, and practically makes its own path. 

 A flash strikes a house at one corner, rushes apparently part way 

 down the conductor, then flashes off sideways to a roof-gutter, sends 

 forks down all the spouts, and knocks a lot of bricks out. An- 

 other branch bangs through a wall in order to run aimlessly along 

 some bell-wires, and then out through a window-frame, and down 

 a spade or something propped up against the wall, to earth. The 

 lightning-tester comes with his galvanometer and Leclanche cell, 

 and reports that the earth of the conductor has one hundred ohms 

 resistance ; and the accident is therefore accounted for. But how 

 much resistance would he have found in the paths which the light- 

 ning seemed to choose in preference to the one hundred ohms .' 

 Something more like a million probably. 



Something has been left out of consideration, and something very 

 important too ; and until that something is fully taken into account, 

 no satisfactory and really undeniable security can be guaranteed. 

 That something is inertia, — electrical inertia. 



The word ' inertia ' one uses as conveying a correct general notion 

 of the behavior of an electric circuit to sudden electro-motive forces, 

 — a behavior which is caused by the influence or induction which 

 every portion of a circuit exerts on every other portion. Consider 

 a conducting-rod as analyzed into a bundle of parallel wires or fila- 

 ments, and let a current be suddenly started in all. The rising cur- 

 rent in any one filament exerts an opposing force on all the others ; 

 and this self-generated opposition electro-motive force, due to in- 

 duction between the different filaments of the conductor, exactly 

 imitates the effects of ordinary inertia as observed in massive bodies 

 submitted to sudden mechanical forces. 



The term commonly employed to denote the electrical inertia- 

 like effect is ' self-induction,' which is becoming gradually shor- 

 tened to ' inductance.' Its original form when first dealt with by 

 Sir William Thomson was the ' electro-magnetic capacity ' of the 

 circuit. 



Now, since electric inertia is due to a mutual action between the 

 filaments into which a conductor may be supposed divided, it is 

 manifest that the closer packed they are, the greater their inertia 

 will be, and that to diminish inertia it is only necessary to separate 

 the filaments and spread them out. 



The main count of the indictment against ordinary procedure is, 

 that too much attention has been hitherto paid to conducting-power, 

 and too little to inertia. In fact, it is not too much to say that prac- 

 tically nothing but conductivity has been attended to, or thought 

 of, in the erection of lightning-conductors. 



Another way of putting the matter is this. A lightning dis- 

 charge is essentially a varying current : it manifestly rises from zero 

 to a maximum, and then dies away again, all in some extremely 

 small fraction of a second, say, a hundred-thousandth or there- 

 abouts. But that is not all : there is a certain amount of energy to 

 be got rid of, to be dissipated ; and it may easily be that a single 

 rush of electricity in one direction does not suffice to dissipate all 

 the stored-up energy of the charged cloud. If the conductor is 

 highly resisting, a single rush is sufficient ; but, if it be well-con- 

 ducting, it is quite insufficient. What happens then } The same 

 as would happen with compressed air or other fluid rushing out of 

 an orifice. If it is a narrow jet, there is a one-directioned blast ; 

 but if a wide, free mouth be suddenly opened, the escaping air 



overshoots itself by reason of inertia, and springs back again, oscil- 

 lating to and fro till the stored-up energy is dissipated. Just so is it 

 with an electric discharge through good conductors : it is not a. 

 mere one-directioned rush; it is an oscillation, a surging of elec- 

 tricity to and fro, until all the energy is turned into heat. 



There is another fact which it behooves us to be aware of. It is 

 one to the importance of which the attention of scientific men has 

 but recently been called. Experimentally it has been discovered by 

 Professor Hughes ; theoretically, by Mr. Oliver Heaviside, Lord 

 Rayleigh, and Professor Poynting ; for, though the necessary theory 

 is really contained in Clerk Maxwell, it required digging out and 

 displaying. This has now been abundantly done, but the knowl- 

 edge has scarcely yet penetrated to practical men ; indeed, it has 

 not yet been thoroughly assimilated by most physicists. The fact 

 is this. When a current starts in a conductor, it does not start 

 equally all through its section : it begins on the outside, and then 

 gradually though rapidly penetrates to the interior. A steady cur- 

 rent flows uniformly through the whole section of a conductor : a 

 variable current does not. It is started first at the surface, and it is 

 stopped first at the surface. 



Remembering the rapidly oscillating character of an electric dis- 

 charge, remembering also the fact that a rising current begins on 

 the outside surface of a conductor, we perceive, that, with a certain 

 rate of alternation, no current will be able to penetrate below the 

 most superficial layer or outer skin of the conductor at all. In the 

 outer skin, of microscopic thickness, electricity will be oscillating to 

 and fro ; but the interior of the conductor will remain stolidly inert,, 

 and take no part in the action. 



Thus we arrive at a curious kind of resistance, caused by 

 inertia in a roundabout fashion, and yet a real resistance, a reduc- 

 tion in the conducting-substance of a rod, so that no portion ex- 

 cept that close to the surface can take any part in the conduction 

 of these rapidly alternating currents or discharges. It must natu- 

 rally be better, therefore, not to make a lightning-conductor of solid 

 rod, but to flatten it out into a thin sheet, or cut it into detached wires^ 

 Any plan for increasing surface and spreading it out laterally will 

 be an improvement. 



Perhaps it may be as well to guard against one favorite miscon- 

 ception. It has long been known that static charges exist only on 

 the surface of conductors. It has also long been known that ordinary 

 currents flow through the whole section and substance of their con- 

 ductors. It is now beginning to be known that alternating cur- 

 rents may be sufficiently rapid to traverse only the outer layers of 

 conductors ; and this last piece of knowledge is felt to be rather 

 disturbing by those who have been accustomed to dwell upon the 

 behavior of steady currents, and seems like a return to electrostatic 

 notions, and an attempt to lord it over currents by their help. But 

 the first and third facts mentioned above — the behavior of static 

 charges, and the behavior of alternating currents — are two distinct 

 facts, independent of each otlier ; not rigorously independent per- 

 haps, but best considered so for ordinary purposes of explanation. 



We have thus mentioned two causes of obstruction met with by 

 rapidly oscillating currents trying to traverse a metal rod. First 

 there is the direct inertia-like effect of self-induction to be added ta 

 the resistance proper ; the resulting quantity being called by Mr. 

 Heaviside ' impedance,' to distinguish it from resistance proper, for 

 there is a very clear distinction between them. Resistance proper 

 dissipates the energy of a current into heat, according to Joule's 

 law ; impedance obstructs the current, but does not dissipate en- 

 ergy. Impedance causes tendency to side-flash ; resistance causes 

 a conductor to heat, and perhaps to melt. The greater the resist- 

 ance of a conductor, the more quickly will the energy of a dis- 

 charge be dissipated, its oscillations being rapidly damped ; the 

 greater the impedance of a conductor, the less able is it to carry 

 off a flash, and neighboring semi-conductors are accordingly ex- 

 posed to the more danger. Resistance is analogous to friction in 

 machinery ; impedance is analogous to freely suspended massive 

 obstruction, in addition to whatever friction there may be. To 

 slowly changing forces, friction is practically the sole obstruction ; 

 to rapidly alternating forces, inertia may constitute by far the 

 greater part of the total obstruction, so much the greater part 

 that friction need hardly matter. 



This is a fairly accurate popular statement of the direct way in 



