_ 
at ihe Island of Balta, and at Woolwich Common. 443 
attached to the lower extremity of the metallic thread. These 
together constitute the vibrating part of the apparatus. 4. A 
metallic circular plate, capable of accurate adjustment to hori- 
zontality, and of very gentle elevation or depression by means of 
a fine screw, so as to be brought into contact with the bottom of 
the ball, when the pendulum is at rest; and thus to furnish the 
actual distance between the fixed plane which supported the 
*¢ knife edge” and the bottom of the ball. 
Now, without adverting to other matters which call for the 
utmost nicety in this experiment, it is obviously requisite that 
the knife-edge with its stem and nut be made to vibrate, not ap- 
parently, but to a considerable degree of precision, in the same 
time as the wire and attached cup and ball: for a very slight dif- 
ference in their respective times of vibration may evidently im- 
pair the accuracy which is assumed in the result. But how can 
the interval of vibration in the upper part of the apparatus he made 
equal to that of the other portion, when this latter is not known, 
being a main object of the experiment itself? How, again, is 
the latter ever known separately, when the experiments deter- 
mine solely the motion of the complex pendulum? But ad- 
mitting, in order to lessen the difficulty, that it 7s known, how 
are the knife-edge, stem, and nut, made to vibrate accurately 
or nearly accurately in the same time? To ascertain whether 
they do so or not, appears a problem of equal diificulty, and re- 
quiring equal care with the other. Sunposing, however, all this 
effected, and the precise period of osci.iation of the knife-edge, 
stem, and nut, ascertained and adjusted for one latitude, and for 
one temperature,—how is it determined through what space the 
nut, N, must move, to ensure the same or any proposed time of 
oscillation in another latitude or at another temperature? And 
it is evident that in so short a part of the apparatus as this which 
we are now considering, where a minute change in the position 
of the nut may change that of the centre of oscillation (that is, 
of NK EB) even inches, the variations of gravity in different 
latitudes, or of temperature whether in different latitudes or the 
same, may so affect the place of that centre, as to eighth the 
vibration of the complex pendulum, and totally to cancel all pre- 
tensions to extreme accuracy in the general result drawn from 
the experiment. All this, then, requires a frank explication be- 
fore the result can be safely adopted, even though it may agree 
with an hypothesis to which our prepossessions tend; unless it 
be decisively confirmed by another experiment of unquestion- 
able correctness, 
But there is another circumstance which may, still more than 
the preceding, affect the precision of M. Biot’s operations in the 
north. To obtain the length of a second or other pendulum, 
froin 
