There is no doubt that lightning generally strikes elevated rather than low objects, and therefore it has been thought that a building surrounded by steeples had nothing to fear. As previously stated, however, the bolt sometimes selects a low object when high ones are at hand. For example, lightning fell upon a house occupied by Lord Tilney in Naples, although it was surrounded on all sides, at the distance of four or five hundred paces, by the towers and domes of a great number of churches, all wet with a heavy rain. In considering the matter of protection from lightning we must bear in mind that trees, buildings, masts and other elevated points exert no attractive power on the thundercloud except in connection with the great plane where they are situated. The primary cause of the discharge is not in the metals of the building, the exact point or line in which the insulation by the air breaks down being determined by a variety of causes. The elevated points of a building or ship may form a channel for the passage of the current, but it is not the only one nor the cause of the discharge, which would take place sooner or later though the ship or building were absent altogether. There has been a difference of opinion in regard to the area protected by lightning-conductors, early notions on this point having been much exaggerated. Leroy's, in 1788, is the earliest positive statement which I have met. It is, that a conductor protects a horizontal space around it equal to somewhat more than three times the height of the metal rod above the building to which it is attached. The physical section of the Academy of Sciences of Paris, on being consulted by the Minister of War in 1823, expressed the opinion that a lightning-conductor protects a circular space of which the radius is equal to the height of the rod. Here, apparently, is a wide difference, but possibly the estimates refer to different elevations. Leroy clearly intended an area at a level with the top of the building: thus, supposing the rod to be attached to a chimney six feet in height and to rise a foot and a half above the chimney, then it would protect a radius of about ten feet on the roof. The estimate of the Academy of Sciences speaks of the total height of the rod, and refers to a horizontal area at the base of the measurement, whether this began at the ground or at the top of the structure to which the rod is attached. In this view the estimates do not differ so much as might appear, the latter being about one-third less than that of Leroy. Other French writers estimate the area protected as having a radius of double the height of the rod above the highest point of the connected structure, being twice the radius allowed by the Academy. Later physicists have been cautious in giving figures, for experience has shown that estimates of protection are not accurately observed by the descending bolt. For instance, when Her Majesty's corvette Dido,