W. Ferret — Law of Thermal Radiation. 



11 



black and filled with mercury at different temperatures through 

 a range of 240°, was determined from the deflections of the 

 galvanometer needle of a thermopile. The third column of 

 the following table contains the deviations y of the needle, as 

 obtained by Rosetti,* corresponding to the absolute tempera- 

 ture T in the first column and the differences d between these 

 and that of the inclosure, 23 "8°, in the second column. 



T 



6 



y 



0-0 



40-4(l-0077 <5 -l) 



22-3(g 4 -l) 



19-6(g 4 - 2 -l 



329-6 



32-8 



100 



+ o-i 



— 1-5 



— 1-6 



-0-8 



369*6 



72-8 



29-5 



+ 0-8 



-0-8 



-1-8 



-o-i 



389-6 



92-8 



42-8 



+ 1-5 



+ 0-9 



-1-2 



+ 1-1 



409-6 



112-8 



55'0 



— 1-2 



-0-7 



-3-6 



-1-2 



429-6 



132-8 



72-5 



-1-2 



+ 1-0 



— 3-1 



-0-5 



449-6 



152-8 



91-5 



-2-3 



+ 1-4 



-3-5 



-0-9 



469-6 



172-8 



116-7 



o-o 



+ 4-9 



-0-7 



+ 1-8 



489-6 



192-8 



141-9 



-0-7 



+ 4-8 



-0-8 



+ 1-2 



509-6 



212-8 



169-5 



-2-1 



+ 3-1 



— 20 



-0-5 



529-6 



232-8 



204-0 



—o-i 



+ 3-2 



o-o 



+ 0-5 



549-6 



2528 



239-5 



-0-4 



— 1-3 



-0-7 



-1-6 



569-6 



172-8 



283-5 



+ 4-3 



-3-9 



+ 33 



+ 0-1 



The usual care necessary in such experiments seems to have 

 been taken. He says that the experiments were all made at 

 least twice, and whenever, between the first and second experi- 

 ment, a diiference of one, or at most two divisions were found, 

 and of five-tenths of a division in the lower temperatures, a 

 third and fourth experiment were made to obtain a correct 

 average. 



Rosetti devised an empirical formula to express the relations 

 between the deviations of the needle, y, and the temperatures, 

 which, expressed in our temperature notation, is 



(16) y=aT t d-bd 



in which 



a=0-00000335131 

 6=0-0636833 



This expression represents the observed values of y in the 

 table above with the residuals, O — C, in the fourth column. 

 But this expression, it is seen, ignores Prevost's principle of 

 exchanges, since it is not composed of two similar functions 

 for different temperatures, the one representing the heat radia- 

 tion of the heated body and the other that of the inclosure or 

 surroundings, as in the expressions of (2) and (7) of which 

 Dulong and Petit's and Stefan's laws are special cases. 



11. Applying Dulong and Petit's law to these observations, 

 the values of y are represented by the expression at the head 



* Memorie della Classe di Scienze Fisiche, Mathematichi, e Naturale della R. 

 Academia dei Lincei. Serie 3 a , vol. ii, 1877-1878. 



