Popular Science Monthly/Volume 25/October 1884/Protection Against Lightning II

From Wikisource
Jump to navigation Jump to search


End of series



IN the year 1875 the Meteorological Society of London was moved to follow the lead of the French meteorologists in reference to lightning-conductors, and to appoint a Lightning-Rod Committee. From the report made to the society by the council in the following year, it appears that the objects contemplated in this action were "an investigation and record of accidents from lightning, an inquiry into the principles involved in the protection of buildings, the diffusion of exact information regarding the best form and arrangement for lightning-conductors, and the consideration of all phenomena connected with atmospheric electricity."[1] It is obvious that in its first conception this committee was intended to be essentially one of investigation and inquiry, and it was for this reason appropriately designated a "Permanent Committee." The meteorologists concerned in its inauguration were actuated by the same consideration that was present to the Section of Physics of the Academy of Sciences in Paris when the following paragraph of the instruction of 1854 was drawn up:

One knows, it is true, a very great number of examples of people being killed or of houses being set on fire; one knows, also, many and diverse instances of metals fused, of timber shattered, of stones and even of walls thrown far away, and many other analogous effects; but what is generally wanting is precise measurements relative to distance, dimensions, the position of the object—both that which is struck and that which escaped. For it is necessary to know what the lightning spares, as well as what it strikes. It is the work of all observers, but especially of officers in the navy and artillery, of engineers, of professors, inventors, and architects, to test these phenomena at the moment they are produced, and to describe them accurately for the benefit of science, as well as that of public economy. Such descriptions, when they refer to a stroke of lightning, should as much as possible point out the track of the lightning from its highest to its lowest point; also they should show, by sufficiently numerous horizontal sections, the relative positions of all objects in a circle wide enough to take in those which have been struck.

In this passage the instruction of the French Academy no doubt touches the one point which is necessary before all else to improve, if not to perfect, the practice of electrical engineering, so far as this is aimed against the destructive powers of lightning. The broad principles upon which the engineer prosecutes his work are happily such as can be referred to actual experiments carried out by the artificial apparatus of the electrician. But there still remain some incidental questions, such as the influence of surface, extent, and form in conductors, the relation of conductivity to tenacity, the area of protection, and the maximum effect of lightning, which can not be settled in this way, and which require an appeal to the larger operations of Nature. This, however, concerns opportunities which can not be arranged at will. The method of the appeal must of necessity be observational rather than experimental. It proceeds upon the lines of close watching and systematic record. Observations where the great operations of Nature are concerned are utterly worthless unless they are made with scientific insight and precision. The plan of investigation that has to be pursued is therefore to collect an exact account of all accidents that occur, and to arrange a system of organization which enables all such chance opportunities to be seized upon and improved by an immediate investigation of concomitant conditions and circumstances. This method of study also must be followed up by patient persistence for a considerable length of time, seeing that accidents from lightning occur at uncertain intervals, and that they are scattered capriciously over the greater part of the surface of the earth. It is for this reason, essentially, that a Lightning-Rod Committee needs to sit in permanence.

The Committee of the Meteorological Society, however, seems very soon to have lost sight of its own excellent design, and to have changed its plan into a mere conference for the preparation of a report, which was drawn up under its auspices and printed and published in 1882, apparently by the conference itself, and which assumes the form of a code of rules for the erection of lightning-conductors, with numerous appendices referring to authorities which had been in some sense consulted. The report is published under the editorship of the secretary, and simply as having been considered and adopted by the delegates of the conference, who seem indeed to have concentrated their attention upon one subordinate object which had been proposed by the Meteorological Society, namely, "the diffusion of exact information regarding the best form and arrangement of lightning-conductors," and to have overlooked entirely the more important work of observation and record which had been contemplated by the society in the first instance, and to which we have drawn attention.

The code of rules put forward by the conference was obviously intended to possess the same kind of authority and position as the "instructions" of the earlier French reports, and indeed its chief value seems to be the approval it accords to the practice of construction which had grown out of those instructions, and which is very generally in use at the present day. It virtually confirms most of the conclusions which had been arrived at by the French commissions.

