PROTECTION FROM LIGHTNING. 105 



the space over which a pointed metallic rod extends its protection, is a CONE, of 

 which the vertex is the point of the rod, of which the rod is the axis, and of which { 

 the section made by any horizontal plane is a circle, whose diameter is four times 

 the distance of such plane from the point of the rod. 



This estimate, which is evidently empyrical, and of which the experimental 

 grounds are not stated, requires much elucidation before it can receive un- 

 qualified assent. Does the conductor extend no protection to any surrounding 

 points at the level of its own points ? To what depth below the point does the 

 surface of the cone bounding the protected space extend ? or what is the posi- 

 tion of the base which limits the protected space taken in the vertical direction 

 downward ? Does the same form of cone limit the protected space for all 

 kinds of structures ? Is the angle of the cone affected by the presence of 

 large masses of metal, such as the guns in a battery, or the machinery used in 

 certain large factories, or the armament of a ship-of-war, or the engines of a 

 large steamship ? 



Theory affords no grounds for the law laid down by M. Charles, and obser- 

 vation is not wanting to show its fallacy. 



The foremast of the ship Endymion was struck by lightning at Calcutta, in 

 March, 1842. The mainmast, not fifty feet distant, had a chain-conductor, 

 which, according to the above law, would protect a circle of one hundred and 

 fifty feet diameter. 



The bow of the ship Etna was struck at Corfu, January, 1830, although the 

 mainmast had a chain-conductor. Other cases of similar character have oc- 

 curred to buildings on shore, one of which has very recently been communi- 

 cated to the French Academy. M. Arago, and many with him, were un- 

 willing to admit so vague a law, and experience confirms their decision. To 

 protect an extensive building, several paratonnerres would be necessary, and 

 the less the height of each, the greater should be their number, which, as well 

 as their position, must be determined by the condition that no part must be 

 more distant from the foot of the rod than twice its height. 



Although lightning falls generally by preference on the highest points, of 

 buildings, it does not always do so. Many cases are recorded in which, with- 

 out damaging the summit, it has struck at the middle of the height. In some 

 cases it has been seen distinctly to move in the horizontal direction, and strike 

 the side of a steeple. Cases are also cited in which it has entered by the 

 ground-floor, where it has struck persons and caused their deaths, doing slight 

 damage to the first floor, and none to the higher parts of the house. Such 

 facts suggest the utility of paratonnerres with points presented laterally and 

 obliquely. 



In some countries the superior extremities of paratonnerres are formed into 

 a group of points, radiating in various directions like a star. This method has 

 been suggested by the supposed advantages of horizontal and oblique points. 

 Experience has not yet supplied data on which any certain judgment can be 

 formed as to the efficiency of this expedient. 



The rod of a paratonnerre, by which it is intended to conduct the electric 

 influence to or from the earth, should be of such thickness that it may not be 

 fused by the most powerful current of electricity which is likely to pass through | 

 it. Experience indicates that this purpose will be sufficiently attained if it be 

 a square of three quarters of an inch in the side, or a circle gf the same di- | 

 ameter. Toward the base, an increased thickness is sometimes given to it, 

 I with a view to its stability. Paratonnerres are sometimes painted to protect 

 [ them from rust, and lampblack is selected as the material of the paint, in con- 

 k sequence of its conducting power. 



It has been already stated, that the inferior extremity of the paratonnerres 



