e diameter of the former. In being
heated a rod contracts in length, and is then liable to fracture by the
shrinkage, but if of sufficient size these results are not likely to
occur. An iron rod, by successively receiving an electrical discharge,
is sometimes reduced in size.

The conducting power of metals likely to be found in buildings is as
follows: taking the power of lead as _one_, that of tin will be
_two_--that is, tin conducts electricity twice as well as lead; iron,
nearly two and a half times as well; zinc, four times; and copper,
twelve times. From this comparison of conducting power the important
fact will appear that when any two of these metals are used in the same
line of conduction, the one of low power should be proportionately
larger. Sir W.S. Harris--perhaps the best authority on lightning-rods
in general--advises that the size of the rod, if of iron, should be
three-fourths of an inch in diameter, although he admits that probably
never in the experience of man has a rod half an inch in diameter been
melted by an electrical discharge. He regarded the extent of surface
rather than quantity of metal in the conductor as the measure of its
power, while many other electricians hold the contrary opinion.

It is important that the conductor should form an unbroken line
throughout its extent, otherwise there is danger that a portion of the
charge may be diverted from it. For instance: a large barn struck not
long since had a conductor at each of three corners. In order to
maintain the uniformity of the four angles of the square hip roof, a rod
was run from the main conductor down the fourth angle to the hip, where
it terminated in an erect point. A heavy discharge struck the main rod
at the cupola, and, descending, divided among the four branches. That on
the short branch jumped from its end to the metal sheathing along the
angle of the roof, which it followed to the gutter, passing along this
to one of the conductors, doing some damage on the way. Had not the
charge found a line of metal on which to continue its course from the
end of the rod, it would have done greater damage, and most likely have
set the building on fire.

Another point of importance is, that the connection of the joints of the
rod be perfect, as explosions and fusion occur wherever the surface in
contact is less than the size of the rod, unless the latter is much
larger than necessary. The hook and the lap joints, if not very
carefull