The "Rules" of the London Conference direct that the main stem of the conductor shall consist of a copper rod or tape, with an ascertained electrical conductivity amounting to ninety per cent of that which pure copper would possess, and weighing six ounces per foot; or that it shall be an iron rod weighing two pounds and a quarter per foot; and that the earth connection shall be made by a copper or iron plate presenting a superficial area of eighteen square feet, imbedded in moist earth, and surrounded with coke. The terminal points are to be more prominent than those usually adopted in England, but they may be less so than the heavy tiges of thirty-three feet employed in France. The rod is not to be insulated from the building, but intimately connected with all large masses of metal used incidentally in the construction. All joints in its length are to be imbedded in solder. Curves are not to be made too sharp, and ample provision is to be secured for free expansion and contraction by varying temperature. Water-mains and gas-mains are to be utilized as means of earth contact wherever practicable, and the conducting integrity of the rod is to be tested every year.

A careful perusal of the French instructions, or of Mr. Richard Anderson's very excellent manual upon lightning-conductors, published in 1879, will show that this is substantially an authoritative acceptance of the measures already advised by the best authorities. It is, however, somewhat remarkable that in the report itself of the London Conference nothing whatever is said of the influence of length in reducing the efficacy of a conductor. This is the more strange, because, in speaking of the care required for the formation of joints in the "final decision of the conference on controverted points," the report categorically remarks that bad joints have the same effect as "lengthening a conductor," and a reference is incidentally made to one instance, in which a bad joint was found to have had the same effect on a discharge of electricity that the lengthening of a conductor to nineteen hundred miles would have had. This nevertheless was a point that was perfectly understood by the French investigators, and it is obviously one in which the London code is behind its predecessors. In the first French instructions, issued in 1823, there is a paragraph which says:

Among the conducting bodies there are none, however, which do not oppose some resistance to the passage of the electric force; this resistance to the passage, being repeated in every portion of the conductor, increases with its length, and may exceed that which would be offered by a worse but shorter conductor. Conductors of small diameter also conduct worse than those of larger diameter.

It follows, as a matter of absolute certainty from this increase of resistance with augmented length, that a conductor which was of ample dimensions for the protection of a building eighty feet high would not be of the same efficacy for a building four hundred feet high. It is for this reason that M. Melsens employed eight main conductors for the Hôtel de Ville at Brussels, and it is for this reason that eight half-inch copper ropes have been carried down from the lantern and cupola in St. Paul's. To use eight main conductors of a given size is obviously, in an electrical sense, the same thing as to use one conductor only of eight times the size.[2] The practice of the French engineers has hitherto been to double the sectional capacity of the rod for each additional eighty feet of the length that is to be protected by its instrumentality. This practice is a sound one, and certainly should be observed.

There is one other particular in reference to the conference report to which it seems desirable to draw attention on account of the erroneous doctrine to which it may possibly give a sanction. Among the appendices which have been added to the report there is a table, obviously prepared at the cost of some labor, which professes to give the sizes of lightning-conductors recommended by various authorities. In order to facilitate the comparison of the several sizes, all have been reduced to what has been termed the equivalent dimensions of copper. But the oversight has been made, in preparing this table, of treating all cases of galvanized iron as if the zinc in the combination had no other function than the protection of the iron from rust. In reality, however, a galvanized iron rod conducts as a combination of iron and zinc, in which the zinc possesses a much higher conducting power than the iron. Zinc surpasses iron in this particular at least three times. All the statements of conductivity that have been drawn from galvanized iron conductors have hence been given much too low. The influence of a too powerful electrical discharge upon a conductor of galvanized iron is, in the first instance, to strip off its coating of zinc by melting this more readily fusible metal. But until this is done the zinc assists very materially in the transmission of the discharge. Practically it is known that galvanized iron ropes effectually transmit discharges which could not be safely carried by ungalvanized ropes of the same diameter. The table is on this account worthless for the purpose for which it was avowedly prepared. It attributes to several of the authorities which are named views on the matter of the size of lightning-conductors which they would certainly not indorse. For instance, Mr. Preece, the eminent electrician, is represented as holding that a copper wire with a sectional area of only the one-hundredth part of a square inch is "sufficient to serve as a lightning-rod for any house." The authority upon which this startling statement is made is a passage in the "Journal of the Society of Telegraph Engineers," in which Mr. Preece says that he thinks "galvanized iron wire one quarter of an inch in diameter is sufficient for the protection of any house." It needs no very large amount of acquaintance with electrical matters to enable the reader to understand that Mr. Preece would not himself have expressed the same confidence in a small copper bell-wire such as is given as the equivalent in the table of the report. Taken in connection with the omission of all reference to the increased resistance in long conductors, it might be inferred from this estimate that Mr. Preece would hold a small copper bell-wire, carried from the golden cross of St. Paul's to the ground, to be a sufficient protection for the great metropolitan cathedral.

