12 Mr. 0. J. Lodge. Experiments on [June 4, 



FIG. 5. 



(as indicated by the dotted line in fig. 5 ; the strong-line shunt being 

 removed), one of the two long thick wires round the room being used 

 to connect up the machine to the same outer coat, as in fig. 4. 

 Under these circumstances, simultaneous sparks could be got at A 

 and at C, and both about the same length, but not when they are too 

 long, say, A = 0*52, C = O57 inch. But now the jar can be made 

 to overflow by either spark if of sufficient length. Thus if A = 0'61 

 or if C = 0*74, the jar lip gets jumped, and sometimes the A spark 

 occurs, sometimes the C, but not both. Another reading: A = 0*69 

 or C = 0'94 ; jar overflows in either case. 



Restore now the usual short wire to the C knobs, and the C spark 

 still often goes, but it has no effect on the jar. The A spark makes 

 the jar overflow as before. 



Bat if the long lead between machine and jar be short-circuited- 

 out (as by the dotted line of fig. 4), while the thick copper spiral 

 still joins up to the C knobs (as indicated by the dotted line in fig. 5), 

 then A cannot make the jar jump, while C can easily. 



Thus overflow, is always easily produced by the action of the spark 

 occurring in a long good-conducting lead, not in a short or bad-con- 

 ducting one. 



Effect of Iron Core. 



10. Using the thick copper spiral as before (6) to make the pint 

 jar overflow, I tried whether inserting large massive iron bars in it 

 as a magnetic core would have any effect. There happened to be 

 three large bars, each about 3 inches in diameter, which were used. 

 They were of soft iron, and intended for the legs of an electromagnet. 



No effect was found. The length of the A spark needed to make 

 the jar overflow was, as near as one could tell, the same, whether the 

 iron was in the spiral or not. Thus : 



Without iron A = O53 



With one bar in spiral O51 



With three bars 0'515 



No difference that one could be sure of. 



