Homoge?ieous Bodies ; tvith illustrative Experiments. 17 



which the electric currents flowed over the surface, there were 

 necessarily two longitudinal lines more active than any other 

 part. These lines passed through the dense and opposite 

 cavernous sides, and may be termed the lines of greatest 

 energy. The neutral lines were also a consequence of the 

 recurved flow of the electric currents by intersecting them at 

 right angles; and as those intersections were in a series of 

 points nearly parallel to the axis, those neutral lines were 

 nearly parallel to the axis also. 



57. When the other end of the cylindrical bar was heated, 

 the lines oi greatest energy were still on the same parts of the 

 surface, but the electric currents flowed in an opposite direc- 

 tion to the former; so that to whatever end of the bar heat 

 was applied, the current uniformly proceeded from the heated 

 end along the dense side c d, and returned over the opposite 

 or cavernous part of the surface a b, a b. 



58. The thermo-magnetic energies never reached to the 

 cold end of the bar, but returned to the point of heat in di- 

 rections indicated by the arrows, and at no great distance 

 from the heated end. The same laws hold good in all cylin- 

 drical bars of antimony of small dimensions, which are not 

 of an uniform density on every side of the axis. I have 

 broken several into fragments for the purpose of examining 

 their internal structure, and have always observed, that when 

 they have displayed phaenomena similar to those last de- 

 scribed, their density is not uniform ; and the side of the cy- 

 linder, which in a transverse fracture will exhibit the most 

 compact texture, may generally be predicted by an observance 

 of the thermo-magnetic phaenomena which it will display 

 whilst whole. 



59. Cylindrical bars of antimony of an uniform density on 

 every side of the axis, and more than two inches diameter, 

 display thermo-magnetic pha?nomena with very great preci- 

 sion, and a rigid observance of certain laws. 



60. When a bar of this description, and 6 or 8 inches 

 long, has been cast in a vertical mould of sand, let its ends be 

 struck off with a sharp-edged hammer, making the sections 

 transverse and not ragged. Apply the flame of a fierce spirit- 

 liunp for a few moments to the convex surface close to one 

 end of die cylinder, and immediately place the heated side 

 downwards, over and parallel to a delicately suspended com- 

 pass-needle. If the heated end of the bar be placed north, 

 the north end of the needle will be deflected eastward, show- 

 ing that the electric current by which it is deflected is flowing 

 along the lower side of the bar, from the point of heat towards 

 the cold end. 



N.S. Vol. 10. No. 55. July 1831. D 61. Let 



