80 Mr. J. P. Joule on the Effects of Magnetism 



required to bring it to a horizontal position indicated the in- 

 tensity of the magnetism of the iron bar under examination. 



After a few preliminary trials, a great advantage was found 

 to result from filling the tube with water. The effect of the 

 water was, as Dela Rive had already remarked, to prevent the 

 sound. It also checked the oscillations of the index, and had 

 the important effect of preventing any considerable irregu- 

 larities in the temperature of the bar. 



The first experiment which I shall record was made with 

 a bar consisting of two pieces of very well-annealed rectan- 

 gular iron wire, each one yard long, a quarter of an inch broad, 

 and about one-eighth of an inch thick. The pieces were 

 fastened together so as to form a bar of nearly a quarter of 

 an inch square. The coil was placed in connexion with a 

 single constant cell, the resistance being further increased by 

 the addition of a few feet of fine wire. The instant that the 

 circuit was closed, the index passed over one division of the 

 micrometer. The needle of the galvanometer was then ob- 

 served to stand at 7° '2.0', while the magnetic balance required 

 0*52 of a grain to bring it to an equilibrium. It had been 

 found by proper experiments that a current of 7° 20' passing 

 through the coil was itself capable of exerting a force of 

 0'03 of a grain upon the balance ; consequently the magnetic 

 intensity of the bar was represented by O'-iQ of a grain. On 

 breaking the circuit, the index was observed to retire 0'3 of a 

 division, leaving a permanent elongation of 0'7, and a per- 

 manent polarity of 0'42 of a grain. More powerful currents 

 were now passed through the coil, and the observations re- 

 peated as before, with the results tabulated below. 



Experiment 1. 



