254 MEASUREMENT OF CURRENTS. 



deflections of the two instruments to a convenient magnitude. If, 

 then, the strength of the current is varied, the indications of the 

 graduated instrument will give those of the other. Yet we should 

 , avoid, as much as possible, the use of an auxiliary apparatus. 



871, In experiments on radiant heat we seek to obtain de- 

 flections which represent numbers proportional to the quantities of 

 heat which fall in unit time on a thermoelectrical pile placed in 

 the circuit. The mode of graduation should, in that case, be 

 relative to the quantities of heat, and it only represents an elec- 

 trical graduation if the intensity of the current is proportional to 

 the calorific radiation on the pile, or, more exactly, to the difference 

 of radiations which fall on_the two faces. 



The method used by Melloni* consists in placing on each side 

 of the pile constant sources of heat S and S'. On one of the faces 

 the radiation from the source S is allowed to fall, and a deflection 8 

 is observed. The action of the source S being suppressed, the 

 source S' is allowed to act on the other face. This gives a de- 

 flection 8' opposite that of the first. Finally, the two sources acting 

 simultaneously give a deflection 8", for instance, of the same sign 

 as the first. If we have 8" = 8 - 8', it may be assumed that the 

 deflections are proportional to the radiations. In most of the 

 galvanometers constructed on Nobili's plan, this proportionality 

 holds up to 20 or 25 degrees; beyond that 8">8-8', and a table 

 of graduation must be constructed. If the deflection 8 alone is 

 outside the limits of proportionality, we have 



the observed deflection 8 corresponds to the graduation 8' + 8", and 

 so on from step to step. 



We may also place the two sources on the same side of the 

 pile, and make them act at the same angle ; or even with a single 

 source make use of the law of the square of the distances. M. P. 

 Desaines uses a single source in front of which he places a screen 

 perforated by an aperture consisting of four equal sectors. When 

 one, two, three, or four of the sectors were left open, he obtains 

 calorific radiations, which are in the ratio i, 2, 3, and 4, which 

 he causes to act successively on the pile. If the experiments 

 are repeated at different distances, we shall have all the elements 



* MELLONI. Ann. de Chim. et de Phys. [2], Vol. LIII., p. 5. 1833. 



