[ 38 ] 



VII. On the Alteration of the Arcs of Vibration of Pendu- 

 lums by the Hygrometrical Changes of the Air ; afid on a Com- 

 perisating Pendulum of Deal, ajpplicable to general Use. By 

 Tho. Squire, Esq. 



To the Editors of the Philosophical Magazi7ie and Journal. 



Gentlemen, 



SINCE the time of that eminent mathematician, astrono- 

 mer and mechanic, Christian Huygens, great improve- 

 ments have been made in the construction of horological in- 

 struments, and especially in those most important parts re- 

 lating to the clock — I mean the scapement and the pendulum. 



It is well known that our atmosphere is often subject to 

 great and sudden vicissitudes with respect to pressure, tempe- 

 rature, dryness and moisture; and these causes, operating in a 

 greater or less degree upon the different parts of the mecha- 

 nism of the clock, must always have a tendency to render its 

 motion unequal, and for that reason unfit for the purpose of 

 showing the different portions of time with accuracy. 



To obviate the eff'ects arising from these different causes, 

 many ingenious contrivances have been adopted. 



The variations indicated by the barometer must in some de- 

 gree affect the arc of vibration, by the changes of buoyancy 

 and resistance ; but the error from this source cannot be great, 

 owing to the ponderous materials of which oxw best pendulums 

 are made, their slow motio)i, smooth surface, and lenticular 

 form. 



On the other hand, the effects arising from the hygrome- 

 trical changes of the air do very perceptibly alter the arcs of 

 vibration, they being considerably greater in a moist than in 

 a dry state of the atmosphere; hence, the clock will lose in 

 the ibrmer case and gain in the latter. This is a great evil, 

 and I am of opinion that the most perfect jewelling of the pal- 

 lets will not completely remove it. 



In registering several hundred observations of the semi-arcs 

 of vibration made by a clock having a dead scapement, going 

 fusee work, a motive power of 2 pounds, and a pendulum of 

 12 pounds, I have often found these arcs so variable, though 

 mostly within narrow limits, as to defy calculation, leaving 

 the mind in doubt whether to ascribe the variation in the rate 

 to the arc of oscillation, or an imperfect com])ensation. The 

 pendulum of this clock is so delicately hung, that a maintain- 

 ing power at the pallets of no more than f of a dram, 

 including the friction of the train, is sufficient to keep it in 

 motion. 



I have 



