Chapter XII: Appendix: F

ABSTRACTS OF PRINTED DOCUMENTS.

FRENCH OFFICIAL PUBLICATIONS.

_Preliminary Note._

In 1784, the attention of the French Government having been directed to the desirability of protecting the powder magazines of the kingdom from damage by lightning by the employment of conductors, a system of construction was proposed by two officers of the Engineers and Artillery. This system was referred by the Minister of War to the Academy of Sciences for consideration and report. From time to time subsequently other proposals of a like nature, and other inventions and improvements in the construction of lightning rods were considered by the Academy, and reported upon by various Committees.

At the request of the Conference, I have endeavoured, in the following pages, to give in as condensed a form as possible an accurate abstract of their contents, and to avoid, in all cases, the expression of any opinion, either adverse or concurrent, upon the principles or suggestions contained in them.

E. E. DYMOND.

REPORT _made to the_ ACADEMY OF SCIENCES, _by_ FRANKLIN, LEROY, COULOMB,
DE LA PLACE, _and_ ROCHON.

_24th April, 1784._

Certain proposals for erecting lightning rods for protecting the powder magazines at Marseilles having been submitted to the Academy for their opinion, a committee, consisting of the above-named, was appointed to examine and report.

They begin by enunciating the theory which should regulate the erection of conductors, and they lay down the following rules:—

1. The extent of the building should first be ascertained to decide whether one or more conductors should be used. Electrical experiments had not yet made known anything of the extent to which the action of the point of the conductor reached. But since buildings had been supplied with conductors many observations had shown that those parts of them which were more than 45 feet French (48 English) from the point of the conductor had been struck by lightning.

2. When there are many points or arrows on the building they should be connected together and also connected with all parts of the roof which are covered with lead, and also connected with the weathercocks or ornamental metal points so as to form one metallic system with the conducting bars.

3. It is not less important that these bars should be thoroughly joined together; for a solution of continuity in them produces a resistance to the passage of the electricity according to the extent of their separation.

4. It is necessary that the bars should communicate thoroughly with moist earth or, better still, with water.

As to the height of the points they should be at least 12 or 15 feet (13 to 16 feet English), or even more if the building is a large one. It is certain that the higher they are the wider the extent of their action. They should be 2 inches (2·2 English) square at the base and greater in proportion as their height exceeds 15 feet (16 English). If the conducting bars are 8 or 10 lines (or, say 1 inch) square, it will be more than enough. No case had occurred in which iron bars of this size had been in any way damaged or altered by the passage of lightning.

The reporters then proceed to examine the two proposals for protecting the powder magazines at Marseilles, sent in by M. Ravel de Puy Contal and M. Pierron. They were both for the same building which was 31 toises long and 8 toises wide (about 198 by 51 English feet). The first provided for the erection of three points on the ridge of the roof, and of four others, one at each angle of the building; the second had also three points on the ridge, but the other four were alternated on the two sides of the roof, and iron bars were carried all along and connected with all the points. The manner in which the terminals were fastened to the roof and the conducting bars fastened together and led to the water was the same in each proposal.

The reporters remark concerning the second that the conducting bars laid horizontally along the roof would involve a great and unnecessary expense, but the points should be retained, only instead of placing them alternately they should be set up so that each of them was half way between the middle and the end of the roof, and instead of connecting these points by bars along the length of the roof, they should be connected with the one connecting the three points on the ridge by bars joining it perpendicularly.

As to the method proposed for joining the several parts together the reporters cannot help thinking that in their desire to make thoroughly good connections MM. Pierron and de Ravel had proposed a plan involving too much difficulty and superfluous expense [It seems to have been proposed to screw the bars into each other], and they recommend instead of this to make at the base of each point, immediately above its insertion into the roof, a circular flange about 2 inches in diameter and 2 lines thick, with a hole half an inch in diameter in the middle and at the ends of each conducting bar to make a similar flange and to bolt the flanges together with a sheet of lead between them. Crutches should be fixed on the roof to carry the conducting bars. The points should be fixed three on the ridge and two on each side of the roof half way between the point in the middle and that at each end. These four should be connected with the conducting rods running along the ridge and should overtop the ridge by at least 6 feet (6 feet 5 inches English). By this arrangement all parts of the roof would be well protected.

The reporters highly approve of the way in which the conducting bars are connected with water by being led into the sea, but if at the other end of the building there is sufficient earth on the surface, and the soil is not entirely rock the conductor from the point placed at that end might be led into it. It is recommended that the points of copper should be screwed to the terminals for convenience of removal when necessary.

REPORT _made to the_ NATIONAL INSTITUTE, _by_ LEROY, LA PLACE, _and_
COULOMB, _on a Lightning Rod for Powder Magazines proposed by_ REGNIER.

_6 Nivose, Year 8_ (_23rd December, 1789._)

The reporters think it desirable to make some general observations on lightning rods, the rather that it appears that some persons have had fears as to the certainty of their effect.

It is impossible to reject the theory upon which Franklin had proceeded in providing lightning rods for the purpose of protecting buildings from damage by lightning. Still, as the theory needed to be confirmed by facts, it might at first have been doubted whether the lightning rods were really effectual; but now that observation, and experiment had proved the truth of the theory there was no longer any room to question their utility. It may even be remarked that observations had not only proved that they were effective when well constructed, but that they conducted the lightning down without accidents, even when they had some defects, which might have caused one to doubt their efficacy. The defects alluded to were a blunted point and a break in the continuity of the conductor. With reference to these two cases observations have shown—1st. That although the points have been blunted, they still attract the lightning from the clouds to themselves in preference to the surrounding objects. 2nd. That although the several parts of the conductor are not thoroughly joined together, the lightning will still, if the break be not too considerable, pass along the conductor without accident.

In support of the first proposition they quote the observations of Doctor Rittenhouse, of Philadelphia, who had examined several of the points in that city, and had found them melted, showing clearly that they had been struck by lightning, and probably more than once, as it had been shown by many observations that where, from local circumstances (not then fully ascertained,) lightning had struck in certain places or on certain buildings, it was not uncommon to see it strike again; and a number of observations of a different sort had shown that lightning was attracted by metals on buildings even when they were but slightly pointed, such as tin weathercocks, or iron crosses, and even plain sheets of iron.

