10 Mr. W. Sturgeon on the Tkermo-tnagnetism of 



can be exposed to tlie needle by liaving the bismuth in those 

 varieties of shapes ; for when the metal is in the shape of a 

 circular ring, it is impossible to bring any more than a very 

 small portion of its surface at any one time sufficiently near 

 to the needle to examine the character of the magnetism 

 which it displays. With rectangles, or other straight-sided 

 frames, it is quite otherwise ; for with those, the metal may not 

 only be brought parallel and close to the needle for a consi- 

 derable extent, but, if the frame be sufficiently long, it can be 

 made to operate on both poles of the needle at the same time, 

 — an advantage of considerable importance in experiments of 

 this delicate character*. 



38. Fig. 7. will illustrate the thermo-magnetic character of 

 one of these rings, and will be sufficient, also, to give a toler- 

 ably good idea of the thermo-magnetism displayed by similar 

 rings of bismuth generally : although some trifling difference 

 as to the energy, number, and situation of the active points will 

 frequently be found amongst them ; for, in the present state 

 of the inquiry, it is next to impossible to procure two exactly 

 alike; nor is it easy to predict an active or an inactive point 

 in a ring of bismuth which is 33'mmetrical in all its parts. 



39. It will be observed, by contemplating fig. 7, that the 

 gate o is a neutral point. This I have also shown (28) is the 

 case in I'ectangles ; and, indeed, I believe that the gate will al- 

 ways be found in that state, into whatever form of open frames 

 bismuth may be cast. 



40. For the convenience of bringing a greater part of the 



* Should it be asked why the multiplying galvanometer was not resorted 

 to in these delicate experiments, the answer would be, that that instrument, 

 however valuable it may be for some purposes, is quite inapplicable in 

 these inquiries; where every metal, excepting that under examination, 

 should studiously be avoided, and on no account be permitted to enter 

 the thermo-magnetic circuit. Errors frequently occur by employing the 

 multiplier whilst examining the thermo-magnetic character of very small 

 specimens of metals. Besides, the circuit in that instrument is frequently 

 much too long to be penetrated by the feeble energies which are sometimes 

 displayed in homogeneous bodies, but which are easily detected in short 

 circuits, through which they will pass with very great freedom. Doubts, also, 

 regarding the correctness of the results, would necessarily have presented 

 themselves, had any other metal been permitted to enter the circuit ; and 

 with very great propriety indeed, might every experiment have been ques- 

 tioned, had the galvanometer, with its copper multiplying wire, mercurial 

 cups, &c. been employed in an inquiry which professes for its object the 

 contemplation of the thermo-magnetism of homogeneous bodies alone. 

 Moreover, these researches, as will presently be shown, have led to dis- 

 coveries which could never have been made by the employment of the 

 multiplying galvanometer; and the character of several of the experiments 

 is such, as entirely to preclude the use of that instrument in their exhibi- 

 tion. 



edges 



