IMPROVEMENTS IN CLOCK-ESCAPEMENTS. 425 



1/7 1 



which is about - a second. Even if y = - , the daily rate will be only - sec. ; below that 



At o • y 



angle it increases rapidly. (It may be observed that this error is of the same sign as the 

 circular error, and therefore cannot be employed to correct it.) 



It appears therefore that, within moderate limits, there is no need to be particularly 

 careful about the proportions of a and y. And this is very convenient, because it happens 



that, in small clocks especially, the escapement is easier to make safely with y less than — — • 



V 2 



because when it is so large, the arms can hardly be made light enough for a pendulum of the 



usual weight, without making them so thin that they are liable to bend (so little impulse does 



a pendulum require when it is delivered from friction) ; and moreover the smaller the lift 



and the heavier the arms, the more they tend to protect themselves against tripping. 



Lastly, it is essential to the success of any gravity escapement that it should be easy to 

 make and cheap. If it does not render the pendulum so independent of the train that the 

 'movement' of a common house-clock will do for an astronomical clock, it may as well do 

 nothing; and if the escapement itself is so complicated or delicate, and therefore expensive, as 

 to bring up the total cost of the clock to much the same as the present price of a first-rate 

 clock, it will have very little chance of coming into use, however good it may be. 



All these conditions are satisfied by the escapement of which a drawing is given on the next 

 page. The gravity arms are lifted by the three pins set in the three legs of the scape-wheel. 

 In this drawing the upper pin is about to lift the right arm as soon as the pendulum, now 

 going to the left, has lifted the left arm, and so carried the stop away from the long tooth 

 which is stopped or locked by it. The pin which is lifting always moves from the vertical line 

 through the center of the scape-wheel to 60° beyond it ; not as in the three-legged dead escape- 

 ment from -30° to + 30°; and consequently there is no occasion here for such delicate work as 

 in that escapement. By the time the pin has got to 60° from the vertical, the tooth belonging 

 to it will have reached the stop on the right arm, and so the scape-wheel will be stopped until 

 the pendulum returns to that side and releases it ; and then the pin at the bottom will lift the 

 left arm. Each arm, you observe, (as in all gravity escapements) descends with the pendulum 

 to a lower position than that at which the pendulum begins to lift it, and thus the impulse 

 is given. The business of the scape-wheel is merely to replace the arm at the proper height 

 for the pendulum to take it up and unlock the wheel ; and this raising of the arms by the 

 wheel must be done so slowly, that there may be no risk of the arms being sent out too 

 far and allowing the tooth to slip past, or ' trip,' instead of being caught by the stop. 



1. This first condition is satisfied in this escapement by the simple addition of a fan-fly 

 set with a friction-spring on the scape-wheel arbor, just like the striking-fly of a common clock. 

 Probably such a fly could not be used with advantage unless where the scape-wheel has very 

 few teeth, and consequently moves through a large angle at each beat. The construction of 

 the scape-wheel also requires the pallets or gravity arms to be much longer than they are in 

 any other escapement, and consequently the depth of the locking, which must occupy some- 

 thing less than the angle a - y, is greater than usual, and this also is an additional security 

 against tripping. 