One of the most striking examples in support of the second proposition, was the case of an American ship, reported in the Phil. Trans. for 1770. During the night, in the midst of a storm, the crew reported that there was a stream of fire in the rigging, just above the middle of the lightning conductor. The captain saw a stream of fire, sometimes in sparks, and sometimes only a steady light; and on examining the conductor next morning, found that one of the links of its chain was broken. Fortunately the two pieces, being kept in place by the fastening to the shrouds, were only about three quarters of an inch apart. These two broken ends formed a sort of points, and on its passage between them the lightning had become visible. But this was all; no shock was felt, nor anything which caused any suspicion that the fracture of the conductor had in any way hindered the passage of the lightning. Franklin also had shown by experiment that in a lightning rod where the upper end was only connected with the part entering the ground by a very fine brass wire, although the wire was melted by the passage of the lightning, it still was conducted from top to bottom without any damage to the house; and in other instances metallic wires, though partly melted by the lightning, had still served as conductors. But it is not contended from these examples that a very exact and continuous connection of all the different parts should be dispensed with.

The lightning rod proposed by Regnier consisted of a piece of wood, coated with resin, rising 2 metres (6 feet 7 inches) above the roof, and having fixed on its top a sort of inverted funnel of copper, at the upper end of which was fixed the point. To the lower edge of the funnel were fastened ropes formed of twenty-seven annealed iron wires well bound together, which were, at a suitable distance, connected with iron bars, fastened to masts, and leading to moist earth. The point had a small piece of platinum at its upper end.

The reporters observe that the wooden support may be employed by way of extra precaution, though there was no known instance of lightning leaving metal for wood; but it should be strong enough to resist the wind. They approve of the method proposed for connecting the point with the metal bars, metallic ropes being very suitable for this purpose, and keeping them well away from the building was quite right; but they add that the metallic bars should not only communicate with moist earth, but also with water in wells or otherwise.

INSTRUCTIONS FOR ERECTING LIGHTNING RODS FOR POWDER MAGAZINES, _adopted
by the_ FORTIFICATIONS COMMITTEE.

_25th August, 1807._

A lightning rod is an electrical conductor terminating in a point and carried down to the common receiver. It may be regarded as a metallic tree, and divided into (1) the upper terminal, (2) the trunk, and (3) the roots.

1. The upper terminal is a very pointed, conical or pyramidal spike of metal having a base 3 or 4 centimetres (1½ inches) in radius. The point is of gold or platinum, soldered to a copper rod 1 or 2 metres long (3 feet 3 inches to 6 feet 7 inches). This rod is joined to the rest of the upper terminal, which is of iron, either by solder, a screw, or a pin. It is important that all the parts of the upper terminal should be joined with care so as to prevent fracture; at the bottom of the terminal are several feet by which it can be leaded to the vault or bolted to the framing of the roof. Several devices for giving some play to the terminal so as to diminish the effect of vibration have been proposed, but it is better to make the terminal strong enough to resist. At the bottom of the terminal is joined the piece connecting with the conductor; this ought to be very complete and continuous, especially at the point of junction with the terminal. Frequently the terminal is enlarged at this point to facilitate the passage of the lightning. To preserve the terminal from rust it is sometimes gilded—it has been proposed to tin it—more frequently it is merely painted; experience shows that this is sufficient. Instead of making the whole terminal conical or pyramidal, a square bar of iron, finished with a point of copper tipped with gold or platinum, is sometimes used. This plan may usually be adopted without danger, but they are more liable to be broken or bent by vibration.

FIG: 1.
]

FIG: 2.
]

FIG: 3.
]

FIG: 4.
]

2. The trunk or conductor is made of iron bars 13 to 20 millimetres square (½ to ¾ inch) notched at the ends and bolted together with a plate of lead between the two (Fig. 1). For powder magazines a bar of 27 millimetres (1 inch) square is recommended. They follow the outline of the roof, cornice, and wall, and each bar is fixed by a half collar (Fig. 2) or cramp placed in the middle of the bar or as far as possible from the junction of two bars. Instead of the iron bars ropes of copper or iron wire, or even of hemp, may be used; these last may be used provisionally, but for permanent conductors they have no advantage either in economy or conductivity. The copper rope conducts the lightning better, but its smaller size and cylindrical form, by diminishing its absolute and relative surface, counterbalances its superior conductivity. The great and real advantage of metallic, and especially of copper ropes, is in their continuity and their flexibility. The conductor is led down to the surface of the ground where it is bent and led parallel to the surface towards a pit full of water, or deep enough to allow the end of the conductor to rest in damp earth; from 2 metres (6 feet 6 inches) above the ground to the pits the conductor is enclosed in a channel or trough like the fuse of a mine, the object of this is to protect the conductor from the dampness of the soil and from contacts. These would be unimportant so long as there is a perfect connection between the point of the conductor and the common reservoir, but this continuity may be destroyed by degradation of the conductor, and it is chiefly at the joints that this discontinuity is to be feared. When the conductor has to be buried it should be in an oaken trough, well put together and tarred or charred or surrounded by powdered charcoal so that the metal cannot be rusted by infiltrations or humidity; in some soils it is better to make the subterranean part of the conductor of lead, taking care by increasing the surface to make up for its inferior conductivity. Sometimes water pipes may be made use of, but only when they serve to lead water away and when they terminate in an isolated reservoir. It is important to lead the conductor far away from water pipes carrying water to public fountains or into the interior of houses.

3. If the conductor leads to a well full of water the roots (Fig. 3) need not be more than a few spindles terminated in points and long enough to be always immersed. When the conductor only leads to a bed of earth it is supplied with a system of roots (Fig. 4), having for its object the multiplication of the points for the escape of the lightning, and these are increased in number according as the soil is a less good conductor. The pits should be some distance from the foundations of the building, so that the lightning may not damage them, and it is important, by all possible means, to increase the natural humidity of the soil. When the wells cannot be closed it is necessary that the conductor should be insulated and plunged deeply in the water for fear that the communication of the electricity to the well-chains or pump-rods might cause accidents or alarm. After some other instructions it is added that the dispersal of the electricity in the common reservoir is, next to the continuity of the conductor, that which most deserves the attention of the physicist and the engineer.