In his "Notes et Commentaires sur la Question des Paratonnerres," printed in 1882, Professor Melsens complains that no notice of his system of numerous conductors of weak or small section has been taken in the code of laws of the Lightning-Rod Conference of London, even as a possible alternative of construction, a silence which he interprets as equivalent to a formal condemnation. He says:

Still, I believed that the silence which the conference observes in its code of law upon the possible application of my system was equivalent to a condemnation; I should have been glad to see the conference pronounce, distinctly, without any reticence, either for or against the system as a whole, or in regard to its adoption concurrently with the lightning-rods which it prescribes or which it commends; the eminent savants who were a part of it would not have failed in that case to discuss the essentials, with great profit in the elucidation of the scientific and practical question, particularly on the points still subject to discussion, and on which we still meet very opposite opinions. I have to regret deeply, especially in consideration of the ancient savants who are members of the English commission, the silence which they have thought it their duty to keep respecting my new system of lightning-rods, while giving the regulations and laws which, according to it, secure the most efficacious protection, aside from all consideration of the constructors who advertise so largely, or who are protected by letters-patent.

The distinguished electrician of Brussels is not without good ground for this complaint, but he may console himself for his disappointment in the approval of his system that has been accorded by other highly competent authorities. In his "Report on Static Electricity and Paratonnerres" at the International Exhibition of Electricity at Paris in 1881, Professor M. E. Rousseau says:

From the comparative examination that I have made, I am convinced that in each of the three constituent parts of which the lightning-conductor is composed, namely, the point, the rod, and the root or earth contact, the system of M. Melsens has a marked superiority over the old system; and, as MM. Angot and Nardi have remarked, must be regarded as efficacious as the old system, if not more so, besides being at the same time less costly.[3]

M. Angot, the author of an able treatise on "Elementary Physics," printed in Paris in 1881, speaks of Professor Melsens's system of lightning-protection as being "more efficacious, as well as less costly, than the older plan, and sure to come soon into general use." M. Nardi, in a memoir on "The Parafulmine of Melsens," printed at Vicenza in 1881, describes the multiple system of points and rods and the large earth contacts adopted by Professor Melsens as being "the most rational, the most efficacious, the most easy to construct and fix, and the least costly of all the alternative systems of construction." M. Mascart, Professor of Physics in the College of France, in his excellent treatise on "Static Electricity," describes the entire system devised by Professor Melsens as "forming, without any doubt, the most beautiful model of the paratonnerre that has been realized." The frank and outspoken acceptance and praise of France, Italy, and Belgium may, therefore, fairly be placed as a set-off against what Professor Melsens feels to be the discourteous, if not condemnatory, silence of London.

Since the appearance of the report of the Lightning-Rod Conference a small volume has been published by "Major Arthur Parnell, of the Royal Engineers,"[4] entitled "The Action of Lightning, and the Means of defending Life and Property from its Effects." In this little book the author has been at the pains to compile a reference to a very large number of accidents that have been occasioned by lightning. This, however, has been done for an ulterior and somewhat insidious purpose. He has a new theory of his own to propound, and a revolution in the practice of lightning-rod engineering to propose. He wishes to do away altogether with the lightning-rod as a dangerous and superfluous expedient, and to establish in its place a system of earth-buried plates and short earth-points surrounding the building. Space does not here permit an allusion to the various fallacies which are involved in this heretical scheme. It will be enough for all practical purposes to say that the proper answer to the dangerous heresy is an appeal to the argument of facts. There are innumerable instances on record in which lightning has been seen to strike lightning-conductors with a luminous flash, and there are still more in which the extremity of the rod bears the traces of the passage through it of lightning; but in every case, if the rod has been of due size and properly constructed and fixed, the building associated with it has been entirely uninjured. The truth obviously is that the question of efficiency and safety entirely hangs upon the amplitude of the dimensions, the number and position of the points, and the completeness of the earth contact, of conductors. In any case where these are insufficient the lightning-rod is a source of danger. In every case where they are ample, and where the system of their establishment is sound, the protection is complete. It will be time enough to enter upon a consideration of the merits of the retrograde course which is advocated in this ill-advised scheme when any single case of failure in a lightning-conductor of satisfactory dimensions, and of tested perfection of construction, has been established before a competent jury on incontrovertible grounds. The failures incident upon defective work—as all unbiased and properly trained thinkers are aware—are among the weightiest of the arguments that tell in favor of the employment of conductors.

