412 Prof. F. Rosetti's Experimental Researches 



part of the pile received rays, not from the whole of the sphere, 

 but from a spherical segment, the thermal effect of which was 

 equal to the radiation which a circular disk of 21*42 millims. 

 diameter would produce if placed so that the back point made 

 a tangent to the sphere, i. e. at 170 millims. from the face of 

 the pile. By the law of inverse squares, we find that if at 170 

 millims. distance the thermal effect of the radiation of the 

 ball was 235*9 divisions of the galvanometer, then at 2296*5 

 millims. it should have been 1*322 division. 



There are two methods of comparing the results of the calo- 

 rimeter experiments with the galvanometer observations ; and 

 they yield similar results. We can either introduce the values 

 T and T — 6 (given by the calorimeter experiments) into the 

 formula, and obtain thus the value of y, in which case we 

 should have 



T=273 + 762-l = 1035-i; 

 (9 = 273+ 11-0= 284*0, 

 T-6>=751*1, 

 g ivin g y =1-3943; 



or we can equally well introduce into the formula the value 

 y— 1*322 given by the galvanometer observations, and deduce 

 from this the value of T and then of t. Doing this and re- 

 membering that the surrounding temperature was 11°*0, and 

 consequently 



0=273 + 11*0 = 284, 

 we obtain T=1019*0, £=746*0. 



Comparing the two values of y and of t, 



C. G. 



y= 1*3943 and 1*322, 



£ =762*1 and 746*0, 

 we see that they do not differ much • but quite enough to 

 show that the formula would not be applicable at high tempe- 

 ratures. It struck me that the slight difference might possibly 

 be owing to the emissive power of the red-hot ball, and that, 

 since the experiments which were employed for the calculation 

 of the constants of the formula had been made with a surface 

 covered with lampblack, all discrepancies would probably have 

 disappeared, if even at high temperatures the radiating surface 

 were covered with lampblack. I invented, therefore, a me- 

 thod of comparing the thermal effect of very hot bodies whose 

 surface is bare with the same bodies when equally hot but 

 with their surfaces covered with lampblack ; so that, on com- 

 paring the two values, I obtained a number representing the 

 emissive power of the plain body compared with the emissive 