It has been remarked that a point extended its sphere of activity as far as 10 metres (32 feet 9 inches), that beyond this distance its effect became less sensible, and that when the points were too near together they neutralised one another. So upon a building of a given size it is necessary to set up so many that all parts shall be covered by their spheres of attraction, which should meet and not overlap each other. Lightning in passing from a cloud to the earth does not always take a vertical direction, it sometimes follows the path of the rain drops, which is inclined by the wind, so when a magazine is very lofty, or on an elevated spot it is not useless to fix horizontal or inclined points on the gables or angles. In some places the magazines are dominated by other buildings; in these cases the neighbouring buildings should be protected, or the magazines should have horizontal points towards them. If the ramparts dominate the magazine it will be prudent to set upon them a lightning rod on a mast. Trees are only struck by lightning because their tops serve as points but their trunks are bad conductors, hence it is prudent not to have plantations, especially of lofty trees near magazines. However many points may be set up on a magazine they should all be connected together, and all joined to the principal conductor, and it would be well to have more than one principal conductor so that if one loses its continuity the lightning may have a path by the other. Stone, wood, and gunpowder are bad conductors, and pieces of metal may without danger be used in the inside of magazines, provided they are connected with the principal conductor by branch conductors of suitable size: still it is prudent to keep the metal outside.

Reference is then made to “Regnier’s System of Lightning Rods,” Appendix F., p. 53, which is thought to be much too expensive.

REPORT _on the foregoing Instructions made by_ LA PLACE, ROCHON,
CHARLES, MONTGOLFIER, _and_ GAY LUSSAC _to the_ NATIONAL INSTITUTE.

_2nd November, 1807._

The reporters say that experience has taught that the point of a lightning rod 4 or 5 metres (13 to 16⅓ feet) does not effectually protect a space round it greater than one having a radius of 10 to 12 metres (32¾ to 39¼ feet). That when there are points or considerable masses of metal on a building having a lightning rod it is absolutely necessary to connect them by branches with the principal conductor. That it is not less important that the metallic bars should be thoroughly well connected together so that the electricity may find no resistance in its path from the point to the common receiver. And lastly, that it is necessary that the conductor should have a perfect communication with moist earth, or better, with water. They then proceed to discuss the instructions, or that part of them which relates to the construction of the lightning rods. They recommend the use of gilded copper points, notwithstanding the doubt concerning them which had been raised in consequence of their deterioration by oxidation, and their being blunted by lightning. They say that experience has shown that an iron rod 20 millimetres (·8 inch) square is more than sufficient to carry the most violent discharge of lightning, and that it is consequently needless to make them larger, as recommended in the Instructions; that it is only at the joints that there is any cause for fear because, in spite of the insertion of the piece of lead, the contact is not perfect; that it would be easy by enlarging the bars at their junctions to increase the number of points of contact, and by lengthening the bars to make fewer joints. That in this respect the use of iron wire ropes would be very advantageous, but they fear that the ropes would be easily destroyed, and that the use of copper wire rope instead of iron would be too expensive.

When the conductor reaches the ground too much care cannot be exercised in making a free communication between it and the soil. It is upon this that its good effect principally depends, for houses have been struck although provided with a conductor, because it only communicated with a very dry soil. M. Patterson, of Philadelphia, in the fourth volume of the American Phil. Trans., has published a means of making a good contact which seems useful. He proposes to lay the conductor in a bed of galena worked into a paste with melted sulphur. The galena is a good conductor, and would have the advantage of protecting the iron from the damp. He has also proposed a simple means of providing for the easy dispersion of the electric fluid in cases where the soil is not very damp, which consists in making a hole in the ground and filling it with charcoal, into which the conductor is plunged. But M. Guyton used the conducting power of charcoal for this purpose more than thirty years ago, and it has been applied in many ways. Charcoal, like galena, is a good conductor, and this property renders its employment desirable in cases where the soil is dry.

Upon the proposal to fix inclined or horizontal points they think that vertical points will suffice; and with reference to the Regnier system, they remark that it would certainly be very expensive, and that it would not be necessary to adopt it until the usual system had been found insufficient.

INSTRUCTIONS _about_ LIGHTNING RODS _adopted by the_ ACADEMY OF
SCIENCES.

_First Part, 23rd April, 1823._

PREPARED BY A COMMITTEE CONSISTING OF MM. POISSON, LEFEVRE-GINEAU,
GIRARD, DULONG, FRESNEL, AND GAY LUSSAC.

After some theoretical remarks the Committee describe the conductor they recommend, giving the name of _tige_ (upper terminal) to the part rising into the air above the roof, and that of conductor to that part extending from the upper terminal to the ground. The upper terminal is a square or round bar of iron tapering from base to summit. If from 7 to 9 metres (23 feet to 29 feet 6 inches) high, which is the smallest height to be used on large buildings, it should be 54 to 60 millimetres (2·1 to 2·3 inches) square or diameter at the base, if 10 metres (32 feet 9 inches) high, it should be 63 millimetres (2·5 inches). About fifty-five centimetres (1 foot 9½ inches) of the upper end is cut off and replaced by a point of copper either gilded at the end or tipped with a little piece of platinum. At the lower end of the terminal (A), 8 centimetres (3·15 inches) above the roof, is fixed a base (B) to throw off the rain which would run down the terminal, and above this base the terminal is clasped by a collar (C), as shown in the drawing, to which is bolted the conductor (D). The engraving shows the modification of the arrangement as adapted to both round and square terminals. The conductor is a bar of iron 15 to 20 millimetres (·59 to ·79 inches) square, joined firmly to the upper terminal by bolting it tightly between the two ears of the collar. The best way of joining the bars together is shown in figure 1, p. 55. It is to be held up at a distance of 12 to 15 centimetres (4·7 to 5·9 inches) from the roof by crutches, and to be kept at a like distance from the walls of the building. At 50 or 55 centimetres (19·6 to 21·6 inches) below the surface it is turned away perpendicularly from the wall for a distance of 4 or 5 metres (13 feet 1 inch to 16 feet 5 inches) if it does not sooner meet with water. To avoid rusting the rod is carried in a trench filled with charcoal, and then turned down a well so as to have at least 65 centimetres (25·7 inches) in the water when at its lowest level, where it terminates in three or four branches to facilitate the exit of the electricity from the conductor.

