COMPENSATION PENDULUIVi. 



rods together while the adjusting screw ia removing, taking 

 care to release it again afterwards; and 4thly, That it can 

 be maaufai'tored for less expense than any other compen- 

 sation uenduluni hitherto published. 



N. B. The compensation of this pendnlnm which I now 

 send to the Society of Arts is properly adjusted, at least 

 very t ear the truth. The holes for the adjtistinL' screw are 

 made at snch a distance from each other, that by removing 

 th.e screw one hole, it will produce an alteration in the go- 

 mj; of the clock of about a quarter of a second per day 

 with a change of thirty degrees of Fahrenheit's thermome- 

 ter. 



59 



SIR, 



PERMIT me to slate to you the observations I have 

 made since my compensation-pendulum was laid before the 

 Society. 



The regulator, with the hammered zinc rod, and ball of P<^nfliilumwith 

 forty-six pounds weight, was firmly fixed to a brick wall at Q^^the lower 

 the top of my house. The adjustment of the length of extremity. 

 the rods, by means of a lamp, was repeated as before. 

 There was, however, an alteration necessary to be noticed; 

 the ball of the pendulum rested on its lower extremity, in- 

 stead of being suspended bvits centre. 1 prefer this method, 

 as being less liable to errour, if the rods should be affected 

 by heat or cold qiiicker than the ball. The length of the Lengthofzinp. 

 zinc rod, as ascertained by the lamp, was now found to be 

 !l0| inches. 



The clock was then set to mean time, and suffered to 

 go without alteration; the result is exhibited in the foliovv- 

 inji' table. 





Errour of Clock at 



Number of Days 



Daily rate. R^te of going. 



1806 



time of observation. 



between ihe O'b- 

 seivatioiis. 



March 21 



0" 



18 



-f 0"- 15 



April 8 



+ 2- .8 



32 



+ 0. 18 



May 10 



+ 8-7 



16 



+ 0- 80 



26 



+ 21- 5 



26 



+ 1- 10 



June 21 



+ 50- 





Increased 