In a very large majority of the cases in which accidents have occurred to buildings which have been furnished with lightning-conductors, the mischief has been actually traced by competent inquiry to some easily recognized fault or deficiency of construction. A very instructive illustration of the accuracy of this remark has quite recently presented itself in a form which is worthy of notice. Shortly after midnight, on the 26th of November, during a thunder-storm of some severity, a flash of lightning struck the lightning-conductor attached to the spire of Chichester Cathedral, and scattered a considerable portion of it into fragments. A letter from "A Fellow of the Royal Astronomical Society" forthwith appeared in the "English Mechanic and World of Science," drawing attention to the accident, and commenting upon it in the following words: "This seems to open a very serious question indeed, because, if so elaborate an affair as the Chichester conductor proved so much worse than useless when a thunder-storm came, what security have we that a similar disaster may not befall at, say, the Government magazines at Purfleet or elsewhere?" In reference to the accident which called forth this note of alarm, it may be at once, however, said that it belonged essentially to the class of occurrences which have been pointed at in the beginning of this paragraph. The conductor which was attached to the spire was not adequate and competent for the protective work which it was intended to perform. It had been put up sixteen years ago, when a new spire was erected in the place of the old one, which fell in consequence of having been added as an after-thought to a tower that had not been prepared to bear its weight, and was of a form which is, happily, now obsolete. It originally consisted of twelve No. 15 gauge[5] copper wires arranged in a double series, side by side, and held together by a double strand of zinc and copper wire crossing them transversely, and acting as a kind of weft to the longitudinal copper warp. The conductor was thus a sort of ribbon of copper wire, with transverse binding-threads of zinc. The weight of the metal in this compound conductor was ten and a quarter ounces per yard, instead of being thirty-six ounces per yard, as it ought to have been at the very least if it had fulfilled the conditions that are now required for such a task as it had been required to perform. But, besides this, in consequence of having been exposed for sixteen years in its sub-littoral situation to the blasts of the moist sea-wind, the copper wires were in many places eaten into by corrosive action where the zinc wire of the woof crossed them, so as to reduce to some considerable extent their original conducting capacity. The conductor was so fixed that it descended from the summit of the spire along the slope, and along the face of the tower, then crossed the lead flashing of the roof, passed down the main wall of the building near the intersection of one of the transepts with the nave, and was finally plunged into a well dug into the grave-yard about twenty feet from the place where it reached the ground. At the time of the storm a flash of light was seen to pass along the upper part of the track of the conductor, and this flash was accompanied by an instantaneous crash of thunder, that awoke most of the slumbering inhabitants of the close. The destruction of the conductor, however, was not discovered uutil the second morning after the storm, when some shattered fragment was observed projecting from the tower. It was then found that about forty feet of the conductor at the top of the spire still remained uninjured in its place, but that for the next one hundred feet below this the woven metallic band had been scattered into a shower of short fragments of copper wire, which were strewed thickly upon the roof of the tower and of the lower building. These fragments were three quarters of an inch long, corresponded in length with the materials of the transverse crossings of the zinc wire, and bore unmistakable indications of galvanic corrosion upon their ends. The lower portion of the conductor was uninjured, but one of the iron rain-pipes, which descended from the roof of the transept a few feet away, had been shattered by the discharge. It was therefore manifest that from the leaden covering of the roof downward the incompetent conductor had been assisted in its work by the roof and its numerous iron rain-pipes, and this intelligibly accounted for its own preservation through that portion of its course; and it was also clear that the earth communication of the conductor was not ample enough for the transmission of the entire discharge, as, if it had been, the lower part of the conductor would have been shattered like the upper part, and the rain-pipe would have remained uninjured. The resistance of the earth communication of the conductor, measured through the uninjured fragment, was sixty-five ohms—that is, some twelve or sixteen times greater than under any circumstances it ought to have been. So far, therefore, from this maligned conductor being open to reproach, it had done exactly what it was scientifically bound to do, and what any expert could have foretold that it would do, under the circumstances which have been described.