If there is no well convenient, a pit should be made 13 to 16 centimetres (5·1 to 6·3 inches) in diameter, and 3 to 5 metres (9 feet 10 inches to 16 feet 4 inches) deep, down the middle of which the conductor should be led and the hole filled with charcoal tightly rammed. As the iron bars forming the conductor are not easily bent to follow the lines of the building a metallic rope may be used. It is made of four strands, each composed of 15 iron wires, and forming a rope of 16 or 18 millimetres (·62 to ·7 inches) in diameter. Each strand is tarred separately, and the whole also well tarred when put together. It is attached to the upper terminal in the same way as the bars by pinching between the ears of the collar (c). At 2 metres (6 feet 7 inches) above the ground it is joined to the bars which form the earth connection by being pinned into a socket formed at the end of the first bar. Ropes of copper or brass wire may be used, and they need not be more than 16 millimetres (·62 inches) in diameter.

It is necessary to connect any considerable metallic masses (lead roofs, metal gutters, or tie rods) with the conductor, because if this be not done, and the conductor be broken, or have a bad earth connection, the lightning may leave the conductor for the metallic mass.

Modifications of this form of conductor for use on churches, ships, and powder magazines (for the latter carrying the conductors on masts is recommended) are then described.

The report says that the terminal of a conductor protects efficiently a circular space round its base, having a radius equal to twice its height; but that it is prudent to estimate that a conductor on a church spire only protects a circle having a radius equal to the height of the conductor.

The conductor should go the shortest way to earth. It should be on the side most exposed to the weather, especially on spires.

_Second Part, 18th December, 1854._
PREPARED BY A COMMITTEE CONSISTING OF MM. BECQUEREL, BABINET, DUHAMEL,
DESPRETZ, CAGNARD DE LATOUR, AND POUILLET.

Notwithstanding the considerable advance in knowledge since 1823, the instructions of that date have no need to be altered, at least in their essential principles; but the methods of construction of buildings having materially altered, and metal having largely replaced wood and stone, buildings had, so to speak, become metallic masses, which would have incomparably greater attraction for thunder clouds. The Palais d’industrie in the Champs Elysees, for example, nearly 3 hectares (7·4 acres) in extent, and 40 metres (131 feet) in height had everywhere enormous masses of iron, brass, and zinc.

The company undertaking the building had sought the advice of the Academy as to the means to be employed to protect it from lightning, and it had been found necessary to revise the instructions of 1823, in order to introduce such modifications as were necessary.

Quoting the passage referring to the connecting of metallic masses with the conductor, the Committee think that the time had come to enter into fuller details on this point.

Formerly the use of metal was almost restricted to ridges, gutters, and tie rods; now metal was used everywhere, and what is important, in large surfaces and great masses; and this new system realised on a large scale the first objection to lightning rods—it attracts the lightning.

When this objection was applied to lightning rods, it had only the appearance of truth, but when applied to the masses of metal then used in buildings, it was not only specious, but true, and founded upon well established laws; these buildings do attract the lightning, and render its effects more disastrous.

In the case of two buildings alike in size and shape, situated on the same soil, one made of wood and stone as formerly, the other with much metal as now, and both without lightning rods—if the conditions are such that the lightning must discharge itself, it will always strike the latter, and never the former; in the same way as on bringing to the conductor of an electrical machine a ball of wood or stone, and one of metal, it is always the latter which will receive the spark. Lightning rods, therefore, are so much the more indispensable as the buildings contain greater surfaces and greater masses of metal.

The nature of the soil must be taken into account, as well as the buildings and other objects upon it. A dry soil, with a subsoil of dry sand, chalk, or granite, does not attract the lightning, because it is a bad conductor. Unless when accidentally wetted the buildings on it participate to some extent in this immunity, at least if they are not built in the modern style, and are not very large. But if there are at a moderate depth underneath this dry ground, large metallic veins, vast caverns, sheets of water, or only abundant springs—these will attract the lightning, which will destroy everything in its path unless protected. If the wet or metallic strata are very deep, the danger of an explosion is diminished by the difficulty of passing the intervening envelope, and by the weakening of the action of the cloud by the increase of distance.

On the 19th April, 1827, the packet boat _New York_ was twice struck by lightning. On the first occasion, having no conductor, it received considerable damage; on the second, the conductor was fixed; it was made of a pointed bar of iron, 1·2 metre (about 4 feet) long, and 11 millimetres (·43 inch) diameter at the base, and a surveyor’s chain about 40 metres (131 feet) long, forming a connection between the foot of the rod and the sea; the chain was made of iron wire 6 millimetres (·24 inch) in diameter; the links were 45 centimetres (17·7 inches) long, ending in loops, and joined together by two round rings. When struck the chain was dispersed in burning fragments and globules, which set the deck on fire in many places, notwithstanding the hail upon it and the rain which fell heavily; the bar at the top was melted for a length of 30 centimetres (11·8 inches) from the point, and down to a diameter of 6 millimetres (·2 inches). The rest of the rod remained with about 8 centimetres (3·1 inches) of the chain attached to it, the longest piece of chain found was less than 1 metre (3 feet 3 inches) long, and was blistered as by fire.

On the 13th June, 1854, the _Jupiter_ was struck by lightning. The conductors were in place; that of the mainmast which was struck went 2 metres (6 feet 6 inches) into the sea, and had at its end a ball 2 kilos in weight. After being struck the conductor had disappeared and the pieces of it were scattered everywhere. The conductor, about 70 metres (230 feet) long, was a cable of three strands formed of sixty brass wires, each one half or two-thirds of a millimetre (·019 or ·026 inches) thick. The cable was mostly in bits no bigger than pins, but there were some pieces a few decimetres long, these had been turned violet colour as by fire, and those first touched were still burning hot.

