104 Description of a Mercurial Pendulum. 



deep, that receive the edges of the side bars M M ; a piece of 

 white writing paper should be gumiiicd on upon the bottom part 

 of the Hd. I is the index at bottom, and SS a scale of inches 

 divided into tenths, and about -44 inches from C, the top of the 

 pendulum. Try on this scale the arc of escapement; and if it 

 should be l-,",y inch from e to e^ then apply such a pulley weight 

 that the pendulum shall vibrate from V to V, or Z-{'^ inches. The 

 diameter of the glass being 2 inches inside very nearly, and the 

 depth of the quicksilver 6y'^j- inches, the weight will be lOlbs. 

 fully, and therefore it will be best to buv 1 1 lbs, 



" From accurate calculation I find, that if such a pendulum 

 should go perfectly true when the thermometer is at 30% but 

 that at 90' it should go one secon<i slower in 24 hours, it would be 

 remedied by pouring in 10 ounces more quicksilver; or taking 

 out that quantity, if it went one second faster in 24 hours at 90% 

 than 30^ of the thermometer; and for y'^ of a second that the 

 clock may go faster or slower in 24 hours, its compensation is 

 obtained by putting in or taking out one ounce of quicksilver for 

 those temperatures : and this is all easily done, 



" Paradise Row, Islington, June 1804. GavIN Lowe." 



" Let the glass be wiped out with a large old clean linen rag be- 

 fore the quicksilver is poured into it. A paper funnel folded so 

 that the end of it will reach the bottom of the glass is best for put- 

 ting in the quicksilver. The extremity of its end should be closed 

 by wrapping a bit of thread twice or thrice round, and knotting it. 

 About 15 or 20 pricks should be made above the extremity of the 

 funnel with a needle of the size that carries six-penny thread. 

 During the tilling in of quicksilver the person that holds the fun- 

 nel should keep the pricks in the paper below the surface of the 

 quicksilver that is in the glass, raising his hand as it fills up." 



(Copied August 24, 1804.) 



T. FiRMINGER. 



On the Calculation and Construction of a Merc2irial Pendulum 

 for a Clock. 

 " The absolute quantity of expansion of the following metals 

 was determined by the late Gen. Roy : viz. 



" 10 inches of plate brass with 60^ of Fahrenheit^ Q.QAp3 

 expanded .. .. .. .. ..J ^ 



" 10 inches of steel 0038 



" 10 inches of white flint glass . . . . 00027 



" In 1792, an accurate experiment was made to find the dif- 

 ference between the expansion of white flint glass £ftid quicksilver; 



for 