But the critic who sounded the note of alarm in "The English Mechanic" was also egregiously wrong in another by no means unimportant particular. The unfairly maligned conductor had not "proved worse than useless when a thunder-storm came." As some more appreciative commentator figuratively but not inaptly remarked at the time, it had "gallantly died at its post in the efficient performance of its duty." Although the lightning-conductor was destroyed, the exceedingly beautiful stone spire remained absolutely uninjured. It had not even a scar upon its face. This circumstance of the destruction of a lightning-rod of too narrow capacity without injury to the building to which it is attached is by no means of infrequent occurrence. About five inches of the top of the second conductor which Franklin himself erected in Philadelphia were destroyed by a discharge, which was seen to strike the rod, and which also made itself visible in a luminous blaze in the dry earth around its base; and Franklin adroitly claimed the incident as a proof that Nature itself had borne testimony in favor of his invention. The brass-wire conductor of the war-ship Jupiter was struck at sea on June 13, 1854, and the sixty brass wires of which it was composed were shattered into fragments the size of a pin. But no injury was done to the vessel. A large number of instances of a kind very similar to this well-known and altogether typical case might be adduced did space permit. But it must not therefore be inferred that so desirable a result is in the proper order of events. When a lightning-rod "dies at its post" in a successful defense, as in the memorable Chichester case, the auspicious issue is due to the accidental circumstance that no better extraneous earth contact is within the striking reach of the discharge. If this were the case, the lightning would certainly be diverted from the course of the conductor into the more facile way, and, in making its devious leap into the more available path, would be quite sure to leave the marks of its divergent passage in some undesirable form. It is on this account, as well as because of the wasteful outlay which is required to supply a new rod when an old one has been destroyed, that lightning conductors of insufficient dimensions, and of bad principles of construction, are by no means to be looked upon with tolerance, to say nothing of favor, notwithstanding the occasional good service that may be entered to their account.

Irrespective of all theoretical considerations, and upon purely experimental and demonstrative grounds, it is possible in the present state of electrical science to definitely state what it is that an electrical engineer has to do when he undertakes to protect buildings against the destructive force of lightning. He has, in the first place, to make sure that, wherever the lightning can fall, it shall find an open and practically unobstructed path to traverse in its passage to the ground. He is quite sure that the electric discharge will confine itself to the track of a conductor, and will pass quietly and harmlessly along it, provided its dimensions are adequate to the task of transmission, and provided the inlets and outlets are sufficiently capacious for its unimpeded reception and escape. It is a thoroughly established and altogether indisputable canon of electrical science that when a discharge has to pass through a conductor of too narrow size, and with obstructed inlets and outlets, it, of necessity, accomplishes its passage as a turbulent and ill-regulated force all the way, with a tendency at every step to make a devious outburst or overflow; and that when it passes through a conductor of ample dimensions, and with unimpeded ingress and egress, it is devoid of all erratic impulse, and traverses the appointed channel as an obedient and well-trained power. The task of the engineer, therefore, resolves itself primarily into so arranging his apparatus as to keep the lightning in its well-ordered and harmless state so long as it is in the close neighborhood of buildings that might be injured by any uncontrolled outburst through a devious path. There are three ways in which he can seek to accomplish this purpose. He can multiply and, as it were, enlarge the gates of ingress by increasing the number of his air-terminals and earth contacts through which the discharge may have to be gathered into the conductor. He can augment the dimensions and the carrying capacity of the conductor, and he can amplify the outlets of escape, whether in the direction of the cloud or earth. Where these conditions have been properly secured, there is not the most remote probability that the conductor will fail in its appointed task. This is not a question that is now open to doubt. It is as certain that the lightning will traverse a well-arranged and competent conductor, rather than the building to which this is attached, as it is that the electric spark from the charged conductor of an electrical machine will strike a brass ball and rod, and will not strike a stick of sealing-wax or of dry wood, when these are presented side by side. As a matter of fact it is sometimes imperfectly insulated tracts of the surface of the earth that are inductively charged by the propinquity of an overhanging storm-cloud, and sometimes the overhanging cloud that is inductively charged by disturbances originating in the ground. But the conductor provided by the electrical engineer acts in precisely the same way, and with equal efficiency, in either case. It provides the means by which the electrical disturbance may set itself at rest in a quiet and unexplosive way. The chief danger that has to be feared is the purely economical one that there is always a tendency on the part of the imperfectly informed public to limit too narrowly the cost, and in that way to impair the efficacy, of the engineer's work. The duty of the engineer is, summarily, to see that his building is adequately covered above by the lines of the conducting network, that the main channel of his conductor is ample for any storm overflow that it can, by any possibility, be called upon to accommodate, and that the outlet to the earth is capacious and free. Even in the present state of electrical science it can, with the utmost confidence, be affirmed, not only that wherever destructive accidents have occurred in association with lightning-conductors, such accidents have, in every case, been due to the circumstance that the conductors have been of faulty construction, but also that in by far the greater number of instances the fault has been in the least conspicuous and least obvious part of the apparatus, where the earth contact has to be established. In his report on the lightning-conductors of the Paris International Exhibition, Professor Rousseau states that it is in this particular that lightning-rods most generally and most flagrantly fail. In one passage of the report he says:

