14 Mr. W. Sturgeon on the Thcrmo-magnetisin of 



ridian, the greatest deflection was again produced in the op- 

 posite direction to the former. 



46. Considering this as one step at least towards the dis- 

 covery of the cause of this novel phaenomenon, I proceeded 

 to examine the opposite side of the frame, which 1 supposed 

 might possibly present similar effects. For this purpose I 

 heated one of the points in that side which by previous trials 

 was found to be very active, and then placed it directly over 

 the magnetic needle, keeping the plane of the frame in the 

 magnetic meridian, and the same end north as had been in 

 that position whilst experimenting with the other side. The 

 needle was deflected in precisely the same direction as I had 

 always found it to be by a similar position of the frame, and 

 with the same point of heat. 1 afterwards inclined the plane 

 of the frame sometimes eastward, at others to the west ; but 

 in no instance could I obtain the least indication of results 

 similar to those which I had noticed whilst the opposite side 

 of the frame was nearest to the needle. Thinking that the 

 different active points might possibly operate under different 

 laws, I next heated the point which had before presented 

 the extraordinary phtcnomenon, still keeping the other side 

 nearest to the needle ; but nothing remarkable was noticed by 

 this variation of the experiment. The needle continued to be 

 deflected in one uniform direction, whatever inclination was 

 given to the plane of the rectangle. I tried the other sides of 

 the rectangle in precisely the same way, but obtained no un- 

 usual results. It now appeared evident that one side of the 

 frame was endued with peculiar properties which the other 

 sides could not be made to exhibit, which soon proved to be 

 the fact. 



47. When this side was heated at one particular point, two 

 distinct electric currents were called forth ; one of which may 

 be distinguished by the name of general current, because it 

 pervaded every part of the metallic frame ; the other cur- 

 rent was perfectly local., and could be traced only to a short 

 distance from the point of heat. It never reached further 

 than the angle, and returned into itself on the opposite face 

 of the solid prism which formed that side of the frame. 



48. From this singular result, it appeared likely that some 

 inaccuracy might possibly have occurred in the conclusions 

 I had already drawn from the former experiments. Fortu- 

 nately the rectangles were not broken, and I had an oppor- 

 tunity of examining them again: but it seems that they were 

 of too small dimensions to ]noduce local currents, as I found 

 them to operate exactly as at first. On trying one of the 

 circular rings, however, another curious fact was discovered, 



which 



