^^ TIiMEKEEPERir. 



tract or become smaller by cold, and instead of compensating 

 the error of the spring, it would add to it.) 

 Adjustment of The balls gg, being made of equal weight, may be placed 

 movin^^the ^^ ^^ ^^^ ^^^ of the taps at e, and if the timekeeper, being in a 

 pieces "ttt the situation where the thermometer will rise to 100 degress 

 amg"* ^^"'' or more, should go faster than when placed in another situa- 

 tion where the th°;rmometer will fall to 32 degrees or lower, 

 it is a proof that the expansion pieces do too much, and that 

 the balls are too heavy. Supposing this to be the case, sc.ev/ 

 the balls up close to the ends of the expnnsion pieces at ii\ 

 and their effect will be less; because, notwithstanding the 

 same degree of heat will occasion the expansion pieces to move 

 inwards, the same quantity, or to describe the same angle 

 from c, yet the balls will move through Ics space at d than at e. 

 For it is evident, that if they could slide up to the ends of the 

 expansion pieces, next to the arms of the balance, they would 

 not move at all, or, at least, their motion could not be disco- 

 vered by any effect that it would produce. If the timekeeper 

 still gains in heat, reduce the balls, and screw them home 

 ^gain to d. In the next trial, should it lose in heat more 

 than in cold, contrary to what it did before, it is a proof that 

 the expansion pieces do not do enough, and the balls must be 

 imscrewed toward the ends of the taps at c, until it keeps 

 the same time in heat a's in cold. If the balls being at the 

 ends do not do enough, and the timekeeper still loses in heat, 

 increase their size until the adjustment is brought within 

 the compass of the length of the taps, where there is gene- . 

 rally room sufficient to correct for a minute of difference 

 in heat and cold per day. By screwing the balls up and 

 down, it may be soon seen how much of error two or three 

 turns will correct in a given time, and by that means discover 

 their proper situation. 



Of Positions, or the Mode of adjust ivg the Timehecper to go 

 alikct or nearly so, in dlferent Positions. 



Method of ad- The long and short vibrations being adjusted, and also the 

 lauce 'so as to' ^^^^^ *"^ cold, I shall next shew how to adjust the different 

 make equal vi- positions. Let US suppose that the two mean time screws 

 hh,'V\ale III, Fig. X, when the balance is at rest, stand at those 



points 



bratious lu all 

 poiiitions. 