These two examples show that a conductor may be destroyed, but they also show that it is not useless even then, since it will have received the discharge and directed it, and so prevented greater mischief. The _Jupiter_ received no damage; whilst not far off, a Turkish vessel, which also had a conductor (but the chain of which did not reach the water) having been struck by lightning in the same storm, had a hole more than 30 centimetres (11·8 inches) deep, and almost such as would have been made by a cannon ball, in her side just above the copper, and near the water line.

The question is, are such accidents to conductors inevitable, or are they the result of faulty construction? All the facts established in the accounts of lightning and its phenomena, leave no doubt on this point. All the lightning rods which have been destroyed were of bad materials, insufficient, badly constructed, not in accordance with the principles which theory has deduced from experience.

The conductor of the _New York_ had several faults; its upper terminal was too small, and too much drawn out; its conductor had much too small a sectional area; and the use of a chain in such cases should be strictly excluded.

There is no example known in which lightning has been able to melt iron rods 2 centimetres (·78 inch) in diameter, or 3 square centimetres (1·18 inch) in section; and copper may be used in still smaller sizes.

The conductor of the _Jupiter_, although better than the former, had also a radical defect. The fragments of the conductor which were examined bore but few traces of fusion, and none of these traces extended to the entire thickness of the cable; they were also limited to a group of some of the sixty wires of which it was composed. This seemed to show that the discharge was not carried equally by all the wires, and that those wires which it followed being insufficient to carry it, were the ones melted, and the others were broken or volatilised with explosion. Hence the breaking of the cable and dispersion of fragments of some decimetres in length, which, though too hot to be touched, were not hot enough to set wood on fire. This explanation, however, raises a singular question, whether, in a cable of similar wires twisted and bound together, the lightning can choose some wires in preference to the rest, even when the whole of them are hardly sufficient to give it a free passage.

Undoubtedly, yes; at any rate under certain conditions. No doubt if at both ends of the cable, for the length of a decimetre, the wires first tinned separately are afterwards soldered together, so as to make a sort of metallic cylinder, electricity, whether natural or artificial, having to pass along the cable, will not show a preference for one wire over another; but where this is not done—if at the two ends, or, more generally, at the two points of junction with other conductors, the wires are isolated by layers of dust or oxide—if, in addition, the cable only touches the terminals by its outside wires, then things happen very differently. The electricity takes those wires that are in contact with the terminal; these reduced to few in number become incapable of carrying it; and the whole cable broken by the explosion exhibits the phenomena shown in the case of the _Jupiter_.

The deficiency in each case was due to one cause—insufficiency of sectional area. In the first case the insufficiency is apparent, the iron wires 6 millimetres (·24 inch) thick were nine or ten times too small; in the second, the insufficiency is more hidden, it results from badly made junctions.

The two most fundamental rules for the construction of the rod and conductors are—1st. That they shall have a sufficient sectional area. 2nd. That they shall be continuous and without a break from the point of the upper terminal to the common receiver (the earth). But this continuity may in strictness be interpreted in two ways: it may be said that two pieces of metal in contact form a sufficiently continuous connection; and it may be said, on the other hand, that most frequently this simple contact is no more than a break in consequence of oxidation and the interposition of foreign bodies.

The instruction of 1823, without adopting the first interpretation, does not appear to have sufficiently recommended the second, which should exclusively regulate all construction of lightning rods. No doubt it is possible, by taking great care, to join and bolt together two pieces of iron or copper closely enough to make a practically continuous conductor, but when there are many joints we fear that evil might arise from the negligence of workmen, and still more from the chemical alteration of the surfaces, the deposition of foreign matter, and the mechanical dislocation produced by time and repeated shocks.

Hence, the three following practical rules should always be observed:—1. To reduce as much as possible the number of the joints. 2. To make all the joints with hard solder, and they should be upon surfaces of at least 10 centimetres (3·9 inches) square, and further strengthened by straps and bolts. 3. Not to make the upper terminal so gradually pointed as usual. The upper terminal of iron should be not less than 2 centimetres (·78 inch) diameter, the end should be filed down and a screw tapped 1 centimetre (·39 inch) high and 1 centimetre diameter, and to this a cone of platinum 2 centimetres diameter and 4 centimetres (1·5 inch) high, and consequently having an angle at the point of 28° or 30° should be fitted, screwed, and carefully soldered.

In other respects the instructions of 1823 should be followed; no fact which leads to a modification of the general rules there proposed 1, for the sectional area of the conductors; 2, for the method of fastening to buildings; 3, for the method of making the earth connection, has since come to light.

The subject, however, is not exhausted, there still remains the important and difficult question: what is the circle of protection afforded by a well constructed lightning rod? The opinion generally received at the end of the last century was that the circle of protection had a radius of twice the height of the terminal, and the instruction of 1823 adopted this opinion, but with some restrictions as in the case of spires. It is important to remember that these rules rest upon a more or less arbitrary basis, and this is said not to condemn them, but only to prevent there being attached to them a value which they do not possess.

More observations are required, and it is only with reserve that these rules are admitted. They are neither general nor absolute, they depend upon a variety of circumstances, and especially on the materials of the buildings. For example, the radius of the circle of protection, which would be sufficient for a building having only wood tiles or slate on its upper portion, would not be sufficient for a building in which the covering or the framing of the roof was of metal. In the former case the active portion of the thunder cloud, although further from the lightning rod than from the roof, would exert a greater action on the rod, whilst in the latter the action on the rod and on the roof would be almost equal at an equal distance.

A special note upon ships, and another on the Palais de l’Exposition close the report.

SPECIAL REPORT FOR THE NEW BUILDINGS OF THE LOUVRE, 18 DECEMBER, 1854,
BY THE SAME COMMITTEE.

Referring to the subject of the earth connection the Committee say; in the earliest instructions, it is said that the conductors should communicate with the water in a river, a pond, or wells, or at least with moist earth. This rule, although quite correct in itself, frequently leads to erroneous practice. It is sometimes thought that lightning is extinguished by water, as fire is; and when water is scarce the conductors are plunged into a well-cemented cistern. This is a most dangerous mistake; the conductor should be in connection with the common receiver, that is, the great water-bearing strata (nappes d’eau,) of much greater extent than the thunder cloud. At other times where wells are possible but costly, advantage is taken of the alternative allowed by the instructions. Instead of wells the conductors are put in connection with the earth, without being careful to see that it preserves sufficient moisture in times of drought when storms are most to be expected, and without being careful to see that the moist connection is sufficiently large. They specially note this latter error, as it appears to be still more common than the former. They do not hesitate to say that recourse should never be had to this method of connection with the common receiver. They recommend that in default of rivers or very large ponds, the conductor should always be connected by large surfaces with the inexhaustible subterranean water-bearing strata.

