Emission and Transmission of Heat 49 



bored for the insertion of the thermometers, thus producing points 

 of reduced sectional area, which must have had some influence 

 on the results. 



830. Since the numbers found by M. Despretz were relative 

 only, they could not serve for any applied calculation so long as 

 the absolute conductivity of no one of the metals was known. 

 In 1841 I took up the question and in order to determine the con- 

 ductivity in heat units I made a great number of experiments of 

 which a detailed description will be found at the end of this work; 

 I shall limit myself here to a brief description of the apparatus 

 employed and to the setting forth of the results. 



A cylindrical vessel full of water and surrounded by non- 

 conducting material was provided with a bottom formed by a 

 disc of metal of which the circumference was separated from the 

 walls of the vessel by a ring of cork. The lower surface of the 

 disc was heated by steam, and the heating of the water contained 

 in the vessel was observed. The water was kept in constant 

 motion by an agitator. If we admit that the amount of heat 

 traversing the disc is proportional to the difference of temperature 

 of the two surfaces, and remembering that this amount of heat is 

 proportional to the rate of cooling (776) we have dT=aTdt; 

 and m log. T=Ca t, m being the modulus of the table of loga- 

 rithms , 2 . 3025 and T the difference of temperature after the time /. 



If we designate the temperature corresponding to t=o by A 

 we find C=m log. A, since in this case T=A, and the formula 



m. 

 gives a= (log. A log.T) 



The logarithms are those of the tables and a is the cooling 

 which would take place in one second for a difference of tempera- 

 ture of one degree. Two observations gave a value of a, and the 

 identity of these values deduced from different pairs of observa- 

 tions proved the truth of the hypothesis on which the formula 

 was based. These experiments repeated with discs of lead, zinc, 

 tin, sheet iron, and cast iron, gave the same results. But what 

 at first astonished me very greatly was the fact that the values of 

 a were sensibly the same for these different metals, no matter 

 what their thickness was, although it varied from one to twenty 

 millimeters. In all the experiments, I noticed that the speed of 



