210 M. Arago on the Thermometrkal State 



rent parts of the mass shall elongate themselves from the axis of 

 rotation. 



There are few who do not know that heat dilates all known bo- 

 dies, and that cold contracts them. The more then, that the 

 wheel, of which we have been speaking, is heated, the more will 

 it enlarge and extend itself; that is to say, the more will the ma- 

 terial particles of which it is composed, elongate themselves from 

 the centre of rotation. ,The contrary effect will manifest itself 

 during the decrease of the temperature. It follows then, that 

 under the action of the same force, a given wheel will revolve so 

 much the more rapidly as it becomes colder, and so much the 

 more slowly as it becomes hotter. 



In machines intended for exact measurements, as for example 

 in watches, the difference of the rate arising from the change in 

 the dimensions of the wheel, consequent upon the natural varia- 

 tions of the atmospheric temperature, are so great, that they 

 must needs be remedied. 



The moving power of all watches is a steel-spring rolled upon 

 itself, which incessantly acts upon the whole system of indented 

 •wheels, of which the watch is composed. It is to be observed, 

 however, that these wheels have not an even continued motion ; 

 it is sometimes interrupted; the stopping of the seconds-hand upon 

 each of the divisions of the dial plate, shews this. Well then, the 

 interval which elapses between two of these consecutive stops, 

 and consequently the duration of the second as told by the watch, 

 (and on this depends the duration of the minutes and hours) is 

 regulated by the time which a metallic wheel, called the balance 

 wheel, occupies in making a turn upon itself. If the principles 

 above laid down be true, then we may deduce from them, sup- 

 posing the watch regulated to the ordinary temperature, that by 

 heat the balance-wheel becoming larger, will oscillate slower, — 

 that the second will be too long, — that the watch will get slow. 

 By cold, on the other hand, the balance-wheel will go too ra- 

 pidly, it will stop the wheel-work at intervals too near each other, 

 the seconds will be too short, and the watch will go too fast. 



These different results are confirmed by experience. Those 

 watches in which no remedy is applied to the defects of the 

 balance-wheel, by the help of an arrangement, the description 



