326 Prof. F. Rosetti's Experimental Researches 



sun investigated with any degree of accuracy. In so doing I 

 hope to overcome one of the chief difficulties which the problem 

 presents. Like my predecessors, I have been obliged to have 

 recourse to an extrapolation in order to obtain the value of 

 the solar constant within the limits of our atmosphere ; but the 

 method which I have followed renders it much more precise, 

 and I think that the numbers which I have obtained represent 

 the temperature of the sun in a satisfactory manner. This will 

 be rendered evident by a perusal of the present memoir. 



I. Description of the Instruments employed in the Researches. 



1st, Thermoelectric Piles. — In my experiments I have used 

 two piles — one made by M. Duboscq, and the other by M. 

 Gourjeon. 



The one made by M. Duboscq (No. 1) was composed of 

 25 antimony-bismuth elements arranged in the form of a 

 straight prism with a square base ; the junctures of the metals 

 were placed at the base of the prism, and were covered with 

 lampblack. The first rod of bismuth and the last of antimony 

 were in communication with two insulated coils of wire in which 

 the rheophors were fixed. The pile was protected by a small 

 brass case with double wall (fig. 1) in which were two apertures 



opposite to two sides of the pile. The pile was situated in 

 the centre of the case and was kept in position by a piece of 



