1 7 * Thfrmmetrkal Flrfi movers. 



VI. 



Concerning tho/e perpetual Motions which are producible in Machines, by the Rife and Fall of 

 the Barometer, or the Thermometrical Variations of the Dimenfuns of Bodies, W. N. 



(Concluded from p. iz8.) 



OUPPOSE this apparatus to be put together at a certain temperature in the day-time; and 

 that in the night the temperature becomes colder; in this cafe, the curvature of all the bars 

 will diminifh, and the diftance between A and B will be increafed by the aftion of the inter- 

 mediate fpriiig : but as the plate A is prevented by the click C from receding, the plate B 

 will be pufhed forward, and the interior cylinder will gather a certain number of its teeth upon 

 the click D. The next day, when the temperature again rifes, all the expanfion bars will 

 bend, and the fpace between A and B will be diminifhed ; this, however, cannot happen by the 

 motion of B, which is held fall by the click D. The external part will,confequently, be now 

 .carried forward, and will aft upon the apparatus E: a fecond lowering of the temperature, by 

 whatever caufe, will occafion the interior part again to advance; and, in this manner, the 

 accumulations of force may be inceflantly reiterated. 



Experience muft determine how far the properties of thefe compound bars may be changed 

 in the courfe of time*. It feems probable, that the mere changes communicated by the 

 atmofphere, could fcarcely produce any fenfible efFedl; and whether this efFeft would be 

 detrimental to the general refult, may alfo be queftioned. Confiderations of this nature, lead 

 to the enquiry, whether our obje£l: may not, with equal facility, be obtained by the direft pufh 

 , or pull of bars of metal, as in the gridiron pendulum, or that of Ellicottf . 



If a fucceflion of bars of brafs were difpofed round a cylindrical face of lefs expanfible 

 metal, fo as to form an helical line from the one end to the other ; or, otherwife, if we fuppofc 

 a brafs clock chain, with a right-lined edge, to be wrapped round fuch a cylinder ; or again, if, 

 inftead of the cylinder, we fuppofe the chain to pafs over a fucceflion of rollers, whether dif- 

 pofed in a cylindrical fyftem, or according to the form of a pully, the refult will be the fame; 

 that is to fay, the chain will contract and expand about the ten-thoufandth part of its length, 

 for every ten degrees of Fahrenheit's thermometer. In the way of rough eftimate, let us, 

 therefore, aflume a cylinder of caft-iron, one foot in diameter, and one foot in length, 

 having a groove turned in its furface, like a fcrew, with twelve turns in the inch, for the 

 purpofe of lodging a fyftem of fridtion rollers to receive the brafs chain, wrapped round it. 

 Such a chain J, confifting of 144 turns, would meafure 450 feet§, and would contradt nearly 

 0,S4 inches for every 10 degrees, or one-twentieth of an inch for each degree of change of tem- 

 perature ; but as the cylinder itfelf contracts, the whole effedl will be fomewhat lefs than half 

 cthat quantity ; that is to fay, each degree of the thermometer will be one-fortieth of an inch. 



* Philof. Journal, I. 62. »f Ibid, p. 59, 60. 



■^ And would ccA about »5l. § For pyrometrical data, fee Philof. Journal, I. 58. 



The 



