8 Mr. W. Sturgeon on the T/iermo- Magnetism of 



energies of which are directed in one and the same way round 

 the rectangle. 



28. In the side cd there are two neutral points; one of 

 which is exactly at the gate, or that point at which the metal 

 was poured into the mould. I mention this circumstance more 

 particularly, because I have observed that, when the point of 

 heat is situated on different sides of the gate in any of these 

 frames of bismuth, there are generally opposite currents eli- 

 cited ; or, in other words, the gate is generally a neutral point. 

 The letter o in figs. 4. 5. and 6. denotes the gate in each. 



29. Experiments voith the rectangle B. — By consulting fig. 5, 

 it will be observed that the greater part of the side ab \s neu- 

 tral. No deflection of the needle could be produced when the 

 point exceeded one inch from either of the angles a or b. 



30. The side cd has three neutral points, one of which is 

 at the point o, where the metal entered the mould. The two 

 short sides, b c and d a, have each one neutral point. In a d 

 it is at equal distances from the angles a and d. \nbc the 

 neutral point is nearer to c than to b. 



31. The opposite angles a and c are decidedly neutral 

 points ; but when the angles d or c were heated, very power- 

 ful curi'ents were excited. The arrows on each side of the 

 angle b are both directed the same way, and consequently 

 would represent that angle of one uniform character ; but it 

 was found by repeated trials, that there are two neutral points 

 very close to the angle b, but on different sides ; and so close 

 to each other, that it required a very fine-pointed flame to 

 heat one of those points without heating the other also. 



32. Experiments 'with the rectangle C. — There are three 

 neutral points in each of the long sides of this rectangle; one 

 of which is at o, the point where the metal entered the mould. 

 In each of the short sides there is one neutral point, situated 

 nearly at their centres, as is shown by fig. 6. 



33. The angles c and d are perfectly inactive when uni- 

 formly heated ; but the angles a and b, although represented 

 by the arrows as causing conspiring currents on both sides of 

 each angle, have, in fact, each of them two neutral points very 

 close to each other ; so that the needle will be deflected vari- 

 ously by heating different adjoining points about either of 

 those angles. 



34. When those parts of the rectangle which are opposite 

 to the arrows i i are simultaneously heated, the energies of 

 the conspiring currents become very powerful indeed ; and 

 the needle may be driven round on its pivot through a whole 

 circle by following it up with the side of the rectangle. The 

 electric currents excited in this rectangle are more powerful 



than 



