adii as the impulse roller. To locate the
center from which the arc which defines the back of the teeth is swept,
we halve the space between the teeth _A^2_ and _a^4_ and establish
the point _n_, Fig. 141, and with our dividers set to sweep the circle
representing the impulse roller, we sweep an arc passing the point of
the tooth _A^3_ and _u_, thus locating the center _w_. From the center
_k_ of the escape wheel we sweep a complete circle, a portion of which
is represented by the arc _w v_. For delineating other teeth we set one
leg of our dividers to agree with the point of the tooth and the other
leg on the circle _w v_ and produce an arc like _z u_.

ORIGINAL DESIGNING OF THE ESCAPEMENT.

On delineating our chronometer escapement shown at Fig. 141 we have
followed no text-book authority, but have drawn it according to such
requirements as are essential to obtain the best results. An escapement
of any kind is only a machine, and merely requires in its construction a
combination of sound mechanical principles. Neither Saunier nor Britten,
in their works, give instructions for drawing this escapement which will
bear close analysis. It is not our intention, however, to criticise
these authors, except we can present better methods and give correct
systems.

TANGENTIAL LOCKINGS.

It has been a matter of great contention with makers of chronometer and
also lever escapements as to the advantages of "tangential lockings." By
this term is meant a locking the same as is shown at _C_, Fig. 141, and
means a detent planted at right angles to a line radial to the
escape-wheel axis, said radial line passing through the point of the
escape-wheel tooth resting on the locking jewel. In escapements not set
tangential, the detent is pushed forward in the direction of the arrow
_x_ about half a tooth space. Britten, in his "Hand-Book," gives a
drawing of such an escapement. We claim the chief advantage of
tangential locking to lie in the action of the escape-wheel teeth, both
on the impulse stone and also on the locking stone of the detent.
Saunier, in his "Modern Horology," gives the inclination of the front
fan of the escape-wheel teeth as being at an angle of twenty-seven
degrees to a radial line. Britten says twenty degrees, and also employs
a non-tangential locking.

Our drawing is on an angle of twenty-eight degrees, which is as low as
is safe, as we shall proceed to demonstrate. For establishing the angle
of an escape-wh