4 
; 
4 
F. A. P. Barnard on the Pendulum. 185 
pendulum-rod. And though, in such a case, if the clock should 
run more slowly than in warmer weather, we might justly infer 
an over-compensation to exist, we should be unable to determine 
exactly the excess. . 
The partial or total failure in practice of plans of compensa- 
tion theoretically perfect, has been sometimes attributed to the 
unequal rapidity with which changes of temperature take place 
in different parts of the same pendulum. EHyery plan of com- 
pensation is necessarily founded on the supposition that, under 
all alterations of temperature, the temperature throughout the 
entire instrument is simultaneously the same. It is easy to see, 
for example, that if the mercury @mployed to compensate the 
dulum of a common astronomical clock were to be wholly 
inclosed in a cylinder of some material entirely impervious to 
heat (supposing such a material to exist), it would be altogether 
useless for the purpose intended. And that which would thus 
be true, on the supposition that the mercury could not change 
its temperature at all, must be measurably so, if its changes of 
temperature lag behind those of the rod. Glass jars for contain- 
ing the mercury in pendulums of this construction have been 
objected to, on the ground that they do introduce an irregularity 
of this sort; and accordingly Mr. Dent, the distinguished practi- 
the 
woul appear to be well worth making, before pronouncing the 
mercurial compensation to be unsatisfactory, or even condemning 
the glass cylinders, 
The escapements called remontoirs apparently set the pendu- 
lum free from most of the liabilities to disturbance which the 
train Introduces. In clocks provided with these escapements, 
nee force of the train itself is exerted not in impelling the pen- 
uum, but in raising a small weight, or bending a slight spring, 
Which subsequently acts—the former in its descent or the latter 
i 1f8 Tecoil—in moving the pendulum. The gravity-remontoir. 
fparently furnishes an impelling force which is perfectly uni- 
orm, there being nothing but the very slight friction on the 
Pivots of the arms carrying the remontoir weights to disturb this 
of fata, The spring-remontoir is free from even this source 
avence of the varying elasticity of the spring occasioned. by — 
Change temperature. Both of these contrivances, however, * 
