NEW PENDULUM. 229 



To remedy this evil I contrived the one here treated of, and —is prevented 

 m my own judgment, have fucc^eded as near perfection as^^i^gg; 

 any effort of mine ever did. The tubes are drawn flraight 

 and regular in an engine, in the manner that wire is done : 

 and their ftrength alTure fuccefs, preventing the poffibility of 

 any fenfible bending, while the freedom of a6lion prevents any 

 irregular motion whatever. 



I have no farther quarreled with my own conftru6iion of the Troughton's 

 mercurial pendulum, the principles of which are faultlefs, and JJ,^iy^^"j perf^£l 

 form elegant; than that it is quite unfit for carriage, and an but not porttble. 

 article too hazardous for me to deal in ; but after all, candour 

 obliges me to give the preference to the original one of Gra- 

 ham, which on account of fome fenfelefs criticifms about one 

 part taking heat fooner than another, evaporation of the mer- 

 cury, &c. was too haftily laid afide, Graham undoubtedly Graham's pen- 

 made his pendulum before (he expanfions of either brafs or ^"'""^ ^PP'^°^^'^- 

 mercury were accurately known; but an ingenious friend of 

 mine, has from the beft experiments, computed the different 

 parts to mathematical precifion, and introduced it a fecond 

 time to the world ; and already, tj^ere are not wanting proofs 

 of its pradical excellence. 



Smeaton's pendulum with tubular bob is a good one, but Smeaton's 

 on account of the glafs rod, like my mercurial one, is totally goo^i ^^^"ot 

 unfit for tranfportation, and like Graham's, is unfightly; while 

 all three labour under the unavoidable difadvantage of not 

 being applicable to the pyrometer in their compound and pen- 

 dant fiate. 



I am about to conftru6l a pyrometer fit to try the tubular 

 pendulum in its finifhed flate; a pin fixed in the centre of of- 

 cillation, (liall be contrived to ad on a nice fpirit level, and ' 



ftiew the expanfion, if any, in angular meafure : by which I 

 fliall be enabled to adjuftit; for I have no idea of adjufting 

 one by the going of a clock, that complicated apparatus will 

 always, I fear, have too many errors of its own, to be ufed as 

 a criterion of the merits of this more perfe<fl inftrument, 



I will here fubjoin the lengths of the different parts of tu- Dimcnfions ani 



bular pendulums in inches; alfo, the expanfions of brafs and i""®""^^ °^/*y** 



^ 1 ftruCtion of the 



fteel from which I computed them. The length of the inte- tubular pcndu- 



rior tube is 31,9, and that of the exterior one 32,8, to which ^**'"* 



muft be added 0,4- the quantity which in this pendulum, the ; 



centre of orcillation is higher than the centre of the bob; thefe 



all 