Secondly, where these strata are at a moderate depth below the surface, the Committee consider it necessary to make use of a conductor with two branches, the principal to descend to the subterranean water; the secondary, leaving it at the ground level, is put in connection with the surface. And for this reason; after great droughts thunder clouds exert but a feeble influence upon a dry, badly conducting soil. All their energy is felt by the subterranean waters; and the electricity will be carried by the principal branch. On the other hand, after a summer shower, when the surface soil gets moist, it is at once made a good conductor. It is that which is affected by the thunder cloud: while, at the same time, it screens the subterranean water from electrical influence. In such a case it is indispensable that the surface of the ground should be in direct connection with the conductor; and this the secondary branch supplies.

There is a final question how the conductors should be connected with the various metallic portions of the building. The ridges are throughout of iron; but the interior arrangements require that, in some portions of the building, there should be, properly speaking, only one floor, whilst in other parts there are six. Each floor may be regarded as a great metallic network, composed of several strong plate girders, crossed by numerous joists analagous to rails, while these are, in their turn, crossed by a multitude of smaller iron rods; and the meshes of this network are filled with tiles. In enquiring into the effect of a thunder storm upon those portions where there are six such floors one above the other, it is easy to see that if the roof were a great continuous sheet of metal, it would take up the whole electrical energy of the cloud, at any rate, as far as the floors underneath it are concerned. In this case it would be amply sufficient if the covering were well connected with the lightning rods. But in this case the roof is metallic, only in a very small portion; it may be said that the ridges only form a network with very large meshes, and, consequently, is an insufficient shield, through which the upper floor may still receive a considerable shock. Therefore the Committee propose the following arrangements:—1st. The principal pieces of each floor should be put in connection with the conductor. 2nd. It is very desirable that all the joists of the upper floors should be connected together by a rod bolted, and, if possible, soldered to each, which rod should be connected with the conductors. 3rd. It seems probable that, in general, the roof frames are in good connection with each other, and, consequently, it would suffice if all the upper terminals are connected with them. If, however, it happens either by changes of level in the gutters, or from other causes that the connections become doubtful special iron connections must be made. 4th. The zinc gutters and ridges should be connected with the lightning rods.

REPORT _on the points of upper terminals made by Messrs. Delieul, by a
Committee consisting of_ MM. BECQUEREL, BABINET, DUHAMEL, DESPRETZ,
CAGNIARD DE LATOUR, REGNAULT, DE SENARMONT, _and_ POUILLET.

_5th March, 1855._

The committee examined the points presented by Messrs. Delieul, one of platinum, made exactly as described in the report of the previous 18th December; the other, a cone similar in form, size, and external appearance, but rather less costly, being made of a cap of platinum, fixed with hard solder upon the conical end of the iron rod. It was thought that this second arrangement would not practically be inferior to the other; but it must be made by a skilful workman, who knows how to insure that the solder should take to the whole of the surfaces brought together. They see no objection to the substitution of palladium, or gold or silver of a standard of ·950 for the platinum. But all these metals are costly; few workmen know how to work in them, or at least to employ that precision, and take that minute care, which are indispensable to success. These reasons have raised again a proposition that was discussed in the former commission, which consists in making the points of copper. The copper point is 2 centimetres (·78 inch) in diameter, like the upper part of the iron rod, to which it is screwed and brazed; its length is about 20 centimetres (7·87 inches), and it terminates in a cone 3 or 4 centimetres (1·1 or 1·5 inches) high.

They see no reason why this should not be used with almost the same confidence as the preceding forms. If there is ground to fear that it may undergo changes from atmospheric influences, this is counterbalanced by certain advantages. 1st, copper is with palladium, gold, and silver, among the best conductors of heat and electricity; and the point of the cone will be much less heated than the platinum point; and 2nd, the terminal, with a copper point, is much less expensive, and can be made everywhere.

On the report being put to the vote M. Despretz could not approve the proposal to employ copper points, fearing that the deposition of carbonate or some other badly conducting matter would diminish the efficacy of the lightning rod.

INSTRUCTIONS _upon_ LIGHTNING RODS _for_ POWDER MAGAZINES, _by a
Committee consisting of_ MM. BECQUEREL, BABINET, DUHAMEL, FIZEAU, EDM
BECQUEREL, REGNAULT, _le Maréchal_ VAILLANT, _and_ POUILLET.

_14th January, 1867._

After referring to some general principles, and to the construction of lightning rods recommended in the reports of the earlier Committees: the Committee recommend, that the upper terminal including the copper point should be from 3 to 5 metres (9 feet 10 inches to 16 feet 5 inches) high; that the junction of the conductor and the upper terminal, and also the several joints of the conductor, should be covered with solder, and insist very strongly upon the necessity of communication with the _nappe d’eau souterraine_, which they define as “the water level in neighbouring wells which never dry up, and which retain at least 50 centimetres (19·68 inches) in depth of water in the most unfavourable seasons.”

The special arrangements to be adopted in setting up lightning rods for powder magazines are: not to fix them on the building itself but outside the surrounding walls. For each large sized magazine (27·89 metres, by 20 metres, and 11 metres high, equal to 91 feet 6 inches by 65 feet 7 inches, and 36 feet high) there should be three conductors—two near the ends of the long side of the enclosing wall most exposed to storms, and the third in the middle of the opposite side. The upper terminals should be only 5 metres (16 feet 5 inches) high, and should be raised on a pier, a mast, or other support 15 metres (49 feet 2 inches) high, down which the conductor should be led to the ground. There should be a circuit which the Committee call _circuit de ceinture_ carried entirely round the enclosing wall to which each conductor should be joined, and a conductor should be carried from the most convenient point of this circuit to the underground water. For middle sized magazines two terminals and supports, and for small magazines one terminal and support will suffice; but in all cases there should be a _circuit de ceinture_. This need not be deep below the surface, nor covered over; it may even be in an open gutter, but a conductor must be led from it to the underground water, even if in order to do this it is necessary to carry the conductor several hundred metres or several kilometres. It need not, however, be made of bars and carried all the way in a trench, but it may be made of six wires 6 or 7 millimetres (about ·25 inches) in diameter, and carried on posts like telegraph wires, except that they need not be insulated.

