196 F. A. P. Barnard on the Pendulum. 
by constantly reducing the impulse weights, they may perhaps — 
become inconveniently small. 3 
To return to the subject of compensation for variations of 
temperature, it may be observed that, while every pendulumis 
liable to be disturbed by the forces other than grav Se 
upon it, and while these forces are not all of them subject to 
law, so that their effects can be exactly predicted and allo a 
for, it is not surprising that methods of compensation theoretr 
cally good should have failed to satisfy in practice. In the eles 
ted; and the proper mode of correction may be made a subject 
plaint made of the performance of the mercurial compan | 
and of the glass jars as connected with it, is well founded. — | 
greater rapidity. If the smaller bulk of the rod in proportion 
its the - 
dient last suggested is adopted, of introducing the mercury 
or the Co; i «tervals 
provided, which will be substituted for each other ne oe 
