450 Dr, Traill on some Thennomagnetic Experiments, [Dec. 



on the subject that I have perused led me to expect. Should 

 they appear to you sufficiently important, I transmit a short 

 account of them for insertion in the Annals. 



The apparatus which I have found most convenient consists 

 of a bar of antimony 4-1 inches long, half an inch broad, and 

 one-quarter of an inch thick. To this, a slip of copper equally 

 broad, and bent as in the figure, is firmly attached by a few 

 turns of copper wire. This method of connecting them is better 

 than by solder; because the joinings can then bear a higher 

 temperature. A spirit lamp is the source of heat, and the devia- 

 tions are observed with a delicate pocket compass, the needle 

 of which is 1-i- inch long, and poised on an agate centre. 



When the apparatus is placed in the magnetic meridian, with 

 the slip of copper uppermost, as in the figure, and the lamp is 

 applied to the N end of the bar, the needle placed within the 

 rectangle, always deviates to the W ; while the compass placed 

 in contact with any part of the outside of the rectangle (whether 

 copper or antimony) deviates to the E. 



These effects are reversed when the lamp is apphed to the 

 south end of the apparatus, other circumstances remaining the 

 same. While the metalHc surfaces in contact are bright, the 

 deviation often amouuts to 75° within the rectangle, while 

 without it, the deviation is usually 45°, or upwards ; and the 

 effect produced by the upright portions of the copper connecting 

 piece, is less than of the horizontal parts of the same metallic 

 slip. 



Absolute contact of the compass with the metallic apparatus 

 is not necessary. The effects were apparently as powerful when 

 the compass was placed on a plate of glass about half an inch 

 in thickness ; or even when held in the hand, without touching 

 any part of the apparatus, especially when held within the paral- 

 lelogram. Hence the magnetic power of such a combination 

 extends to some distance from its surface, like the magnetism 

 of a common magnet. 



Inclining the apparatus at different angles from 20° to 72°, 

 produced no change in the deviations, provided the direction of 

 the apparatus was in the plane of the magnetic meridian. 

 - The inversion of the apparatus so as to place the antimony 

 above the copper was then examined. When the N end of 

 the bar was heated, and the compass on the upper, now outer 

 surface of the antimony, and on all the outer surfaces of the 

 rectangle, the deviation was still to the W, and all the interior 

 surfaces of the rectangle showed deviation to the E. When the 



