The arrangement usually adopted consists in erecting- upon each of sev eral chimneys a terminal spike or point, and connecting these points with a single main lightning conductor or leading them directly and independ ently, by separate rods, to the earth. The first plan is the preferable one. In the case of a building of considerable extent, two or more rods may be advisable, but ordinarily one main conductor, properly grounded, will be found to afford ample protection, provided connection has been made by metallic rods extending over every portion of the roof. A metallic roof, when connected properly with the earth through the medium of rain lead ers, will afford immunity from damage by lightning as completely as the most elaborate system of rodding. When the roof is surmounted by orna mental crestings of metal, it is obvious that these, when suitably connect ed with the ground by a metal conductor of ample area, will serve every purpose.
Rod Termination. —The rod is usually terminated either by a point (N. 7, Jigs. TS, 23, 25) or by a sphere, which it is customary to gild or platin ize, though the last-named refinement is quite unnecessary. The commis sion of the French Academy which lately examined and reported on the subject recommended as a very satisfactory form of termination " a cylin der of copper T86- of an inch in diameter and S or ro inches long, the sum mit being- tapered off into the form of a cone r IA or r y, inches high." For practical purposes, it makes, really, little or no difference whether the rod be terminated by a point or by a blunt or spherical extremity, although the former style is most generally adopted, as affording better conditions for the silent discharge of an electrified cloud. The continuity of the con ductor should not be broken at any point, and the safest plan, in the case of rods of solid metal, is to join the successive sections to one another by screw sleeves of the same metal, so applied that the ends of the united sec tions shall abut directly ag-ainst one another.
Insulators and use of insulators of glass or porcelain in making the attachment of the rod to the building, which is often prac tised, is an unnecessary precaution; for it is absurd to assume that a light ning flash which possesses sufficient intensity to break through hundreds of feet of air will be prevented by the interposition of a few bits of glass an inch or two in thickness from leaving the rod and seeking a better path, should one exist. The fastening of the rod to the building, therefore, should preferably be made as direct and simple as possible. Professor Phin, a reliable authority on this subject, recommends that " if the rod be flat it may be pierced with small holes and tacked directly to the building; but a better way for either round, square, or flat rods is to employ properly shaped staples of stout wire. These staples may be driven into the stud
ding of wooden houses or into the joints of brick walls, and when properly painted will not present an unsightly appearance." Where something better than staples is desired, he proposes a strap of metal of the same kind as that of which the rod is made, and "pierced with two holeS, whereby it may be attached to any structure by means of a couple of screws The advantages of this cievice are that it does not weaken the rod, is not un sightly, permits the rod to slide on the building as it expands and contracts by heat and cold, and permits it easily to be applied or removed without injury to the building." Several illustrations exhibiting the details of construction will be seen on Plate 7 (figs. all of which are so suffi ciently explanatory as to make a special description unnecessary.
Ear111 manner in which a lightning conductor will perform its intended service will depend absolutely upon the perfection of its connection with the earth. The good features above enumerated may all be present; yet if the ground termination be imperfectly made, the rod will be worse than iiseless. Defective ground connection, as attentive ob servers have frequently demonstrated, is the cause of the great majority of accidents that happen to buildings presumably protected by liglitning rods and supposed to be safe. There are two conditions relating to the ground connection of a lightning conductor which mnst be regarded as essential. The end of the conductor must be made to terminate in ground that is per manently wet, and it should be made to expose to this wet soil as large a surface as practicable. In the erection of a conductor, the importance of these conditions should not be underestimated, for npon them will depend the ability of the rod to conduct away the electrical discharge and dissipate it harmlessly in the earth.
The practice of placing the lower end of the rod in water is much less effective than to bury in the earth a considerable length of the rod, leading it away from the buildin,g, and surrounding it with coke. Phil], in his excellent treatise on lightning rods, approves this plan. His advice is as follows: " If a trench To feet long be sunk to the depth of permanent moisture and filled to a depth of 12 inches with coke, it will be ready to receive the end of the rod, and will furnish a path for all the electricity that will ever tend to escape front the clouds to the earth."