I do not know whether I have defined with sufficient precision what is implied in a good communication with the earth, but I think the principle, at any rate, may be laid down that the communication of a lightning-conductor with the earth can not be considered good if it is inferior to that of any masses of metal that lie in its close neighborhood. If this is the case, it may be anticipated, as has so frequently been found, that the lightning will quit the paratonnerre to pass to the object which is in better communication with the earth. It is thus that buildings have been frequently set fire to by lightning which has leaped from paratonnerres to gas-pipes. In one notable case, after striking the conductor of a church in New Haven, United States, the lightning left the conductor to pierce a brick wall fifty centimetres (nearly twenty inches) thick, to get at a gas-pipe which rose twenty feet out of the ground a little distance off.

We ourselves some little time ago investigated the nature of an accident occasioned by lightning, which so strikingly confirms the views expressed by Professor Rousseau that it is worthy of being specifically brought under notice here. In the year 1865 the tower of the church of All Saints, in Nottingham, was struck by lightning during a severe thunder-storm. The tower was one hundred and fifty feet high, and had a small rope of copper wire, intended to serve as a lightning-conductor, descending along its west face from one of its corner pinnacles to the ground, where the rope terminated by being coiled round a stone buried, a few inches in the dry soil. On the inner face of the same wall of the tower, near its base, and only separated from the conductor by a solid stone wall four feet six inches thick, there was fixed a gas-standard of iron, which was used in lighting the church. The lightning in its descent left the conductor at this point, and passed through the solid mass of masonry, to reach the standard, knocking out a large circular breach in the stone-work by the way. It preferred to take this devious path, and to avail itself of the facilities which the capacious gas-main connections of the town afforded it for the accomplishment of its escape into the earth, rather than to embarrass itself with the still more onerous task of forcing its way into the dry soil at the bottom of the tower, through the too briefly terminated coil of the rope. The floor and pews of the church were found to be on fire the day after the storm, and some considerable mischief was done before the conflagration could be stopped. This fire was almost certainly due to the circumstance that the gas-pipe from the standard was connected with the meter and the mains by means of a short length of soft fusible gas-pipe in a small basement-room under the floor of the church; But, when an investigation into the cause of the fire was subsequently instituted, no one seemed to be able to say whether an escape of gas from the injured pipe had been lit up at the time of the lightning-discharge, or whether the actual lighting of the gas was due to some subsequent introduction of a burning flame into the neighborhood of the gas-meter.

The obvious method of guarding against accidents of this class is the simple expedient, wherever gas-pipes are concerned, of connecting the termination of the conductor directly, by means of a sufficiently ample metallic band, with one of the large iron pipes of the general system of the mains. If this had been done with the lower extremity of the rope, in the case of the tower of All Saints Church, instead of merely twisting it around a stone in the dry surface-soil, the injury to the wall at the bottom of the tower, and the consequent train of accidents which culminated in the burning of the floor of the church, would have been physically impossible. The lightning would then have gone through the large, open, and direct route to the mains instead of piercing a stone wall four feet six inches thick, and leaping across a small fusible gas-pipe to get there.

