A PENDULUM FOR THE REDUCTION TO A VACUUM. 
465 
be attached to those zealous officers, surrounded as they must be with diffi- 
culties of every kind for carrying on such delicate experiments. In fact, 
amongst the multitude of experiments that I have myself made, I have seldom 
found, after I had dismounted a pendulum, and then replaced it (even on the 
same day, under all the favourable circumstances of equality of temperature &c., 
and with all the conveniencies of manipulation,) that I could make it tell the 
same story in the next series of experiments. Even the same pendulum, furnished 
with two different knife edges, rendered synchronous or nearly so, similar to 
those described in the above enumeration as convertible pendulums (No. 25 — 
38), where the trifling difference in the results of each pair of knife edges, 
ought, after proper reductions, to be a constant quantity, will frequently differ 
by an amount much greater than can be attributed to the errors of obser- 
vation. 
The fact, I believe to be, that the pendulum furnished with a knife edge and 
agate planes, as at present constructed, is a very inadequate instrument for the 
delicate purposes for which it was originally intended : and a more rigid exa- 
mination and adjustment of that part of the instrument are requisite, before 
we can depend on the experiments made with it, either for the determination 
of the length of the seconds pendulum, or even for the comparison of results 
obtained in different parts of the world. The knife edge is seldom or never 
perfectly straight ; the planes are seldom or never perfectly true : at least, I 
have never found one so, amongst the number of those on which I have experi- 
mented. The consequence is that, as there is generally a little play in the Y’s, 
the knife edge is not always let down on the same parts of the agate plane. 
This may be best detected by holding a lighted candle behind the knife edge 
when it is resting on the plane : by which method the smallest inequalities in 
the points of contact are readily discernible. But, the fact is rendered still 
more evident by reversing the pendulum in the Y’s, when a sensible difference 
in the result generally takes place. Amongst the numerous pendulums in my 
possession, I have not met with more than one, that does not differ in the 
results by an appreciable quantity, when the pendulum is reversed in the Y’s, 
or turned half round in azimuth. If the knife edge and planes were perfectly 
correct and true, there ought not to be any difference in the results, whichever 
side of the pendulum is placed fronting the observer : how then are we to 
