220 
required to compensate for the expansion of steel than of 
wood, it would be inconvenient to place the centre of the bob 
upon the upper end of the zinc tube. The latter is, therefore, 
passed through the hole in the bob, its upper end being about 
nineteen inches above the upper end of the bob ; a brass 
collar is fixed upon it, from which two steel rods descend 
(one on each side of the central pendulum rod); two holes 
are made from top to bottom of the bob to receive them, 
and they are secured halt-way down by screws. Thus, as 
the zinc expands or contracts, it raises or lowers the centre 
of the bob as before described ; or rather a change of tempera- 
ture, which causes the zinc tube to expand upwards, causes 
the steel rods (which are all three suspended by their upper 
extremities) to expand downwards, and thus the centre of 
oscillation, which coincides very nearly with the centre of the 
bob, is kept constantly at one height. 
As the centre of the bob of a pendulum to beat seconds of 
mean time in this latitude should be about 39’ 14 inches from 
the point of suspension, the following proportions were found 
by calculation for the component parts of this pendulum : — 
Length of central steel rod, from the bend of the spring to 
the bearing surface of the regulating nut, 45*37 inches; 
length of zinc tube, 30 inches. 
Length of side rods, extending from top of zinc tube to 
centre of bob, 22*87 inches. 
The proportions of these lengths to one another were 
determined by Mr. Lowe from data found by experiments 
upon pieces cut from the steel rods and zinc tube of which 
the pendulum is made. 
Objections had been made to the use of zinc, on the 
ground of the uncertainty of its ratio of expansion and con- 
traction at different temperatures. Mr. Lowe was, therefore, 
induced to construct an apparatus by which to ascertain 
whether this objection was valid. A considerable number of 
sj)ecimens of zinc were examined, but it was invariably found 
