August 25, 1887] 



NATURE 



393 



of Chigvvell, and Fig. 2 ^ shows his plan. 2 may be con- 

 sidered the same as the square to which the key would be 

 applied in winding, and the wheel A is fastened to it and 

 is geared to B, which in turn engages the pinion 3, which 

 is part of the stalk 4, passing through the pendant 8, and 

 terminating in the crown-piece 5. On turning 5, A will 

 revolve, winding up the watch in doing so ; the clicks 6 6 

 prevent A from returning. It is now clear why the fusee 

 must be dispensed with. With a fusee (whilst the watch 

 is going) the square which you wind travels backwards, 

 and it would naturally turn the crown-piece in doing so ; 

 this latter, meeting with resistance in the pocket, would 

 obviously stop the watch. Fig. 3 shows the mechanism 

 which takes the place of the fusee. It will be seen that 

 the main wheel is attached to the barrel ; the shaft (squared 

 at its extremities) which passes through the barrel is con- 



nected with the main-spring ; when the shaft is turned 

 the main-spring is wound. The shaft, being held by the 

 intervention of the clicks, cannot return, and the out- 

 side of the barrel being urged to follow it by the pulling 

 of the main-spring, impels the main wheel and drives the 

 train. Overwinding is prevented by means of the star 

 wheel and finger-piece shown in the diagram. Every 

 turn of the shaft causes the star to move on one division, 

 but on passing the last division the circle, out of which 

 the finger is cut, meets a convex instead of a concave 

 surface, and further movement is arrested. There is much 

 less difference between the pull of the main-spring when 

 the watch is wound up and nearly down than might be 

 expected. To obtain as much uniformity as possible a 

 long thin main-spring is used, it is tapered, and very few 

 turns of it are brought into service. 



Fig. s. — Chronograph with Swiss Keyless wjrk. 



About twenty-five years elapsed before any satisfactory 

 method was established of causing the keyless mechanism 

 to set the hands of the watch in addition to winding 

 it. The method which was first adopted had the draw- 

 back that the hands could not be put backward when the 

 watch was fully wound. At present two systems are 

 principally employed, they are known as the English and 

 Swiss. Fig. 4 shows the English or rocking-bar plan. 

 Wheel b is in connexion with the crown-piece, and com- 

 municates with the square of the shaft passing through 

 the barrel by means of the wheels / and g. Wheels / 

 and h are on a lever, or rocking-bar, pivoted about the 



' We are indebted to Mr. David Glasgow, Vice-President of the British 

 Horological Institute, and the Messrs. Casseli for the use of Figs. 2, 3, and 

 4. and to Mr. F. J. Britten, Secret.-iry of the British Horological Institute, 

 for the use of Figs. 5, 6, and 7. 



centre of d. ;> is a push piece acting against c, which is 

 a part of the rocking-bar. When / is depressed by the 

 finger or thumb, it lifts / and forces down h into con- 

 nexion with /, which communicates with the pinion of the 

 minute-hand c. If the crown-piece now be turned, the 

 hands will follow ; no winding is performed, because / has 

 been lifted away from g. When the pressure is removed 

 from/, a spring, s, puts the rocking-bar back again into 

 its normal position, / engaging g, and h quitting /. The 

 Swiss system is different in this : that connexion with the 

 winding or set-hands wheels is made by a pinion faced 

 with teeth on both sides, sliding up and down the stalk 

 of the crown-piece. The normal position (as in the 

 English system) is engagement with the winding-wheels, 

 but when the push piece is depressed the pinion moves 

 away from its engagement with the winding-wheels, and 



