that are mechanically bad, but those that are so electrically. A joint is said to be electrically bad when it offers resistance to the passage of electricity through it. There should be no resistance whatever. A careful inspection by Capt. Bucknill, R.E. (Appendix M, p. 243), has proved that bad joints in lightning rods are very abundant, though they appear perfectly sound; and everyone who has measured the electrical condition of conductors confirms this fact. Bad joints have the same effect as lengthening the conductor; and, in one case, one bad joint was found to have the same effect on a discharge of electricity as a conductor 1,900 miles long. It is evident that such rods may be worse than useless, for other parts of the building may offer easier paths for the discharge to the earth. If the joint be imperfect, and the rod convey a charge to earth, heat will be generated at the joint, the rod may be fused, and the discharge be diverted to the building.
Screwed, scarfed, and rivetted joints, however well they may be made mechanically, are certain to rust and corrode in time, owing to the expansions and contractions due to changes of temperature admitting moisture, and thus causing corrosion and resistance. No joint can possibly be electrically perfect that is not metallically continuous, and careful soldering, in addition to screwing, scarfing, or rivetting, is the only certain mode of securing this. Soldering is a method that has borne the test of experience, and its success as a means of securing perfect joints leaves no excuse for its omission. The fewer joints the better, but where there are joints they can only be made electrically secure by careful soldering.
PROTECTION OF ROD.—The lower part of copper rods is sometimes stolen for the sake of the metal. This can be guarded against by putting it inside a length of iron gas-barrel, extending from some distance below ground to 10 ft. above it.
PAINTING.—Iron conductors, even if they are galvanized, should be painted throughout, except at the points, which should be gilded or nickel-plated.
In France and Belgium painting is resorted to to a considerable extent, and the practice was recommended by the late Professor Joseph Henry, and followed very largely in America. [Appendix F, pages (99) and (113).]
ATTACHMENT TO BUILDINGS.—The evidence against the use of glass or other material in order to insulate the conductor, is overwhelming, and insulation may be regarded as unnecessary and mischievous. The essentials are (1) that the rod be attached to the building by fastenings of the same metal as itself, (2) that the fastenings be of adequate strength, (3) that they be of such form as not to compress or distort the rod, (4) that they allow play for its expansion and contraction, (5) that they hold it firmly enough to prevent all the weight falling on any one bearing.
Where practicable it is well to take the rod down that face of the house which is most exposed to rain.
EARTH PLATES.—This portion of the lightning conductor is of the utmost importance, but has hitherto been the most neglected. The majority of cases in which lightning has caused injury very near to or upon conductors are traceable to those conductors having imperfect earth terminals. We know of many cases in which the earth terminals have been miserably imperfect, or entirely neglected, when the above-ground portion has been perfectly satisfactory. In fact, though it may be admitted that the case found by Dr. Mann,[2] of the lightning rod of a church tower, the lower end of which was thrust into an empty glass bottle, is an exceptionally bad one; yet there are sadly too many, of which the Middlesboro’ case, Appendix I, page (217) is a perfectly fair type.
2. Quarterly Journ. Met. Soc., Vol. II., p. 420.
A convenient earth connection is often afforded in towns by the iron mains for gas and water—arguments both for and against the utilisation of both water and gas mains will be found in the Appendix—we, therefore, need only state our opinion in favour of connection with both. But no connection should ever be made with soft metal pipes, because of the risk of their fusion; and the conductor should be kept as far as possible from internal gas pipes on account of the risk of lighting the gas at an imperfect joint.
As a general rule we advise the soldering of a plate of metal, copper to copper, iron to iron, to the lower end of the conductor. The earth plate should always be of the same metal as the rod, otherwise destructive galvanic action sets in. This plate, which may be flat or cylindrical, must not have less surface than 18 square feet, i.e., 9 square feet on each face; there is no advantage in notching or pointing it. A hole must be dug, or well sunk, to receive this plate, and the hole must be so deep that the earth surrounding the plate shall never be dry. Any available drain or other water should be allowed to soak into the earth, over the site of the plate. After the hole has been dug, and the plate lowered into position, it should be filled with cinders, or coke. In extremely dry rocky localities, it is sometimes impossible to fulfil these conditions: then the best thing to do is to bury three or four hundredweight of iron at the foot of the conductor, still using the earth plate and the coke, and taking especial care that the rain-water and sink pipes discharge over it.
All drains, water-courses, in fact, everything which will assist in distributing the charge over a large extent of moist earth should be utilized by leading branches from the earth plate to them, or a long length of the rod may be laid in a drain if it be one which will be constantly wet.
SPACE PROTECTED.—The question as to the extent of the space which will probably be protected by a lightning rod is one which is of very great practical importance, because it governs the number and height of the upper terminals which are required for the protection of any given building. The index to the Appendix shows that “Protection, Area of,” is discussed upon twenty-nine pages in different parts of the Appendix. It has been laid down that the space protected was a cone, having the point for its apex, and a base whose radius was equal to twice the height of the point, while the latest French official instructions, Appendix F, p. (67), state that a point will “effectively protect a cone having the point for its apex, and a base whose radius is 1·75 of its height.” The English War Department instructions considerably reduce this space by asserting, Appendix F, p. (71), that “no precise limit can be fixed to the protecting power of conductors. In England the base of the protected cone is usually assumed to have a radius equal to the height from the ground; but though this may be sufficiently correct for practical purposes, it cannot always be relied upon.”[3]
3. On page (96) two instances are recorded in which, if the evidence can be trusted, the stroke fell within a radius equal to the height, but it is only right to say that the facts are not very clearly recorded.
According to this rule, the church of Ste. Croix (see Appendix F, p. (141)), would require four upper terminals, one on steeple, one on chancel, and one in the middle of each half of the transept.
From theoretical considerations stated by Mr. Preece, Appendix F, p. (137), he arrives at the conclusion that “A lightning rod protects a conic space whose height is the length of the rod, whose base is a circle having its radius equal to the height of the rod, and whose side is the quadrant of a circle, whose radius is equal to the height of the rod.”
At present we have not sufficient data to enable us theoretically to calculate the space protected by a lightning rod, and therefore we are compelled to draw up our rules upon the question entirely from experience, and here we find, that with the doubtful exceptions already mentioned, there is no recorded instance of a building being struck by lightning within a conical space, the radius of whose base was equal to its height, and we think that the adoption of this rule may reasonably be expected to yield that security in the future, which as far as we know, it has done in the past.
HEIGHT OF UPPER TERMINAL.—This matter is one which may be left entirely to the option of individual architects and engineers, subject, of course, to the opinions expressed under the heading “Space Protected.” In France extremely long tiges, or upper terminals, generally 33 feet long, are used; but it is obvious that they are