142 W. Ferret — Weber* 8 Law of Thermal Radiation. 



From (3) we get with the value of 273 Ce 273 above and put- 

 ting F=l, 



^S 37S>273 =0-452(g, 1 ^«l) (5) 



This expression with #=*0064, which was required to satisfy 

 the observations in the table of § 5 and the condition of (4), 

 we get 



^S 373 , 273 =0*7192 



This was denoted byH 100 — H in §7, paper A, and was found 

 from the experiments of Dulong and Petit to be 0*9092 by the 

 one formula and 0*8926 by the other. It was also determined 

 by Lehnebach directly from experiment to be 0*912. This 

 quantity, then, must be regarded as being pretty accurately 

 determined, and therefore the law of Weber, which gives only 

 0*7192 when applied to the same experiments does not hold for 

 the temperatures between 0° and 100°, even with the value of 

 #=•0064 instead of -0043, and indicates that the value of a 

 must be still greater for this range of lower temperatures ; and 

 this is in accordance with what we have already shown, namely, 

 that the lower the temperature, the greater must be the value 

 of a. 



10. While nearly the same values of a in Weber's formula 

 are required for all temperatures in Schleiermacher's experi- 

 ments in which the loss of energy by the wire is measured by 

 the electric method, in all other experiments in which it is 

 measured by the observed rates of cooling and by the ther- 

 mopile, the value of a required seems to be much greater, and 

 to increase with decrease of temperature. This may arise in 

 part from the uncertainty in the temperatures of the wire as 

 determined from the observed resistances at different tempera- 

 tures, especially for high temperatures, but a more probable 

 explanation is found in the want of a perfect vacuum in the 

 tubes through which the wires, heated by the electric current, 

 are stretched ; for the heat conduction from small wires is 

 enormous for all ordinary, and even very low, air pressures, 

 and so it is perhaps impossible to have a vacuum so nearly 

 perfect, and to maintain it during the whole time of the ex- 

 periments, that the amount of heat lost by conduction is so 

 small in comparison with that lost by radiation that it may be 

 neglected in comparison. * 



Putting 



A=the heat conducted from each unit of surface of the wire, 

 r x , r 2 =the radii of the wire and the internal part of the tube 



respectively, 

 T l9 r 2 =the corresponding temperatures at r 1 and r 2 from the 

 center, 



