// CI.OCK-COXIKOl.l.l.n IVXIXG FORK 157 



Ki is not sutVirli'iit tn n[MT.iif it. As soon as tin- ri-la\' R-i roli-ases, 

 nirront will pass throut;h all the windinjjs on lM)th rt'la\s which in 

 turn rlosi-s the relay R*. This opens the circuit through the driv- 

 ing coil of the clock. The impulse given to the pendulum is, theri-- 

 lore, the duration of the operation of the relay /?:, or the time during 

 whii'h the light is cut off the photo-electric cell during the swing of 

 the pendulum to the left. When the pendulum swings to the right 

 and the rela\' R-^ o(H>rates, Ri is short-circuited and releases, Rt being 

 held up by winding H'l. When /?; releases, it releases /?< bringing 

 the circuit back to normal. Since the circuit through the driving 

 coil c)f the clock is closet! only when the relay R3 is dosetl, and the 

 relay Ri is released, there is only one impulse per second given to 

 the pendulum. 



During a periotl of operation by this nnithod covering several days 

 the clock gave as satisfactory performance as with the mechanical 

 ilrive, but while the present gear train is connected to it, no appreci- 

 ably Ix-tter ix'rformance can be obtained than at present, and ac- 

 cordingly it is proposed to carry out further work along this line 

 with an axperimental pendulum having no mechanical connections. 

 By using a good compensiited pendulum and mounting it suitabh' in 

 a constant temperature hermetically sealed case, it appears probable 

 that a photf)-electric cell dri\'e would produce a more constant rate 

 of oscillation than the l)est clocks of existing types. The advantage 

 of this tyf)e of drive over other types is the fact that the pendulum 

 is absf)lutely free from all mechanical constraint at all parts of its 

 swing. The problem of supplying an uninterrupted current for the 

 light and power could readily be solved by the use of duplicate ap- 

 paratus. 



The general method outlined in this paper for synchronizing a fork 

 with a clock has a ver>' wide field of usefulness, and is not limited to 

 the particular application described. For instance, in place of the 

 clock we may substitute another fork and distributor, and we are 

 thus enabled to hold 2 forks with their distributers in exact syn- 

 chronism by means of an impulse transmitted at a constant time 

 interval of aljout once every half-second. 



By substituting the field coils of a motor for the damping wind- 

 ing on the fork, we are able to hold the sp>eed of the motor in syn- 

 chronism with the clock, the only requirement being a step down 

 gear on the motor to furnish the desired contact. 



The general principle involved is not dependent on the use of a 

 vacuum tube, and if other means of control based on this principle 

 be adopted, very large powers may be controlled in the same way. 



