HOROLOGY. 
ceives the succeeding tooth of the wheel af- 
ter t!ie escape from the point of D. In this 
manner the alteration may be conceived to 
go on, without limit. 
The celebrated George Graham improved 
this escapement very much, by taking off 
part of the slope furthest from the points of 
the pallets ; instead of which part, he form- 
ed a circular or cylindrical face, having its 
axis in the centre of motion. Pallets of 
this kind are seen on the opposite side of 
the wheel at E and G, having H for their 
centre or axis. A tooth of the wheel is 
seen resting upon the circular inner surface 
of the pallet G, which is not therefore af- 
fected by the wheel, excepting so far as its 
motion, arising from any other cause, may 
be affected by the friction of the tooth. If 
the vibration of the pendulum be supposed 
to carry G outwards, the slope surface will 
be brought to the point of the tooth, which 
will slide along it, and urge the pallet out- 
wards during this sliding action. When 
the tooth has fallen from the point of this 
pallet, an opposite tooth will be received 
on the circular surface of E, and wall not 
affect the vibration, excepting when the 
slope surface of E is carried out so as to 
suffer the tooth to slide along it. In 
the two former escapements, there is al- 
ways a certain portion of vibration takes 
place after the drop which drives the pal- 
lets back, and causes the index also to re- 
cede through a small arc : this has been dis- 
tinguished by the name of a recoil. Other 
considerable objections, besides that of the 
continued action of the maintaining power, 
have been made against escapements with 
a recoil ; but it would lead us too far into 
the minute departments of this subject, to 
discuss them. The escapement of Graham, 
and all such as have no recoil, have been 
called dead beat escapements, because the 
index for seconds falls directly through its 
arc, and remains motionless on the line of 
division till the next vibration. It may be 
observed, that the maintaining power in 
Graham’s escapement, may be applied dur- 
ing a small portion only of the vibration ; 
and that an increase of the maintaining 
power tends to enlarge the arc of vibration, 
but scarcely interferes with its velocity. 
The effect of the escapement which has 
been termed horizontal, because the last 
wheel in watches of this construction has 
its plane parallel to the rest of the system,- 
is similar to that of the dead beat escape- 
ment of Graham. In fig. 3, the horizontal 
wheel is seen with twelve teeth, upon each 
of which is fixed a small wedge supported 
above the plane of the wheel, as may be 
seen at the letters A and B. On the verge 
of the balance there is fixed part of a hol- 
low' cylinder of steel, or other hard ma- 
teria], the imaginary axis of which passes 
through the pivots of the verge. C repre- 
sents this cylindrical piece, into which the 
wedge D, may be supposed to have fallen. 
While the vibration causes the cylindrical 
piece to revolve in the direction which car- 
ries its anterior edge towards the axis of 
the wheel, the point of the w'edge will 
merely rub the internal surface, and no 
otherwise affect the vibration of the bal- 
ance than by retarding its motion, But 
when the return of the vibration clears the. 
cylinder of the point of the wedge D, the 
w'hee! will advance, and the slope surface 
of the wedge acting against the edge of 
the cylinder will assist the vibration of the 
balance. When the edge of the cylinder 
arrives at the outer point of the wedge D, 
its posterior edge must arrive at the posi- 
tion denoted by the dotted lines of continua- 
tion; immediately after which the wedge 
or tooth E will arrive at the position <?, and 
rest on the outer surface of the cylinder, 
where it will produce no other effect than 
that of retardation from friction, as was 
remarked with regard to the wedge D, 
until the course of the vibration shall bring 
the posterior edge of the cylinder clear of 
the point of the wedge. In this last situa- 
tion the wedge will act on the edge of the 
cylinder, and assist the vibration, as in the 
former case, until that edge shall arrive at 
the outer or posterior point of the wedge ; 
immediately after which, the leading point 
will fall on the inner surface of the cylinder 
in the first position, as was shewn in the 
wedge D. 
Time-pieces, with a pendulum regulator, 
are certainly the most perfect, when they 
are kept in a fixed situation ; and, for that 
reason, these are the only sort used in 
astronomical observatories. But external 
motion is so contrary to the regularity of 
their performance, that no sea chronometer 
has been since attempted to be constructed 
upon that principle. The balance regula- 
tor remained, as affording the only method 
by which the desired uniformity might be 
obtained in portable machines; and the 
great improvement made in that regulator, 
by the addition of a spiral spring, may be 
considered as one principal cause of the 
perfection which has been since attained in 
them. The first invention of attaching a 
