Mr. Marsh's Ther mo-Electric Apparatus. 325 



It must already have struck the reader as remarkable, that out 

 of the four positions in which the lamp may be applied in the 

 double rectangle (fig. 9), two of them give a rotation in one 

 direction, and one of them in an opposite one, and that at the 

 fourth point there should be no tendency to rotation what- 

 ever, but on the contrarj' a decided direction ; this latter point 

 is to the right of the north pole, or to the left of the south pole 

 of the magnet, the observer assuming his position to coincide 

 with the axis of motion, as already suggested in the preceding- 

 part of this article. A satisfactory illustration of this singular 

 phaenomenon will, it is presumed, be considered a strong test 

 in favour of the hypothesis on which it is founded, particularly 

 if the same hypothesis should be found competent to the il- 

 lustration of every other electro-magnetic phaenomenon hi- 

 therto observed. 



In my " Essay on Magnetic Attractions and on Electro- 

 magnetism," I have shown that, by supposing the galvanic 

 conducting wire to act upon the magnet with a tangential force 

 varying inversely as the square of the distance, we may not 

 only illustrate but compute the effect of any proposed combina- 

 tion ; let us then see how far this same supposition will assist 

 us in explaining the singular anomaly above mentioned. 



The platinum being positive to the silver; when the lamp 

 is applied at the union of the two metals, the fluid will be 

 transmitted through the silver wire under the same circum- 

 stances as in a galvanic appai'atus, with two plates, it is trans- 

 mitted through the conducting wire from the copper side to 

 the zinc ; it ought therefore, when thus transmitted, to project 

 the north pole of the magnet to the left, the observer now 

 coinciding in position with the wire through which the circuit 

 passes. Or the magnet being fixed, and the wire free as in 

 our case, this latter ought to be projected to the right hand. 

 Thus referring to fig. 9, and conceiving the lamp to be ap- 

 plied at E, the point E ought to be projected to the right hand 

 of the observer, looking towards the magnet, while the points 

 F, D, and G (the circuit in these being descending) ought to 

 be projected to the left, the observer conceiving himself coin- 

 ciding in position with these respective wires and always 

 looking towards the magnet. On the contrary, when the light is 

 applied at F, then the circuit being ascending at F, and de- 

 scending in all the other branches, this ought to be projected 

 to the right, and all tlie others to the left; and so on, for any 

 wire to which the lamp is immediately applied : and of course 

 the contrary to all this ought to happen when the south pole 

 is applied. Let figures 11, 12, 13, 14<, represent the four ap- 

 plications of the lamp as stated in experiment 19; and the 



several 



