ON TIMEKEEPERS. 201 



pensate not only for the expansion produced by heat, but also for the dimi- 

 "nution of the elasticity of the spring. Sometimes also a plate has been ap- 

 plied in such a way as to shorten the spring when the temperature is increased, 

 by an operation similar to that which serves to regulate a common watch, 

 the clip, that determines the effective length of the spring, being moved 

 backwards and forwards; and a similar effect has also been produced by 

 dividing this chp into two parts, one of which is fixed to a compound 

 plate, and is made to approach the other so as to confine the spring more 

 narrowly, and thus diminish its length, upon an increase of temperature. 

 (Plate XVI. Fig. 211.) 



Tlie flexure of a compound plate has also been applied in a simple and ele- 

 gant manner by Mr. Nicholson to the pendulum of a clock, by causing it to 

 support the upper extremity of the pendulum. The plate is placed horizon- 

 tally, the brass being uppermost, and carries the pendulum in the middle, 

 while the ends rest on two fixed points, of which the distance may be ad- 

 justed with great accuracy, so that when the temperature is increased, the 

 curvature of the plate may raise the rod of the pendulum, enough to keep 

 the weight or bob at a constant distance below the fixed point, which deter- 

 mines its upper extremity. (Plate XVI. Fig. 212.) 



The resistance, opposed to the motion of a pendulum by the air, affects in 

 some degree its velocity, and the variation of the density of the atmosphere 

 must therefore also produce some irregularities in timekeepers: they are,how- 

 ever, too small to be sensible. Derham found that the resistance of the air 

 accelerated the motion of a half second pendulum about four vibrations in 

 an hour, by diminishing the arc in which it vibrated: and when the vibra- 

 tions were restored to their original magnitude, the resistance of the air pro- 

 duced a retardation of eight vibrations in the same time. But a heavy pen- 

 dulum, vibrating in a small arc, is very little affected by this resistance. 

 i 



Besides these more essential parts of the watchmaker's art, there are se^ 

 veral subordinate considerations which require his attention ; the striking part 

 in particular occupies, in clocks, and in repeating watches, no inconsiderable 

 portion of the bulk of the machine. But the apparatus employed on these oc- 

 casions requires neither refinement of invention nor delicacy of execution. 



VOL. I. D d 



