200 Mr J. D. FORBES on eliciting an Electric Spark 



most effectual for eliciting the spark. These were, 1st, That the 

 spark is more easily obtained at the instant of interrupting than 

 that of completing the galvanic circuit : 2d, That of the combi- 

 nations which I tried, a fine pointed iron- wire suddenly with- 

 drawn from contact with a surface of pure mercury, forming part 

 of the circuit, was the most regular in exciting the spark, and 

 that a good deal depended upon the suddenness of the interrup- 

 tion ; and, 3d, That the spark was easiest obtained from the 

 mercury, not at the horizontal upper surface, but where capillary 

 action attracted it to the sides of the containing vessel ; and that 

 this was independent of the material of the vessel, being the same 

 with wood, glass, and metal. 



These precautions are all easily resolved into the general 

 one that the circuit conveying a weak current should be as ab- 

 ruptly divided as possible, in order to the production of a spark ; 

 and from the greater power of slender wires formed of imper- 

 fect conductors, such as iron, to accumulate the effects of weak 

 currents *, that part of the apparatus is also explained. The ac- 

 tion of the adhesion of mercury to the vessel is probably to ren- 

 der its connection with the wire inserted at that point more 

 perfect. 



It will be unnecessary at present to recount the difficulties 

 which still obstructed my progress, though some of them were, I 

 think, of a kind not unimportant to electro-magnetic science. I 

 shall, however, only briefly notice the arrangement of the appa- 

 ratus with which, on the 1 3th of April, I succeeded in obtaining 

 the spark at pleasure. 



The large natural magnet is represented at A, Fig. 6, 

 Plate IX. A cylindrical connecter of soft iron, ab, passing 

 through the axis of the helix c, was made to connect the poles 



* See Mr HARRIS'S Paper in this volume. 



