in the transmission of force from the main 
spring are certainly increased by the num- 
ber of wheels in the train, it was proposed, 
in the infancy of the art, to detach the last 
wheel, or that nearest the balance, or time 
measurer, from tire rest, and to move this 
by a separate spring or weight; so that 
in this contrivance the time measurer is act- 
ed iipon by one single wheel, and the rest 
of the train is employed in winding up the 
secondary first mover at short intervals, 
such, for example, as every half minute. 
We shall also have to mention some escape- 
ments in 'which the winding up is performed 
in every single vibration. With regard to 
remontoires it may be remarked tliat they 
either greatly shorten, or else destroy the 
periodical irregularities of the train, and 
those of the main-spring; but that with 
regard to the influence of oil and other 
causes of more permanent difference their 
advantage is not very considerable, because 
the remedy is not applied where the motion 
is quickest. 
Whether the irregular action of the main- 
taining power be diminished by the remon- 
toire or not, it is desirable that the im- 
pulse on the balance, through the escape- 
ment, should affect the natural measure of 
its vibrations as little as possible ; or rather 
that it should tend to equalise them when 
the arcs of vibration vary. Some attention, 
but not much, has been paid to the equalis- 
ing quality of an escapement, principally by 
making the faces of the pallets of a figure 
suitable to that effect; but tliese are now 
for the most part abandoned, and the me- 
thod of applying the force constitutes the 
distinguishing feature in this part of our 
modern chronometers. If a balance be set 
to vibrate by the mere action of its pendu- 
lum-spring, its motion will soon decay ; but 
if we suppose a lever or pallet to proceed 
from its axis, and a maintaining power to 
be applied to this, it is obvious that if the 
power meet the pallet in its progress from 
the point of quiescence it will shorten the 
time, and also the arc of that semi-vibration ; 
and, again, that if the power follow the pal- 
let in its progress towards the point of 
quiescence, it will drive it home sooner, and 
consequently will shorten the time of tliat 
semi-vibration ; and that actions contrary to 
these would lengthen the times. If, there- 
fore, the action itself, which may be con- 
sidered as an accellerating force, be not 
applied on both sides of the point of qui- 
escence through a certain arc, determin- 
able from the circumstances, the maintain- 
ing power when it comes to be applied will 
alter the time ; and if this vary, the time 
must also vary. Now, the remedy at pre- 
sent adopted is to make the balance vibrate 
through a very large arc, such as a semi- 
circle or more, and to follow the pallet 
in each returning vibration by a strong 
power exerted through a very small arc, as, 
for example, 15° or 16°. By this contrivance 
the balance will vibrate at perfect liberty, 
out and home, through two semicircles, 
or 360°, excepting the small part during 
which tlie impulse is given ; and it the im- 
pulse vary, the arc of vibration will vary, 
and with it the time, unless the spring be 
made of a certain definite length, or taper- 
ed in its thickness according to the expe- 
rience which many artists in this country 
possess. 
The escapement generally used in our 
best chronometers, as we shall hereafter see, 
consists of a toothed wheel at the end of 
the train, which is prevented from running 
down by a detent or hook, and of two pal- 
lets, a longer and a slrorter, fixed upon the 
verge or axis of the balance. These pal- 
lets are so placed, that when the face of the 
longer pallet has just arrived before one of 
the teeth of the wheel, the shorter pallet 
strikes out the hook, and allows the wheel 
to push forward the longer pallet with its 
tooth, during which action, the hook falls 
again into its place to catch the succeeding 
tooth. The balance therefore proceeds in 
its vibration, and returns again without 
disturbing the train ; because the short pal- 
let does not strike out the hook in its back- 
ward course, but only acts on a slender 
spring, resembling those formerly used in 
the jacks of harpsichords. In tliis manner 
the vibrations are kept up ; and so little do 
the variations in the maintaining power 
affect the rate, when all tlm adjustments are 
made, that if the main-spring be let down 
to only a small part of its ordinary tension, 
these time-pieces will keep the same rate 
for many hours together. 
However perfect, practically speaking, the 
application of the maintaining power may 
be, yet if the balance and its spring be 
subject to vicissitudes from heat and cold 
it will be in vain to expect accuracy. There 
are two ways of correcting this compound 
time-measurer. The first, which was in- 
vented by Peter Leroy, consists in causing 
the balance to enlarge itself, instead of 
contracting by heat ; by which means the 
spring, when in tlie state of greater rigi- 
dity has more work to do ; and the other 
