16 Mr. W. Sturgeon o?i the Thcrmo-magnetism of 



escence, and dispUiyed pl)a noniena precisely of tlie same cha- 

 i-acter as when the bar was heated at the point b only. 



o'l. The neutral end a was now cut off at the point b ; and 

 the remaining bar displayed thermo-magnetism to whatever 

 end heat was applied. The currents still passed over the same 

 faces, and, as regards the point of heat, uniformly set out over 

 the same face and returned by the other. Hence the apparent 

 contrariety in its direction when the bar was heated at dif- 

 ferent ends. These results led to further experiments with 

 bars, and other forms of solid homogeneous metallic masses; 

 some of which have developed vei'y curious phacnomena, which 

 appear to observe an uniformity in the laws of their exhibi- 

 tion. 



55. Experivients with a cylindrical Bar of Antimony. — This 

 bar was 8 inches long and '75 inch diameter. It was very 

 far from having an uniform surface, being much more ca- 

 vernous on one side than the other, from air-bubbles whilst 

 casting. It was heated at various points of its surface by a 

 very fine-pointed flame of a spirit-lamp ; and its magnetism 

 thus excited was traced by its action on the compass-needle. 

 It was ultimately discovered, that whatever point near to the 

 extremities of the cylinder was selected for the point of heat, 

 the electric current invariably flowed over the same parts of 

 its surface ; and when either of the ends was uniformly heated, 

 whilst the other was kept at the temperature of the atmo- 

 sphere, the bar became highly magnetic ; exhibiting phteno- 

 niena similar to those already spoken of as appertaining to 

 the prism of bismuth (54). The electric current constantly 

 passed over the dense and opposite cavernous sides, whilst 

 the intermediate longitudinal lines were nearly neutral. 



5^. Fiff. 12 will £>ive a oood idea of the direction of the 

 electric currents excited in this bar of antimony when one oi 

 its ends was uniformly heated. The cylinder is supposed to 

 be divided into halves in the plane of its axis, and termi- 

 nating on each side by the dense and cavernous parts of its 

 surface. The two halves are placed edge to edge, with their 

 convex sides upwards. The dotted line in each half will re- 

 present the neutral line, or that longitudinal line on each side 

 of the bar, which when placed parallel over the needle, no 

 deflection was produced. The active lines are c d, and a Z>, 

 a b : the two latter correspond to each other when the halves 

 of the cylinder are replaced or brought together. It is to be 

 observed, however, that the thermo-magnetic energies of the 

 cylinder were not confined to two longitudinal lines ; for every 

 part of the surface near to the heated end was more or less 

 magnetic; but in consequence of the recurved manner in 



which 



