Homogeneous Bodies ; luith illustrative Experiments. 5 



16. By consulting the direction of the arrows in fig. 2. the 

 reader will easily ascertain the direction of the electric cur- 

 rents when any particular part of the apparatus is heated; for 

 by heating any point within the range of any individual arrow, 

 that arrow will point out the direction of the electric current. 

 Or, if the whole length of that arrow be heated at the same 

 time, the current still flows in the same direction, and is con- 

 tinued in every part of the frame. The same explanation 

 will also apply to all the other figures, unless otherwise ex- 

 pressed. 



Second Class of Experiments, 



1 7. These experiments were also made with a rectangular 

 frame of bismuth exactly of the same length as the former, 

 and about 1-75 inch broad. More care was taken in the 

 casting, and the frame was a better figure. 



1 8. The direction of the arrows in fig. 3. will point out the 

 course of the electric current when any point opposite to them 

 is heated. It will be observed by consulting that figure, that 

 every angle is a neutral point ; and that the long sides have 

 each one neutral point, and when this is heated, no current is 

 excited. 



19. There is a circumstance connected with these experi- 

 ments which is well worthy of remark. On the side ab of 

 the rectangle, and near to b, there was a pi'otuberance on the 

 inner face : when the point of heat was between this pro- 

 tuberance and the angle b, no current was excited ; but on 

 the other side of the protuberance being the point of heat, a 

 powerful electric current was put into motion ; so that, strictly 

 speaking, the arrow ought not to reach to the angle b. The 

 protuberance was afterwards filed down, and that part levelled 

 with the rest of the side : no difference was produced in the 

 thermo-magnetic character of that part of the frame by the 

 change thus made. Hence it would appear, that, the internal 

 structure of the metal alone, operates in giving direction to 

 electric currents excited by heat. This opinion will appear 

 much better supported by experiments and observations, 

 which will be spoken of in the sequel. 



20. I nmst notice in this place, that the thermo-magnetic 

 energies in this rectangle vary considerably by heating dif- 

 ferent points. When the point of heat is in the side a b, the 

 current which sets in from b io a is much more powerful than 

 the opposite current from a io b; so that by heating any part 

 in the half nearest to a, a stronger current is excited than by 

 heating a point similarly situated with regard to b. 



21. When cither of those halves is uniformly heated, the 



needle 



