bf ' denfity ihould be at 6** of the decimal fcale : but be has 

 taken in this calculation the mean rate of expanfion from the 

 freezing to the boiling point, inftead of the expanfion at tb« 

 boiling point, which, upon the fuppofitton, is twice as great; 

 hence the apparent maximum, on the fame grounds of calcu- 

 lation, ftiould be at 3® of the decimal fcale, or at 37.4^ of 

 Fahrenheit* Mr. Dalton attempts to avoid this difficulty by 

 fuppofing that the thin bulb of a thermometer expands more 

 than the glafs tube employed by General Roy. 



The fame ingenious author has made experiments on the TElaftlcity of Imn* 

 elafticity of different fubftances, by meafuring the depreffion '^^* 

 of -the middle of a bar fupported at its extremities; but if I 

 am not much miftaken, his inferences from them are by nO 

 means accurate. Mr. Leflie gives 67 1625 feet for the height 

 of a column of deal equivalent to its elaflicitj' ; the true height 

 refulting from his experiments, taking into confideration alfo 

 the inequality of curvature, appears to me to" be 4664000 

 feet; which is ftill little more than half as much as would bfe 

 inferred from the experiments of Chladni on the longitudinal 

 founds of fir wood. There mufl: have been fome inaccuracy 

 \v\ Mr. Lellie*s experiments on fieel differently tempered : for 

 it appears from the direct experiments of Coulomb on the 

 flexure of bars, as well as from thofe of Chladni, and fome 

 of my own, upon found, that the ultimate elafticity of fteel 

 in fmall tenfions is the fame, whether it is harder or fofter. 

 This may appear at firft fight paradoxical, but it admits of a 

 fufficient explanation, which, together with many other illuf- 

 trations of various parts of natural philofophy, will probably 

 before long be laid before the public. 



Your very obedient fervant, 



THOMAS YOUNQ, 

 J^^Mbeck Street, Sep, 22, 1804, 



Vol. IX.— OcT«BfiR, 1804-. I Vefcriptiort 



