OF ARTS AND SCIENCES. 



345 



For instance, an iron strip 7 mm. long, 0.17G mm. wide, and 0.004 mm. 

 thick, was made one arm of a Wheatstone's Bridge, and, with a battery 

 of one gravity cell, the successive resistances of the strip were measured 

 as its temperature altered, while the currents through it were made to 

 vary by introducing definite resistances in the circuit. Then having 

 the measured resistances of the strip, from the approximate formula 



t ■=. — ^^^ (where R is resistance of iron at temperature t in Centi- 

 .UU-J: /• ^ 



grade degrees, r the resistance at 0°) we obtain the temperatures which 

 are given below in the fourth column. The temperature of the room 

 was 27° C. 



"We see from the above that, when the temperature of the strip is 

 raised very little above its surroundings, a change of one-hundredth 

 Weber in the absolute current will raise its temperature less than half 

 a degree ; but that when it is raised more than two or three degrees 

 above the surrounding temperature by the current, such a small in- 

 crease of that current is accompanied by a greater rise in the tempera- 

 ture of the strip, and when the temperature of the strip is considerable, 

 though not excessive, the same change of .01 Weber will raise this 

 temperature by eight or ten times the former quantity ; and hence (as 

 it is important to notice) strong currents, and consequent high tempera- 

 ture in the strip, though giving larger galvanometer deflections, involve 

 a yet greater increase of the probable error of an observation on the gal- 

 vanometer, caused apparently by air-currents about the heated strip. 



A number of experiments with a similar iron strip (resistance 0.9 

 ohm) in a Wheatstone's Bridge (whose other arms were 0.9, 0.4, and 

 0.4 ohms) showed that with a half-ohm galvanometer a deflection of 

 about 204 divisions could be obtained by exposure to lamp radiation 

 as before described. The total current was 0.58 Weber ; and, as one 

 division of the galvanometer scale corresponded to about .000 0002 