INSTRUCTIONS _by the Committee consisting of_ MM. ALPHAND, BELGRAND,
FIZEAU, _Comte du_ MONCEL, ED. BECQUEREL, DESAINS, CH. SAINTE
CLAIRE-DEVILLE, DUC, BALLU, MAGNE, DAVIOUD, FELIX LUCAS, _and_ R.
FRANCISQUE MICHEL, _appointed to inspect the_ LIGHTNING RODS _on the_
MUNICIPAL BUILDINGS _of_ PARIS.

_20th May, 1875._

The Committee find that platinum tips are useless, and recommend instead that the point of the terminal should be made of pure copper, 50 centimetres (19·7 inches) long, and terminating in a cone, forming an angle of 30°. This should be scarfed, pinned, and soldered to the end of the terminal. The terminal should be of wrought iron in one length, and where possible galvanized; but on no account painted. The connection with the conductor should be by a piece fitted and bolted; and, lastly, the whole joint should be well covered with solder.

The Committee consider that on an ordinary building a terminal will effectively protect a cone, having the point for its apex, and a base whose radius is 1·75 of its height. But in practice the terminals may be much farther apart, if there is a _circuit des faites_. This is defined as a metallic conductor, which extends without break over the ridges of all the buildings which it is intended to protect, and which is joined by metallic contact to all the upper terminals and to the conductor, and consequently to the underground water which alone forms the common reservoir. All pieces of metal of any considerable size should be connected with the conductor.

If the conductor is made of iron bars, they should be galvanized if possible, and the joints should be fitted, bolted, and finally covered with solder. If the bars cannot be galvanized, they should be well painted. The Committee recommend the employment, especially in the _circuit des faites_, of an arrangement for compensating for the lengthening and shortening of the bars by the variations of temperature. This is made by inserting in the circuit a curved band of copper which will yield to the movement of the rods. If the conductor is made of galvanized iron wire rope, each wire should be 2·5 or 3 millimetres (·09 to ·11 inch) in diameter, and there should be such a number of them that the sum of their sectional areas shall be equal to one-fifth more than that of a bar of iron 20 millimetres (·78 inch) square. The rope should be all in one piece, and the joints with the terminal and earth connection should be covered with solder.

The supports should not be insulated, and there should be as few as possible of them. At the underground end of the conductor should be fixed a large sheet or hollow cylinder of metal, and this should be always, even in the greatest droughts, plunged at least 1 metre (3 feet 3 inches) into the subterranean water. If from any cause this water cannot be reached, the conductor may be joined to one of the main water-pipes of the city; but if the conductor cannot be led either to the subterranean water or to a main water-pipe, no lightning-rod should be erected. It would do more harm than good.

In the case of buildings of any importance, two or more conductors leading to the subterranean water should be employed. It should be so arranged that the underground part and the earth connection may be easily inspected and cleaned from rust, and the whole should be inspected and cleaned at least once a year, at the end of the autumn. The Committee is of opinion that it would be better to put all the lightning-rod work in the hands of special workmen, under the control of an agent appointed by the administration, and not to trust it to the blacksmiths and locksmiths usually employed. The Committee lastly recommend that they should be permanently appointed, and meet every year after the inspection, to report and decide upon the steps to be taken to remedy any defects that may be discovered.

REPORT _by the joint Secretary_ (FRANCISQUE MICHEL) _of the Lightning
Rod Committee to the Prefect of the Department of the Seine_.

This report gives a detailed description of the state of the lightning rods attached to the public buildings of Paris.

In most cases the upper terminals were of great length, some of them as much as 9 metres (nearly 30 feet) in height; the conductors were in almost all cases of iron, either in bars or wire rope; the earth connections were of various kinds and extent.

The report frequently states that the points were blunted; that the upper terminals and the conducting rods were deeply rusted; that especially at the joints the conductors were seriously deficient; and that the underground portion was greatly deteriorated by rust.

A description is given of an accident from lightning to the church of St. Sulpice; but this building had no lightning rod. In the case of the church of St. Clotilde there are five upper terminals, two on the two spires, the remaining three along the ridge of the main roof. The building was amply protected as far as its length was concerned, but the transept was not so thoroughly protected. The five terminals were joined to a conductor which went round the building, and was connected with the ground. A second conductor led from one of the terminals to the ground, where it terminated in a second pit. The conductors were made of iron rods 18 millimetres (·71 inch) in diameter, joined by collars and pinned and the whole covered with paint. They terminated in distributors plunged in the underground water in walled pits. They were supported by insulated collars. The building has an iron roof. The church had been struck by lightning at least four times since the lightning rods had been erected. The first time, twelve years ago, the lightning struck the rod placed on the transept, and carried away the platinum tip of the copper point. Since then the rod has received another discharge, and the copper point is bent to the S.W. In January, 1872 or 1873, the lightning struck the western tower, and shattered one of the stones above one of the windows of the staircase.

“One of the platinum tips is gone, and many are blunted. The conductivity of the conductor is very bad, and the joints are very much damaged: hence the accident to the tower. The greater number of the glass insulators are broken, or gone altogether.”

In the case of the church of St. Eloi, which had one terminal on the spire, one conductor, formed of iron wire rope 2 centimetres (·78 inch) in diameter, joined at 3 metres (9 feet 10 inches) above the ground to an iron rod 25 millimetres (·97 inch) in diameter, which entered the ground and ended without branches in a pit filled with charcoal. The soil was dry and calcareous. The conductor was made up of many lengths of rope, old pieces apparently having been used; the joints were in bad condition, and needed soldering. The underground part was deeply rusted.