The case is precisely of the same nature as the accidents alluded to by Professor Rousseau. The earth communication of the copper rope being inferior to that of the neighboring gas-pipe, the lightning quitted the rope to get at the ground through the pipe. No more striking and instructive illustration of the danger of insufficient earth contacts could possibly be furnished.

A still more curious illustration of a somewhat similar kind occurred at Chichester, simultaneously with the destruction of the lightning-rod which has been already alluded to. The boundary of the cathedral close in one direction is marked by a tall and stout iron rail, which divides its precincts from the main street of the town. On the side of this street which is opposite to the cathedral stands the Dolphin, the principal hotel of the city. About an hour after the accident, and while the inmates of the hotel who had been startled by the lightning and thunder were still awake, and in some alarm, a smell of fire was perceived to be pervading the house. The landlord at once rose and proceeded to investigate the cause, and was led by the odor of burning wood to one of the cellars in the basement, where he found the small gas-pipe fixed to furnish it with light melted for several inches, a large flame issuing from the improvised gap, and a beam of wood a little above the blaze already on fire. A thorough and exhaustive examination of the place at the time, and afterward, revealed no trace anywhere else of the passage of the lightning. A water-pipe running in from the outside main, however, transversely crossed, and almost touched, the gas-pipe as this descended from the ceiling to the bracket, and just where the gap had been made. The popular notion among the servants of the hotel was that the lightning had come in through some open cracks in the cellar-door from the pavement of the street, that it had run along the water-pipe, and that it had cut through the gas-pipe as it passed across. The more scientific explanation of the insidious invasion by fire, in the dead of the night, no doubt is that, when the discharge of lightning issued from the cloud to the earth, it had scattered itself in various directions, using such stepping-stones by the way as offered in its path. One part of the discharge, then, first seizing upon the gas-pipes connected with the street lamps, took a course through them to reach the earth, but, coming opportunely by the way across the water-pipe in the cellar of the hotel, transferred itself to that pipe on account of the greater facilities that were offered by it for making an easy and good earth contact through the largely expanded subterranean mains, but "sparked" as it passed from pipe to pipe, and in doing so opened a breach in the small fusible metal wire, and lit the gas as it began to escape. The flame then enlarged the breach by melting a considerable portion of the pipe, and was making good progress toward burning down the house, when its mischievous proceedings were happily discovered, and arrested in the manner which has been described.

The telegraph-wire which, according to the opinion of Mr. Preece, may be sufficient for the protection of any house, is also, it must be remembered, capable of acting as a source of very considerable danger in circumstances that are by no means unfrequently encountered in the arrangements of every-day life. At the time of thunder-storms, portions of the electrical discharge are apt to be conveyed into the interior of buildings by telegraph and telephone wires that are distributed to them for the service of signaling-instruments, and may possibly set fire to badly conducting and inflammable substances that chance to be in connection with them. Instances of this form of accident are now often met with, especially in situations where telegraph-wires are carried to outlying post-offices over high and exposed tracts of land. In such cases it is, most generally, not the full force of the lightning-discharge which effects the mischief, but the partial and secondary discharges which take place in consequence of the influence of induction. The long stretch of insulated wire, having been inductively charged by the near approach of some storm-cloud, sympathetically discharges itself of its accumulated force when the tension of the cloud is relieved by an outburst of lightning in some other direction. The shocks occasionally experienced by telegraph-clerks when handling their instruments during the prevalence of thunder-storms in the neighborhood are due to this cause. It sometimes happens, however, that an actual discharge of lightning does involve a telegraph-wire, and such discharge is then usually distributed so that it passes to the earth in small, broken outbursts wherever it can find an outlet. In such instances enough of the fragmentary discharge may fall to the share of some signaling-office to produce very grave mischief. Telegraph-wires should, on this account, never be carried into the interior of dwelling-houses, or of inhabited places, without appropriate arrangements having been made to neutralize the risk. The plan which is most usually adopted for the protection of instruments and operators in such circumstances consists in the ingenious expedient of arranging two broad metal plates so that their contiguous surfaces be face to face a very small distance apart, one of the plates being in immediate connection with the telegraph-wire, while the other is in communication with the ground. The narrow interval between the two plates is then sufficient to prevent any escape of the ordinary electrical current of low intensity which is employed in telegraph work, but upon the occasion of the wire becoming accidentally charged with an electrical force of high intensity, such as is produced by the agency of the thunder-cloud, this leaps through the narrow space by virtue of its superior explosive power, and so escapes harmlessly to the earth, instead of making its way through some more devious and dangerous route. The plates are, of course, designedly fixed where they serve to intercept the discharge by the temptation of the more open and free passage to the earth, and in that way divert it from the dangerous course which it would otherwise pursue.

