tted plan. The end, however, should
approach as near to the discharging jewel as possible and not absolutely
touch. The discharging (gold) spring _m_ is attached to the blade _i_ of
the detent with a small screw _l_ cut in a No. 18 hole of a Swiss plate.
While there should be a slight increase in thickness in the detent blade
at _w_, where the gold spring is attached, still it should be no more
than to separate the gold spring _m_ from the detent blade _i_.

IMPORTANT CONSIDERATIONS.

It is important the spring should be absolutely free and not touch the
detent except at its point of attachment at _w_ and to rest against the
end of the horn _k_, and the extreme end of _k_, where the gold spring
rests, should only be what we may term a dull or thick edge. The end of
the horn _k_ (shown at _y_) is best made, for convenience of elegant
construction, square--that is, the part _y_ turns at right angles to
_k_ and is made thicker than _k_ and at the same time deeper; or, to
make a comparison to a clumsy article, _y_ is like the head of a nail,
which is all on one side. Some makers bend the horn _k_ to a curve and
allow the end of the horn to arrest or stop the gold spring; but as it
is important the entire detent should be as light as possible, the
square end best answers this purpose. The banking placed at _j_ should
arrest the detent as thrown back by the spring _h_ at the "point of
percussion." This point of percussion is a certain point in a moving
mass where the greatest effort is produced and would be somewhere near
the point _x_, in a bar _G_ turning on a pivot at _z_, Fig. 138. It will
be evident, on inspection of this figure, if the bar _G_ was turning on
the center _z_ it would not give the hardest impact at the end _v_, as
parts of its force would be expended at the center _z_.

[Illustration: Fig. 138]

DECISIONS ARRIVED AT BY EXPERIENCE.

Experience has decided that the impulse roller should be about half the
diameter of the escape wheel, and experience has also decided that an
escape wheel of fifteen teeth has the greatest number of advantages;
also, that the balance should make 14,400 vibrations in one hour. We
will accept these proportions and conditions as best, from the fact that
they are now almost universally adopted by our best chronometer makers.
Although it would seem as if these proportions should have established
themselves earlier among practical men, we shall in these drawings
confine ourse