31srVE:NTI0N OF CHRONOiMETERS, &C. 299 



pieces which we have already quoted ; and we are ratlier sur- 

 prized to learn that it cost Mr. Earnshaw so much trouble to 

 discover it; but, according to his account, that property does 

 not answer completely the object in view, and watches, with Gradual decay 

 isochronal springs, well adjusted at first, will pi"ogressively J^" g J^j^^^,g"°*_^ 

 lose in their rates, from the relaxation which lakes place in serted by 

 them. The remedy for that evil Mr. Earnshaw declares him- '^^Q^*'*^- 

 self to have discovered, by a continuance of the same stub- 

 born application which he had bestowed on the former part 

 of his labours ; and that it consists in making the springs of 

 such a shape as to gain in the narrow vibrations, about five — v/ho disclo- 

 or six seconds per day more than in the wide ones. The Jhat^thVn-^row 

 reason of this contrivance is explained by the author in the vibrations 

 following passage of his statement (p. 10) : " I find the.com- ^lade the 

 *^ mon relaxation of balance springs to be about five or six quickest. 

 '■'■ seconds per day on their rates in the course of a year; 

 *■ therefore, if the short vibrations are made by the shape of 

 *• the spring to go about that quantity faster than the long , 



*' ones, and as the spring relaxes in going by time, so the 

 " watch accumulates in dirt and thickness of the oil, which 

 *■ shortens the vibrations, the short ones then being quicker, 

 *' compensates for the evil of relaxation of the balance spring." 

 But the whole of this explanation, though plausible at first This remedy 

 sight, seems liable to considerable doubts and objections, and 

 it would require a series of decisive experiments to prove the 

 accuracy of the method, and the certainty of its applica- 

 tion. 



Some skilful makers are not satisfied as to the reality of It may be 

 this relaxation of the balance spring, at least to the degree ^h^\i^°"^ 



which is implied in the above reasoning ; and none, we appre- springs do fall 



ofF ? 

 hend, will believe that the relaxation, supposing it to take 



place, will be so uniform as to admit of a remedy, fixed in all — or if S9, 

 cases to the same quantity of difference between the wide and fomiiy p""^'" 

 narrow vibrations. Is it probable that the efiects of the relax- 

 ation, which proceeds from wear, can be ascertained by any 

 other means than very long trials with each individual spring ? 

 Or is it to be supposed, that the progressive alteration of springs —or otherwise 

 is of such a nature, as to be concluded from a short experi- thanb"ex =ii- 

 ment, taking it for granted, that the variations will be always ment ? 

 in proportion to the times ? Some chronometers have been inetrrs"^ffahi° 

 •Q q 2 known on their rates; 



