o resist, without buckling or cockling, the thrust of
the escape wheel, represented by the arrows _p o_. Now we can readily
understand that the lighter we make the parts _i j k m_, the weaker the
spring _h_ can be. You say, perhaps, if we make it too weak it will be
liable to buckle under the pressure of the escape wheel; this, in turn,
will depend in a great measure on the condition of the spring _h_.
Suppose we have it straight when we put it in position, it will then
have no stress to keep it pressed to the holding, stop or banking screw,
which regulates the lock of the tooth. To obtain this stress we set the
foot _F_ of the detent around to the position indicated by the dotted
lines _r_ and _n_, and we get the proper tension on the detent spring to
effect the lock, or rather of the detent in time to lock the escape
wheel; but the spring _h_, instead of being perfectly straight, is bent
and consequently not in a condition to stand the thrust of the escape
wheel, indicated by the arrows _o p_.

OBTAINING THE BEST CONDITIONS.

Now the true way to obtain the best conditions is to give the spring _h_
a set curvature before we put it in place, and then when the detent is
in the proper position the spring _h_ will have tension enough on it to
bring the jewel _s_ against the stop screw, which regulates the lock,
and still be perfectly straight. This matter is of so much importance
that we will give further explanation. Suppose we bend the detent spring
_h_ so it is curved to the dotted line _t_, Fig. 136, and then the foot
_F_ would assume the position indicated at the dotted line _r_. We next
imagine the foot _F_ to be put in the position shown by the full lines,
the spring _h_ will become straight again and in perfect shape to resist
the thrust of the escape wheel.

Little "ways and methods" like the above have long been known to the
trade, but for some reason are never mentioned in our text books. A
detent spring 2/1000" thick and 80/1000" wide will stand the thrust for
any well-constructed marine chronometer in existence, and yet it will
not require half a pennyweight to deflect it one-fourth of an inch. It
is a good rule to make the length of the detent from the foot _F_ to the
center of the locking jewel pipe _j_ equal to the diameter of the escape
wheel, and the length of the detent spring _h_ two-sevenths of this
distance. The length of the horn _k_ is determined by the graphic plan
and can be taken from the plo