6o Ciiiiip,/if,>t'tc:}s fir Change of Temperature in Penduluim. 



two, connetlcil ^t fome didance afunder by a crofs piece above and below, at tiicir ex- 

 tremities, forming a parallelogram. Within and nearly contiguous to thefe were placed 

 two brafs bars, pinned to the lower crofs piece, and proceeding upwards to another crofs 

 piece which connected their fuperior extremities. From this laft piece hung a fecond 

 pair of (leel bars, within and nearly contiguous to- thofe of brafs. This laft pair was alfo 

 conntded below by a fep.irate crofs piece, like the firft, from which a fecond pair of brafs 

 bars proceeded upw.irds, as before, and were connected at top, like the former pair of the 

 fame metal. And lalUy, from the upper connecling piece of thefe brafs bars defcendcd one 

 fmgle bar of (leel, which pifTed clear through the lower framing, and fcrved to fufpcnd the 

 lens or ball. In this ingenious contrivance it is evident that fhe bars of the fame metal, 

 which were ufcd in pairs, performed the oflice of fingle rods, excepting that they mutually 

 fupported each other, fo as to need no bracing. It is Hkewilc clcir that, fiiice the pairs of 

 brafs and ftccl bars compared with each other are nearly equal in length, the repetition muft 

 be more frequent, in order to countera£l the expanfion of the fingle fleel bar in the 

 middle, in proportion as the difference of their expanfions is lefs. Thus Harrlfon and the 

 earlier artifts cut the (leel bar into three parts, and the brafs into two, which, with the 

 doubling of the three outer bars, formed a gridiron of feven bars. But at prefent, by fub- 

 flituting a mixture of zinc and filver in the place of the brafs, the expanfion ot the com- 

 pound rod is found to be fo great, that two pieces of fleel and one of the zinc and filver 

 may be conveniently ufed, which contrivance affords a compound bar of five pieces only. 

 This is the form generally ufed. 



The gridiron pendulum defervedly pofieires a higher degree of reputation with aflrono- 

 mers than any other contrivance upon the fame principle. Cut a confiderabie number of 

 pendulums has been made upon another conftrutlion, known by the n.ime of its inventor, 

 Mr. John Lllicott. This artift formed liis pendulum-rod of iron ; but inflead of attaching 

 the hall to this rod, he caufed it to reft upon the outer arms of two fliort levers, moveable 

 upon centre-pins fiiced in the iron bar. The inner ends of thefe levers were atled upon 

 by the lower extremity of a bar of brafs of the fame figure as the pendulum-rod againft 

 which it was applied, an<,l to which its upper extremity was attached. From this arrange- 

 ment it followed that, upon an increafe of temperature, the brafs b.ir, being lengthened more 

 than the iron, would caufe the inner tails of the levers to dcfcend more than the fiilcrum- 

 plns, and confequcntly would raife the outer tails, and along with. them the ball. By the 

 ingenious application of fcrews to adjuft the bearing parts of the ball upon the levers, the 

 quantity of afcent in the ball itfelf was fo regulated as to counterad the effe£l of lengthening 

 in the iron bar, which is the true pendulum-rod. 



The objctlioUs made by curious reafoners to this contrivance arc, that the levers may be 

 prevented by fridlion from obeying minute alterations of temperature; and that the two 

 bars may be bended during the increafe of temperature. Mr. Gumming, in his work before 

 quoted, has confulerably improved this invention, though, after all, it is liable to more 

 objetlions than the gridiron. Its chief merit appears to confift in the facility with which 

 it may be adjuftcd to temperature, with little or no alteration of the adjuftmeut for time. 



Several writers on this fubjccl have ftrongly infiftcd, that every compenfation for tem- 

 perature, whether in a balance or pendulum, ought to move through tiie air along with it, 

 iji order that the fcveial parts may be heated or cooled together. To this it has been 



anfwered, 



