of Magneto-Electricity. 265 



inclose a stratum of air between it and the tinfoil. Lastly, 

 the flannel was well varnished. By these precautions the in- 

 jurious effects of radiation, and especially of convection of heat 

 in consequence of the impact of air at great velocities of rota- 

 tion, were obviated to a great extent. 



The small compound electro-magnet was now put into the 

 tube, and the terminals of its wire, tipped with platinum, were 

 arranged so as to dip into the mercury of a commutator*, 

 consisting of two semicircular grooves cut out of the base of 

 the frame, fig. 1. By means of wires connected with the mer- 

 cury of the commutator, I could connect the revolving electro- 

 magnet with a galvanometer or any other apparatus. 



In the first experiments I employed two electro-magnets 

 (formerly belonging to an electro-magnetic engine) for the 

 purpose of inducing the magneto-electricity. They were 

 situated with two of their poles on opposite sides of the revol- 

 ving electro-magnet, and the other two joining each other 

 beneath the frame. I have drawn fig. 2 representing these 



Fig. 2. 



electro magnets by themselves, to prevent confusing fig. 1. 

 The iron of which they were made was one yard six inches 

 long, three inches broad, and half an inch thick. The wire 

 which was wound upon them was ^ th of an inch thick ; it was 

 arranged so as to form a sixfold conductor a hundred yards 

 long. 



The following is the method in which my experiments were 

 made : — Having removed the revolving piece from its place 

 (which is done with great facility by lifting the top of the frame, 

 and with it the brass socket in which the upper steel pivot of 

 the revolving piece works), I filled the tube containing the 

 small compound electro-magnet with 9| oz. of water. After 



* I had matte previous experiments in order to ascertain the hest form 

 of commutator, but found none to answer my purpose as well as the above. 

 I found an advantage in covering the mercury with a little water. The 

 steadiness of the needle of the galvanometer during the experiments proved 

 the efficacy of this arrangement. 