The best course for the electrical engineer, who is planning the protection of any building against lightning is, therefore, on account of the various considerations which have been urged, to begin with the arrangement of that which is the primary essential, the earth contact. In towns where there is a large system of water-supply and gas distribution at hand, this is generally an easy task. But it by no means follows that, where the main pipes of water and gas supplies are not available, a square yard of sheet-copper or iron, buried in the ground, can in all cases be accepted as a satisfactory earth connection. It certainly would not have been so in the instance of All Saints Church. In the circumstances which have been described in speaking of the accident there, a yard-square earth-plate could not have been depended upon to prevent the mischief. The lightning would still have preferred the largely developed root of the gas-mains to any such puny substitute, although such an earth-plate, well bedded in moist ground, might have served all purposes in the absence of so formidable a competitor. The condition of safety is that which has been so well stated by Professor Rousseau. The communication of the conductor with the earth must not be inferior to that of any neighboring mass of metal. When the arrangement for the earth connection has been efficiently settled, the conductor may be carried up from it, and this may with equal assurance be done either upon the single-rod system of Gay-Lussac or upon the multiple-rod principle of Professor Melsens, so long as the building is of moderate size and of a compact form. But, if the building is of large dimensions and of irregular form, the single conductor would of necessity have to assume an approximation to the multiple type, as the main stem is branched out above to bring every gable and turret and pinnacle of the structure under its protection. It is only when it has been completed by a broadly cast net of metallic meshes and lines that the old early dogma of the protected area can be now allowed to survive even in the mind of the engineer. When the work of construction has been so far carried out it is still, however, not to be looked upon as complete until the stamp of efficiency has been placed upon it by the application of the final test, which the advance of electrical science has now placed in the hands of the constructor. It is the crowning distinction of this system of defense that by a very easy process it can be at once ascertained whether all the arrangements of the engineer have been properly carried out. By the employment of the ingenious piece of apparatus which is known as the "Differential Galvanometer," the electrician can in a few minutes ascertain what the resistance is that would be offered between the air-terminal and the earth communication of a conductor, if a discharge of lightning fell upon the rod. That resistance must never be left unheeded if it amounts to anything in excess of the quantity which is technically known as two ohms. It is quite possible, indeed, by the exercise of judgment and skill, to reduce the resistance in every case somewhat below that. With a conductor which has recently been erected upon the Hall of General Assembly in Edinburgh, it was found at the final test that the earth resistance was only the 0·7 of an ohm. But the galvanometer test must not only be applied as the last step of the construction; it must also be drawn upon from time to time, and at not too distant intervals, to ascertain how far the originally well-conceived and well-executed work is or is not in process of being injuriously affected by the physical agencies that are at all times in antagonistic operation to the constructive efforts of man. The free and frequent use of the testing galvanometer is, indeed, the natural consummation of the beneficent work which was initiated by Franklin one hundred and thirty years ago. Without this instrument the lightning-conductor is a hopeful and very generally helpful expedient. But, with the galvanometer, it is now assuredly competent to take rank as a never-failing protection.—Edinburgh Review.

  1. See "Quarterly Journal of the Meteorological Society," vol. iii, p. 75.
  2. The solid copper tape which is chiefly used by Mr. Anderson is, to meet the circumstances here alluded to, manufactured of four different sizes, the smallest being 58 inch wide and 112 inch thick, and the largest 112 inch wide and 18 inch thick.
  3. Professor Melsens estimates that the cost of effective protection by the old system amounts to very nearly 412 francs the square metre, but by his system to only 0·66 of a franc the square metre.
  4. Now Colonel Parnell.
  5. That is, of one sixteenth of an inch in diameter.