INVENTION OF CHRONOMETERS, &LCk 283 



that, if he had been in possession of it before he began this ' 

 part of bis labours, he would probably have made use of the 

 same principle, but applying it in a very different manner, 

 tidelicei, to enlarge by cold, and contract by heat, the circum- 

 ference of the balance, preserving the spring untouched. He 

 proceeds to state how a balance may be constructed with com- 

 pound metallic pieces, to effect the compensation, and gives a 

 figure to elticidate his plan, of which Fig. 2, PI. VJI. is a— but he give 

 copy; adding that, according to this method, the balance ^^^^"^'J^^^^f^^^j J 

 maybe easily adjusted by means of small balls, or weights, invention on 

 which are to be attached to the ends of the metallic curves. The onhat'curb^- 

 curves, being made of two j)lates of different metals, with the the arms or 

 one most affected by variations of temperature at the outside, ^^t^bv flexure 

 it is clear, that heat will move the balls placed at the ex- and are adjust-- 

 tremities towards the centre, and that cold will move them i» ^eigjiw.^^ ^'^ 

 the contrary direction ; producing, by this contraction or ex- 

 pansion, the same sort of compensation as that of the mercu- 

 rial thermometer explained before. P. le Roy did not re- He proved the 



main satisfied with the simple suggestion of this contrivance, ^^'^•^ °^ ^"'-'^. 

 , 11 • ■ 1 oo j^j.jjjg jjy expsri- 



but actually put it in practice; and employing a register, suchmentj 

 as is represented in Fig. 3, ascertained by experiment that the 

 mechanism performed well, and corresponded pretty exactly 

 with the other thermometers. After all these investigations, 

 he concludes, by giving the preference to his own in«rcurial —but gave the 

 thermometer, because he thinks it more accurate and steady, meicurial com- 

 as well as more fit to secure, in all temperatures, an uniformity pensation. 

 of weight to the whole circumference of the regulator, than the 

 compound metallic balance; and, under those points of com- 

 parison, he may be right in his choice ; but certainly the last 

 thermometer seems better adapted for small and portable ma- 

 chines, and has answered, during repeated trials, so well, that 

 we must believe it fully eiititled to the favour which it iias 

 obtained in practice. 



The compensations in the balance, applied at present to the Peter k Koy's 

 best chronometers, are essentially the same as that so well ex-g'^^g"^^*^,"'^^^^ 

 plained and published so long ago by P. le Roy ; but, in this used: but 

 country, the invention has been generally ascribed to the i'te y^"|.| "^g, 

 Mr. Arnold, who, in 1782, took a patent for it; and a degree blared in the 

 of merit has been attributed to him on that account, proper- Fjfgnch govern- 

 tioned to the supposed difficulty of the desideratum expressed "™ent and ac«- 



/-\ a , demy in 17*^6, 



^ ^^ by Arnold toek au 



