46 



SCIENCE. 



[Vol. XVI. No. 390 



of a failure; but because the easiest thing to abuse has always been 

 the buried earth connection, that has come in for the most fre- 

 quent blame, and has been held responsible for every accident not 

 otherwise explicable. 



All this is now changing or changed. Attention is now directed, 

 not so much to the opposing charges in cloud and earth, but to 

 the great store of energy in the strained dielectric between. It is 

 recognized that all this volume of energy has somehow to be 

 dissipated, and that to do it suddenly may be by no means the 

 safest way. Given a store of chemical energy in an illicit nitro- 

 glycerine factory, it could be dissipated in an instant by the blow 

 of a hammer; but a sane person would prefer to cart it away 

 piecemeal, and set it on fire in a more leisurely and less impulsive 

 manner. So, also, with the electrical energy beneath a thunder- 

 cloud. A rod of copper an inch or a foot thick may be too heroic 

 a method of dealing with it; for we must remember that an elec- 

 tric discharge, like the recoil of a spring or the swing of a pendu- 

 lum, is very apt to overshoot itself, and is by no means likely to 

 exhaust itself in a single swing. The hastily discharged cloud (at 

 first, suppose, positive) over-discharges itself, and becomes nega- 

 tive; this again discharges and over-discharges till it is positive, as 

 at first; and so on, with gradually diminishing amplitude of swing, 

 all executed in an extraordinarily minute fraction of a second, 

 but with a vigor and wave-producing energy which are astonish- 

 ing: for these great electrical surgings, occurring in a medium 

 endowed with the properties of the ether, are not limited to the 

 rod or ostensible conduit. The disturbance spreads in all direc- 

 tions with the speed of light; and every conducting body in the 

 neighborhood, whether joined to the conductor or not, experiences 

 induced electrical surgings to what may easily be a dangerous 

 extent: for not only is there imminent danger of flashes spitting 

 off from such bodies for no obvious reason, — splashes which, on 

 the drain-pipe theory, are absolutely incredible; flashes sometimes 

 from 3 perfectly insulated, sometimes from a perfectly earthed, 

 piece of metal, — but, besides this, remember that near any con- 

 siderable assemblage of modern dwellings there exists an exten- 

 sive metallic ramification in the gas pipes, that these are in places 

 eminently fusible, and that tbe substance they contain is readily 

 combustible. 



On the drain-pipe theory, the gas-pipes, being perfectly earthed, 

 would be regarded as entu'ely safe so long as they were able to 

 convey the current flowing along them without melting; but, on 

 the modern theory, gas-pipes constitute a widely spreading system 

 of conductors, able to propagate disturbance under ground to con- 

 siderable distances, and very liable to have some weak and inflam- 

 mable spot at places where they are crossed by bell-wires, or 

 water pipes; or any other metallic ramification. 



Above ground we have electrical waves transmitted by the 

 ether, and exciting surgings throughout a neighborhood by induc- 

 tive resonance. Below ground we have electrical pulses conveyed 

 along conductors, leaking to earth as they go, but retaining energy 

 (Sufficient to ignite gas, whenever conditions are favorable, at con- 

 siderable distances. 



The problem of protection, therefore, ceases to be an easy one, 

 and violent flashes are to be dreaded, no matter how good the 

 conducting- path open to them. In fact, the very ease of the con- 

 ducting-path, by prolonging the period of dissipation of energy, 

 tends to assist the violence of the dangerous oscillations. The 

 drain-pipe theory, and the practical aphorisms to which it has 

 given rise, would serve well enough if lightning w^ere a fairly 

 long-coBtinued current of millions of amperes urged by a few 

 hundred volts, or if there were no such thing as electro-magnetic 

 inertia; but seeing that the inverse proportion between amperes 

 and volts better corresponds to fact, and seeing that the existence 

 of electro-magnetic inertia is emphasized by multitudes of familiar 

 experiments, the drain-pipe theory breaks down hopelessly, and 

 only a few of its aphorisms manage to survive it. 



What, then, are we to set up in place of this shattered idol? 

