Mr. Marsh's Thermo-Electric Apparatus. 323- 



wire. The spirit-lamp bemg now apphed at D or E, the ro- 

 tation will commence, either to the right or left, according to 

 circmnstances, and which will be reversed by reversing the 

 pole of the magnet. Fig. 7 shows the same stand widi two 

 magnets. 



This being premised, the reader will easily follow me in the 

 detail of the following experiments, observing that when the 

 motion is said to be to the right or left, he must imagine him- 

 self coinciding in position with the wire about which the ma- ' 

 chine turns. 



Exp. 1. The rectangle, fig. 2, being applied upon the 

 stand, fig. 6, and the lamp at E, the rectangle was projected 

 to the right till D reached the lamp*; it was then propelled 

 back again, and after a few oscillations it remained at rest at 

 right angles to its first position. 



Exp. 2. The rectangle adjusted as before, and the lamp 

 applied at D, the wire was projected to the left with similar 

 results to the preceding. 



Exp. 3. The rectangle adjusted so that D was next the 

 magnet, and the light then applied at D. The wire projected 

 to the right. 



Exp. 4. The rectangle still in the same position, but the 

 light applied at the other extremity. The wire was projected 

 to the left. 



Exp. 5, 6, 7, 8, were made under precisely the same 

 circumstances with the rectangle, fig. 3; and the results were 

 similar, but much weaker, and the motions all reversed. 



Exp. 9, 10, U, 12, were made with the rectangle, fig. 4. 

 The motions the same as with rectangle, fig. 2, except that we 

 generally obtained a rotation when the light was applied as in 

 Experiments 2 and 4. 



Exp. 13, 1 4, 1 5, 16, were still the same experiments, but 

 with die rectangle, fig. 5. The results were as in the above case, 

 but reversed in respect to direction, and inferior in force. 



A similar set of experiments were made with the south pole 

 of the magnet ojiposed to the wire, and Uie results were si- 

 milar, but all in tlie reverse direction. 



As it was obvious from these results that the rectangle, 

 fig. 4, either from its more accurate balance, or from the na- 

 ture of the combination, was the most powerful, it was alone 

 made use of in the following experiments, in which two mag- 

 nets were employed. 



Exp. 1 7. The rectangle, fig. 4, being suspended as shown 



• It ought rather to be said before D reached the lamp ; for wlien l)y 

 chance it did reach it, the wire revolved. 



S S 2 i" 



