ijohen und£r the Electro-Magnetic Influence. 423 



magnet, made in the shape of a horseshoe, was formed of a 

 cyhnder of soft iron of the length of 0"^*335 and the diameter 

 of O'^'OIS; and its electro-magnetic spiral was formed of a 

 copper wire 33"^ long, and of a diameter of a millimetre and 

 a third ; the internal distance of the poles was 0'"'027; the 

 two solid parallelepipedon contacts, forming the interrupted 

 anchor, were 0'»-04 long; and of the sides 0«i'01 1 and 0°^-006 : 

 and the hollow terminal pieces were 0™*035 long ; and of the 

 side 0™*009. They were placed at a distance from one another 

 of four to five millimetres, the magnet being kept in a vertical 

 position with the poles turned upwards. In front of the in- 

 terval of the separation of the contact pieces was placed the 

 flame of a small candle, or of a little oil or alcohol lamp, so that 

 it surmounted with its top by nearly a fourth the thickness of 

 the contacts. The electric circuit was closed by copper wires, 

 and the metallic unions were maintained both at the magnetic 

 poles and at those of the pile by clamps : one of the wires 

 therefore was divided into two equal parts, and the ends 

 being dipped into a tumbler of mercury, allowed the closing 

 and opening of the circuit at pleasure. 



/ have constantly observed repulsion ifi the act of closing the 

 circle, which lasted the xohole time that the magnetism was kept 

 up ; and, when in the act of opening the circle, I saw thejlame . 

 return to its primitive position. Well-satisfied with having 

 in this manner confirmed this important fact which reflects 

 honour on its discoverer, I applied myself to the study of the 

 phaenomenon, and I found — 



I. That this happens with contacts of both solid and hollow 

 soft iron; whereupon I abandoned my suspicion that the 

 movement of the flame was attributable to currents of air; I 

 convinced myself that it was an immediate action of the mag- 

 netism upon the flame, — a fact of the greatest importance to 

 science. 



II. That the repulsion, when it is quite distinct and the 

 flame quite pure, and terminated in a well-shaped top, is ac- 

 companied by depression : repulsion and depression are simul- 

 taneously observed at the closing of the circle ; the return of 

 the flame and rising of the same, at the opening of the circle. 



III. That, ceteris paribus, the greatest effect takes place when 

 the flame is touching the convex of the magfietic curves indicated 

 b\j iron filings. 



IV. That the action is tiull, or almost nullf when the flame 

 is placed in the centre of the interval which separates the two 

 contacts. 



V. That in the manifestation of the effects stated above, it is 

 not necessary for the contacts to be entirdy separated: they may 



