ON HEAT AND LIGHT. ]fjl 



To form our judgment with eafe, and as it were at a tingle 

 glance, of the agreement of our theory with the refults of the 

 experiment, of which I gave an account at the beginning of 

 this memoir, we have only to reprefent thefe refults by a fi- 

 gure in the following manner. 



On the horizontal line A E, Fig. 3. reprefenting the axis of Conftru £i l) * of 

 the cylinder employed in the experiment, we will take three ing by a curve> " 

 points, B, G and D ; one, C, in the middle of the axis, being and its ordinate* 

 thefituation pf the central thermometer, the other two, B and a&uaU^obkrv- 

 D, at the intermediate points which the other two thermo- ed. 

 nieters occupied between the middle of the axis and its two 

 extremities; 



Erecling the perpendiculars A/, B^, C A, Di, and E&, on 

 the points A, B, C, D and E; and taking the ordinate A/== 

 212, the temperature of boiling water; Bg:=162, the tem- 

 perature indicated by the thermometer B; C /j=132|;, the tem- 

 perature indicated by the thermometer C; Di=106f, the 

 temperature given by the thermometer D; and laftly, E kz± ' 

 32, the temperature of water mixed with pounded ice; a 

 curve, P Q, palling through the points/, g, h, i,k, ought to be 

 the logaritlunic; that is, fupporing the temperature of the fur- 

 rouudingair to be conftantly at the temperature of melting ice 

 during the experiment. 



But the experiment in queftion was made when the temper- The curve has a 

 ature of the air was at 78° F. cor.fequently, reckoning fi'om[ oi j t a °J r contr8 " 

 a certain point, taken in the length of the cylinder, where the 

 temperature was at 78°, to the extremity E, the influence of 

 the furrounding air, inftead of cooling the furface of the cy- 

 linder, heated it; and it is evident, that the curve PQ mull 

 fteceflfarily in this cafe have a point of inflexion. 



In fact it appears on a funpie infpeclion of the figure, that It is likewife 

 the curve P Q has a point of inflexion ; but we fee likevvife, irre S ular * 

 that this curve is not regular. That branch which is concave 

 toward the axis of the cylinder is not fimilar to the adjoining 

 portion of the curve, of equal length, which is convex toward 

 that axis; as it ought to be according to our theory; and even 

 the part of the curve which is convex toward the axis A E, 

 differs fenfibly from the logarithmic, particularly toward its 

 extremity P. 



It ought neceflarily to differ from this curve, as far asthede- The deviation is 

 vifions of our thermometers are defedive ; but the deviation be- ^j ""^he de- 

 Vot. XII.— November, 180$* M tween feds of ther- 

 mometers. 



