CH Experiments made to ddcrmlne 



Avith comparative facility, and is worthy of the greatest reliance, 

 if the expansibility of the standard substance has beei* well de- 

 termined. 



The tabulated expansions of few solid bodies can be relied git, 

 on account of the great change in cxpannon caused by the 

 smallest admixture of any foreign substance. Mercury, from its 

 use in the construction of thermometers and barometers, has 

 been examined with considerable care ; yet, among the reported 

 expansions of the fluid metal, there are great differences, mostly 

 to be referred to inaccuracy in the determination of the glass 

 vessels in which it was contained. That account of its variation 

 in which we feel inclined to place the greatest reliance, is given 

 by Laplace in his Systeme dii Monde, where he states it to be 

 100 parts in 54.12, or 1&477 in a million, from the temperature 

 of melting ice to that of boiling water. 



The rod with which we compared the marble was of glass 

 tube, the rate of whose expansion we determined in the foUow- 

 ino- manner. Of a portion of the tube we formed a vessel with 

 a capillary stem. This vessel we filled with recently distilled 

 mercury, when at the temperature of melting ice : and after- 

 wards subjected it to the heat of boiling water, carefully collect- 

 ing all the expelled mercury. The weight of the mercury eject- 

 ed was 66.9 grainsj the weight of that remaining in the ve.'«sel 

 beino- 4312.6 grains, giving for the excess of the expansion of 

 mercury above tliat of glass .015-513 ; whence the expansion of 

 the glass is .002964 in bulk, or .000988 in length. 



This result is rather above the expansion usually given in 

 tables, and is liable to a previol^s error in determining the ex- 

 pansion of mercury. Yet a litde reflection will convince any 

 one, that the tendency of most of tl.e errors of observation is to 

 diminish the apparent expansions, so that an increased result 

 seems to afford some proof of the care with which the expansion 

 of mercury has been determined. The chance of error in the 

 weighings was very small, so that, in all probability, the expan- 

 sion named is very near the truth. 



The glass rod was made to serve as beam to a beam-compass, 

 and was subjected to the same change of temperature with two 

 slabs, one of white Carrara, and the other of black or Lucul- 

 lite marble. At the same time, to afford a check on the process, 



