[ 486 ] 
pendulum, which I intended to place fo, that, by 
adting very near the centre of an index of a confider- 
able length, even the fmallett alteration in the bars 
would be made fenfible, and by the motion of the 
index I fhould be able to form a judgment, whether 
the rods moved regularly and freely, or not : but be- 
fore this was put into execution, I contrived, by 
fattening the two bars to the back plate of a clock, 
not only to make them anfwer the end above pro- 
pos’d, but, at the fame time, to lengthen or lhorten 
a pendulum of a common conftrudtion, in fuch a 
manner, as fufficiently to correct the irregularities 
ariling from the influence of cold or heat upon it. 
The manner of applying them is deferibed by Fig. 3. 
In which, a a a a reprefents the back plate of the 
clock, bbb^ a triangular piece of brafs, ferew’d by 
two ferews, thro’ the flits c c, to the plate, yet fo 
that it may be drawn backwards or forwards by 
means of the ferew at ef is a brafs bar, about 
two feet in length, made fatt at the bottom, by a 
ferew and two pins at /’ to an iron one of equal di- 
menfions, to which it is likewife ferew’d by the 
ferews 1, 2, 3, &c. after the fame manner as the 
rod of the pendulum already deferibed. The iron 
bar is fattened at the upper end to the triangular 
piece of brafs, nearly under that part of the brafs bar 
marked e. g h is a ttrong brafs or iron lever, move- 
able upon a centre at g, and is fupported by the up- 
per end of the brafs bar ; i i is the cock, on which, 
in a common clock, the pendulum is hung ; k 
part of the rod of the pendulum, whofe fpring pair- 
ing thro’ a fine flit in the cock i /, is fatten’d to a 
ftud riveted into the lever at /. The flit in the cock 
mutt 
