ON PENDULUMS. X55 



two cycloids or parabolas a little differing In fize, than be- Properties of 



. ... ,, .," 11/-J lL pendulums, &c, 



tween the cycloid and large circle ; and bendes, as the cy- 



cloidal cheecks ufed by Huygens (which he found to fticceed 



fo well) were liable to the Tame alteration by expanfion, I 



looked on his numerous experiments as confirmations of this 



opinion. 



In reply to the obfervation relative to the fuperior effect of 

 pendulums detached as much as poffible, I beg leave to ob- 

 ferve, that a movement interrupted or intermitting (fuch as 

 more or lefs takes place in clocks with ofcillating pendulums), 

 is fo very different from one which continues equally without 

 any interruption (as that of the circular pendulum), that it 

 does not feem to me eonclufive to argue from the effects of 

 one to thofe of the other. 



In the note relative to the circular veffel enclofing the mer- 

 curial tube, I am inclined to imagine you were not aware that 

 this vedel was directed to be fattened to the i'pindle fo as to 

 revolve with it, by which means there would be no lateral 

 motion given to the air but by the friction of the outfide of 

 the cafe ; for which reafon I mentioned that it fhould be made 

 very fmooth externally : I alfo directed it to be covered, to 

 prevent the effect of the current of air, which would otherwife 

 pais through it, caufed by the centrifugal impulfe of the cir- 

 cular motion. 



As to what I obferved about the fafpending fpring, I per- 

 haps expreffed myfelf foo generally : I ftill think, however, 

 I could point out feveral inftances of gridiron pendulums, 

 where the compenfations were made of one bar againft an- 

 other only, without allowing any compenfation for this fpring. 



I think it neceffary to mention among the emendations of 

 my former paper, that the cycloidal bed of the rolling pen- 

 dulum, propoied in it, fhould be formed of the fame thick- 

 nefs in every part, as it might expand unequally if made 

 thicker in one part than another : As a trial of this fort of 

 pendulum may be made with a common clock at a fmall ex- 

 pence, I hope it will induce fome one better fkilled in fuch 

 experiments than I am, to make it, even if the others fhould 

 not be tried : and as a farther reafon for expectation of fuccefs 

 in tiie trial, it fhould be confidered (in addition to what has 

 been advanced in fupport of the opinion, that the effect of the 

 cycloidal bed would be but little altered by its expanfion), 



that. 



