APPENDIX. 



The following is Mr. Cavendifli's account of the cor- 

 redtlons to be made for Lord Charles Cavendifli's ther- 

 mometer. 



" The Thermometer ufed in thefe experiments is fully 

 " defcribed in the Philofophical Tranfadions, Vol. L. Page 

 "308; fo that I imagine it is unneceflary to mention it 

 " here. But fince the publication of that volume, the late 

 " Mr. Canton difcovered, that fpirits of wine and oth^ 

 " fluids are compreflible; which muft make the thermometer 

 ** appear to have been colder thaait really wasji. and renders 

 " a corre6i:ion neceflary on that account. There is another 

 " fmaller corredion neceflary, owing to the expanfion of 

 ** fpirits of wine by any given number of degrees of 

 " Fahrenheit's thermometer being greater in the higher 

 " degrees than the lower. As the method of computing 

 " thefe two corre£tions is not explained in that paper, it 

 ** maybe properjuft to mention the rule which was made. 

 " ufe of in doing it, 



*' In adjufting the degrees on the fcale of this thermos 

 '* meter, the tube was intirely full of Mercury, or the 

 " Mercury flood at no degrees on the fcale, when its real heat 

 " was 6 c° of Fahrenheit. Let the bulk of the Mercury con- 

 ** tained at that time in the cylinder be called M, and that 

 " of the fpirits, S; let the expanfion of fpirits of wine by 

 "1° of Fahrenheit, about the heat of 65°, be to its whole 



*'bulk. 



145 