 First of all, we can recognize what was virtually suggested by 

 Clerk Maxwell, — that the inside of any given enclosure, such as 

 a powder-magazine or dynamite-factory, can, if desired, be ab- 

 solutely protected from internal sparking by enclosing it in a 

 metallic cage or sheath, through which no conductor of any kind 



is allowed to pass without being thoroughly connected to it. The 

 clear recognition of the exact, and not approximate, truth of this 

 statement is a decided step in advance, and ought to be satisfac- 

 tory to those who have to superintend the practical protection of 

 places sufficiently dangerous, or otherwise important, to make the 

 aiming at absolute security worth while. Similarly, for wire- 

 covered ocean-cables absolute protection is possible; but not for 

 ordinary buildings, any more than for ordinary land telegraph- 

 offices, is such a plan likely to be adopted in its entirety. Some 

 approximation to the cage system can be applied to ordinary 

 buildings in the form of wires along all its prominent portions; 

 and such a plan I have suggested, and I understand it is being 

 carried out, for the entrance towers and part of the main body of 

 the present Edinburgh electrical exhibition, Mr. A. R. Bennett 

 having asked me to recommend a plan to the committee as a sort 

 of exhibit. For chimneys a set of four galvanized iron wires, 

 joined by hoops at occasional intervals, and each provided with a 

 fair earth, seems a satisfactory method; but it is to be noted that 

 a column of hot air constitutes a surprisingly easy path, and that 

 it is well to intercept a flash on its way down the gases of a chim- 

 ney by a copper hoop or pair of hoops over its mouth. Mr. 

 Goolden tells me that he has just applied this method to a new 

 chimney at his works in the Harrow Road. For ordinary houses, 

 a wire down each comer and along the gables is as much as can 

 be expected. At many places even this will not be done. A couple 

 of vertical wires from the highest chimney-stacks on opposite sides 

 must be held better than nothing or than only one. 



Earths will be made, but probably they will be simple ones, en- 

 tailing no great expense. A deep, damp hole for each conductor, 

 with the wire led into it, and twisted round an old harrow or a 

 load of coke, may be held sufficient. And as to terminals, rudely 

 sharpened projections as numerous as is liked may be arranged 

 along ridges and chimney-stacks; but I have at present no great 

 faith in the effective discharging-power of a few points, and 

 should not be disposed to urge any considerable expense in erect- 

 ing or maintaining them. Crowns of points on chimneys and 

 steeples are certainly desirable, to ward oil, as far as they can, the 

 chance of a discharge; but a multitude of rude iron ones will be 

 more effective than a few highly sharpened platinum cones. I 

 find that points do not discharge much till they begin to fizz and 

 audibly spit; and, when the tension is high enough for this, blunt 

 and rough terminals are nearly as efficient as the finest needle- 

 points. The latter, indeed, begin to act at comparatively low 

 potentials; but the amount of electricity they can get rid of at 

 such potentials is surprisingly trivial, and of no moment whatever 

 when dealing with a thunder-cloud. 



But the main change I look for in the direction of cheapness 

 and greater universality of protection is in the size and material 

 of the conducting-rod itself. No longer will it be thought neces- 

 sary to use a great thick conductor of inappreciable resistance: it 

 will be perceived that very moderate thickness suffices to prevent 

 fusion by simple current strength, and that excessive conducting- 

 povver is useless. 



In the days when the laws of common "divided circuits '' were 

 supposed to govern these matters, the lightning-rod had to be of 

 flighty conducting copper, and of such dimensions that no other 

 path to earth could hope to compete against it. But now it is 

 known that low resistance is no particular advantage: it is not a 

 question of resistance. The path of a flash is a question of imped- 

 ance; and the impedance of a conductor to these sudden rushes 

 depends very little on cross-section, and scarcely at all on mate- 

 rial. A thin iron wire is nearly as good as a thick copper rod; 

 and its extra resistance has actually an advantage in this respect, 

 that it dissipates some of the energy, and tends to damp out the 

 vibrations sooner. Owing to this cause, a side-flash from a thin 

 iron wire is actually less likely to occur than from a stout copper 

 rod. 



The only limit is reached when the heat generated by the cur- 

 rent fuses the wii-e, or runs the risk of fusing it; but, in so far as 

 oscillations are prevented, the mean square of current strength on 

 which its heating-power depends is diminished. Accordingly, a 

 fairly thick iron wire runs no great risk of being melted. Its 

 outer skin may, indeed, be considerably heated; for these sudden 



