574 



NATURE 



[Oct. 26, 1876 



be given when the pendulum is falling or rising. In the 

 first case any variation in the friction of the clock-train 

 tending to increase the impulse will make the clock gain, 

 in the second, to make the clock lose. But why could we 

 not so arrange an escapement that the impulse should be 

 given, half when the pendulum is falling, and half when 

 it is rising ? and then any variation in the force of the im- 

 pulse will simultaneously cause the clock to gain and lose, 

 and so correct itself. 



This is what can nearly be done in 'the dead escape- 

 ment, but not exactly, because the condition of so doing 

 is that each tooth of the escape-wheel, shall drop exactly 

 upon the corners C and A, dividing the dead faces c D, w A, 



Fig. 17. 



from the impulse faces c e, a b, and you are bound to 

 allow a little margin for safety. But just in so far as this 

 condition is fulfilled, so the escapement is a good one. 



Gravity Escapements. 



To understand the principle" of these you cannot do 

 better than refer to Fig. 16, which shows the original form 

 invented by Mudge. The tooth of the escape- wheel T^by 

 operating upon the slant Sj has lifted the pallet Si Y^ to its 

 present position, and the tooth is now being, held by the 

 hook at the end of the pallet. Toe pendulum, which is de- 

 tached from the pallets, is advancing towards the arm Y^ A^ 



connected with this pallet, and will by-and-by lift it, freeing 

 the escape-wheel : the other tooth of the wheel will then 

 operate upon the slant Sj, of the other pallet Sj Vg, and lift 

 that until it is in turn detained by the hook at its extre- 

 mity. Meanwhile the pendulum carries the first pallet 

 to the extremity of its swing, and then returns with it ; 

 but as there is now no tooth in the way to receive it, the 

 pallet Sx Yi will drop, until it occupies a corresponding posi- 

 tion to that in which the pallet S2 Yg is at present situated. 



Thus you see the pallet falls upon the pendulum a 

 greater distance than the pendulum lifted it, and this 

 difference — the weight of the pallet multiplied into this 

 space — forms the impulse. This impulse, of course, is 

 constant, and any variation in the force of the clock-train 

 will only cause either pallet to be lifted with greater or 

 less rapidity. 



This escapement used to "trip," that is to say, that occa- 

 sionally the pallets were lifted with so much rapidity, as 



Fig. 18. 



to be thrown completely away, and then the hooks at the 

 extremities of the slants failed to catch the escape-wheel 

 teeth, and the wheel ran on (apparently making the clock 

 gain). 



This escapement was improved by the late Mr. Bloxam,^ 

 who put cogs near the axis of the wheel to do the lifting, 

 which was thus effected more slowly (see Fig. 17). 



That collection of arms, a A a, and the cog-wheel, w, to 

 which they are fastened form the escape-wheel. The 

 pendulum, P P, is now passing through the vertical position 

 or zero, and the pallet p^ p^ has fallen as low as it can, 

 and is resting upon the banking pin B^. But the other 

 pallet, Pg P2, has been lifted, and is now being held by the 

 cog T2, the cog can lift it no further because one of the 

 arms, a a A, is caught by the locking stud Lg (which is 

 situated upon the opposite side of the pallet). By-and-by, 

 however, the pendulum will reach the pallet, push it away, 



^ Mr. Bloxam discovered that the pendulum should quit and take up each 

 pallet at angles depending upon its arc of vibration. The reader who wishes 

 for further information cannot do better than consult Mr. Bloxam's elabo- 

 rate paper. Memoirs R.A.S, vol. xxii. p. 103 ; and another, vol. xxyii. 

 p. 61. 



