Homogeneous Bodies; 'with illustrative Experiments. 3 



the hard and soft ends. In cast steel the current in the bar 

 is from the hard to the soft part ; but in brass it flows in the 

 contrary direction. Fig. 1. (PI. I.) is the shape of the steel or 

 brass bar, which 1 find very convenient in these experiments: 

 the extremities, which are a little bent, dip into the cups of 

 the galvanometer. There is no occasion, however, to employ 

 a galvanometer with brass ; the bar, when heated at the bend, 

 may have its extremities brought into close contact, and one 

 side held over and parallel to the compass-needle ; and the 

 nature of the deflection will indicate the direction of the elec- 

 tric current. When the bar is of steel, the direction of the 

 electric current will be indicated by the arrow : when brass 

 is employed, the current flows in the opposite direction. 



9. I have magnetized the steel in all the various ways that 

 I could think of; but I have not found that its being a magnet 

 has any perceptible influence, either on the direction or the 

 power of the electric current excited within itself by heat. 



First Class of Experiments "with single pieces of Bisnndh. 



10. Bismuth is one of the metals wherein the magnetic 

 powers are finely developed by heat: the energies are 

 promptly displayed, even by small specimens cast into certain 

 forms ; and, as their character can be examined without the 

 aid of a multiplying galvanometer, the phaenomena are very 

 easily exhibited. Experiments on bismuth are therefore well 

 calculated to impress immediate conviction on the mind, as 

 to the distinguished and interesting character of the thermo- 

 magnetism of homogeneous bodies. 



11. The first piece of bismuth which I employed in these 

 researches was cast into the shape of a rectangular frame, not 

 very unlike the rim of an old-fashioned knee-buckle. Each side 

 of this frame was a rectangular prism, the faces of which were 

 each '3 of an inch broad. The sides, however, were not very 

 smooth, but no file was employed to level the inequalities. The 

 length of the frame outside measured 3'2 inches, the breadth 

 1-2 inch. Tlie experiments were made by iieating, at different 

 times, various parts of the frame in the apex of the flame of a 

 spirit-lamp : and, when any selected point was thus heated tor 

 a few moments, one of the longest sides was immediately held 

 over, and j)arallel to a delicately suspended compass-needle ; 

 on which, the magnetism of the earth was pretty accurately 

 neutrahzcd by means of a distant bar magnet: and the de- 

 flections of the needle were taken as an indication of the di- 

 rection iu which the electric currents flowed round the me- 

 tallic frame. 



12. The needle which I have employed is four inches long, 



furnished with an agate cap, and suspended on a fine steel 



B 2 poi'il i 



