344 DESCRIPTION OF A ROTATORY MULTIPLIER. 



cury, communicating, on one side, with a galvanic pair, on the other 

 with the coil of the multiplier. In the next place, by winding over 

 the first coil, another of similar length, but in a direction the opposite 

 of that in which the first coil was wound, I was enabled, by two other 

 globules, situated so as to communicate severally with the lower ends 

 of the pins, at the opposite side from that on which the first mentioned 

 globules were, to cause an impulse at every semi-revolution. 



The one coil being wound to the right, the other to the left, the al- 

 ternate effect of each upon the needle was similar in opposite parts of 

 the orbits described by the pins. Lastly, a second needle, furnished 

 with pins in like manner, being fastened at right angles to the first, so 

 as to form with it a cross, as represented in the engraving, each nee- 

 dle is made to receive two impulses during every revolution. Hence 

 one of Danell's sustaining batteries, as made by Newman, is quite ade- 

 quate to cause a revolution as rapid as consistent with a due degree of 

 stability in the mercurial globules employed. 



One end of each coil, by means of the branching wire A, communi- 

 cates with one pole of the galvanic pair ; the other ends of the coils 

 terminate in mercurial globules contained in cavities on opposite sides 

 of the wooden disc G, upon the centre of which the spindle of the mag- 

 netic needle rests. The branches of the wire K proceeding from the 

 other galvanic pole, terminate in globules situated in the vicinity of 

 those above mentioned, so that as the needles revolve, the pins pro- 

 ceeding therefrom perpendicularly may touch a pair of the globules 

 first on one side and then on the other. Whenever this contact 

 takes place, the circuit is completed, and a discharge is effected through 

 one or the other of the coils of the multiplier. 



Supposing E and F to be north poles, a discharge through one of the 

 coils will cause E to move off a quarter of a circle, or more. As this 

 ensues, the pins of F will come in contact with the globules which 

 those of E touched before. Of course F will be propelled so as to 

 cause the pins of E to reach the pair of globules at G, which, com- 

 pleting the circuit of a coil wound in a way the opposite of that first 

 mentioned, concurs with that coil in its influence, so as to promote 

 the rotation previously induced. The same result ensues when the 