“In September, 1874, lightning struck the spire, twisted the conductor, broke the terminal, threw down the part above the cross, and made great cracks in the apse.”

During the building of the Mairie of the 20th Arrondissement, the lightning struck a fir-pole in the scaffolding. It did not do any damage, being carried away by the chain attached to the pole, from which it took all the rust, and being thence conducted by some pieces of iron roof framing lying on the ground.

There are several other accounts of accidents, but they are mostly represented by the foregoing examples.

INSTRUCTIONS _as to the application of_ LIGHTNING CONDUCTORS _for
protection of_ POWDER MAGAZINES, ETC.

Issued with Army Circulars, dated May 1st, 1875.

(_Abstracted by Prof. T. Hayter Lewis, F.S.A._)

1. The principles adopted by Sir W. S. Harris, as shown in the Appendices A and B, to this paper, still held to be sound.

2. The terminating plane of action of lightning is sometimes beneath the surface of earth, which, if moist, forms good medium for diffusion of electricity.

3. Dry soil is to be regarded as non-conducting matter.

4. Therefore conductor to be taken into soil permanently damp.

5–6. Underground magazines are usually in dry soil, and should

therefore be fitted with conductors as in the case of similar magazines above ground.

8–9. Casemated batteries of modern construction, with magazines in basement should have conductors on the parapet or terreplein from end to end of battery, attached to vertical conductor into earth. Flagstaff should have conductor. In large works there should be several points 5 feet above top of building. Iron verandahs and railings are good conductors when with good earth connections.

10. Iron buildings are good conductors. But if covered with asphalte, concrete, &c., rods or points must be provided projecting above asphalte, &c., and with good earth connections. Iron shields should be connected with conductors.

11. Copper is recommended as best conductor; it is not liable to corrosion, and very durable.

12. But if exposed to injury, or likely to be stolen or corroded, copper may be replaced by iron, provision being made for its smaller conductivity—viz., ⅕th that of copper.

13. Copper rods to be ½ inch diameter; copper tubes to be ⅝ × ⅛ inch thick; copper bands to be 1½ × ⅛ inch thick.

14. If the conductor be of iron, solid rods to be 1 inch diameter; solid bands to be 2 inches wide × ⅜ inch.

15–16. The fusing temperature of copper is 1994° Faht.; whereas that of iron is 2786° Faht. So far there is a marked advantage over copper. But it rusts easily, and then the electrical resistance is immensely increased. Roughly speaking, an equal conducting power may be obtained either in iron or copper for the same cost, the number of iron conductors being greater in proportion to the less cost, and the more conductors being the better.

17–19. Expansion and contraction are to be carefully provided against; _e.g._ by suitable bends at intervals in long lines of horizontal conductors and by bearing collars, allowing of slip in vertical lines.

20. Soldered or welded joints are desirable, but not absolutely necessary.

21. Gives engravings of connections recommended by Sir W. S. Harris, where soldered joints cannot be used, and which fulfil the conditions specified in sections 17–19.

22. Soldered or welded joints to be used where discharge is possible with unsoldered joints, and likely to ignite dust or inflammable substances near.

23. Iron may be connected by similar joints as for copper, or by screw joints as for gas pipes. No white lead to be used, it being a bad conductor.

24. Iron flat bands may be connected by rivets or screws, working in slots, to provide for expansion, each surface in contact being at least six times the sectional area of band.

25. Copper bands to be similarly connected. Joints between different metals may be soldered, screwed, or rivetted, the extent of surface in contact being regulated by the dimensions of the metal of the least conducting power. Access of moisture to surfaces in contact must be prevented, on account of local galvanic action and decomposition.

26. No precise limit can be fixed to protecting power of conductors. In England the limit is usually assumed as being the radius of the height from ground. It may be sufficiently correct for practical purposes, but cannot always be relied upon.

27. Conductors do not attract lightning; they only diminish the resistance due to the air. Even a change in the nature of the soil -over which a cloud passes may produce a discharge.

28. One angle of a building may receive a discharge, though another angle have a conductor. So every prominent part of a building containing explosive material should have a conductor.

29. In buildings of uniform height, provide a solid rod 5 feet above it at each end, and at each 45 feet in length; if the conductor be of iron the top should be gilt.

30. Buildings not over 20 feet long to have one vertical conductor at end, and a horizontal conductor on ridge.

31. If 20 to 40 feet long to have one vertical conductor in centre, and one along ridge, as last.

32. If 40 feet long to have two vertical conductors; if 100 feet long three conductors; in both cases with conductor along ridge.

33. Similar principles to be adopted in larger and more complicated buildings.

34. Each prominent part should have a conductor. The value of three or four points to terminals is not apparent unless the points are widely separated.

35. Conductors are to be connected horizontally, _e.g._, by ridge or eaves, which, when of metal, should be invariably connected with conductor. All metal surfaces whatever to be also so connected.

36. Sir W. S. Harris considers the relative conductivity of the several metals as being—of lead 1, tin 2, iron 2½, zinc 4, copper 12. So lead cannot be altogether depended on.

37. Avoid long lengths of horizontal conductors without earth contacts, as the currents might leave the conductor, and pass to earth, causing danger. Avoid sharp angles.

38. Good earth connections most important. Conductors are to be led into springs or wells or earth permanently wet. Not into watertight tanks. Shingle, dry sand, or dry mould are not sufficient. Provide several earth connections in all large systems of conductors as a precaution.

39. Lead conductors into ground in trenches 18 inches deep. Not less than 30 feet of metal to be in contact with moist earth.

40. Lead a flow of water over trenches if possible, _e.g._, from rain water pipes.

41. Trenches in rocky or dry soil to be 30 to 120 feet long, so as to obtain all moisture possible.

42. Connections in trenches may be of old iron, forming continuous metallic surface, the trenches to be filled with cinders or coal ashes. Water pipes form excellent earth connections, but gas pipes are dangerous.

43–44. Frequently inspect conductors, especially as to joints connecting different metals and defects in iron from rust.

45–46. Galvanize iron, care being taken that the coating is good.

47. Great care to be taken in case of contact of zinc coating with other metals, especially copper.

_April 8th, 1875._

FRED. E. CHAPMAN.
_Inspector General of Fortifications._