of Light hy the Discharge of a Leyden Jar. 343 



a battery of ordinary Leyden jars and by single jars, which 

 altogether raise the capacity to "074 microfarad or 670 metres. 

 To vary the capacity in the ratio 1, 2, 4, I use a switch con- 

 sisting of 6 glass pillars, each with a bit of close-fitting glass 

 tube cemented on the top to serve as a mercury cup. The 

 leading wires are held in the mercury cups by indiarubber 

 umbrella-rings round the glass pillars ; and insulated wire 

 bridges easily make connexion between one pillar and the 

 next. 



In the annexed figure dotted lines indicate sufficiently the 

 permanent connexions. M being , , 



machine and discharger, L being j \ ,.. "^ ,. \ \ 



circuit and coils. The movable ""]"""'•...' '-J r 

 connexions are made in pairs as / \ 



follows : — / \ 



For series or cascade, connect / \ 



2, 3; 4, 5. j /-"T--^ j 



For one condenser only, con- \/^ q Ny*' 

 nect 3, 4. X i /' 3 A 



For both condensers in parallel, / JP / O \ 



connect 1, 2 ; 3, 4 ; 5, 6. M^ f% j 



As regards circuit, one of the I '--K K) 4 / 



coils I use is a hank of 440 yards \ ^9 \ O I 



of thickly covered No. 16 G. P. \;' b^ \/ 



wire as it came from the maker, /\ '^\ 



with a self-induction of '048 ( ,.,^ ^,_^^ / 



secohm and a resistance of 3*75 ( l ') i^ l )' 



ohms. ''"' 



The coils on the CS2 tube consist of the same kind of wire ; 

 they have a resistance rather greater than the above, and a 

 combined self-induction of about '008 secohm. 



A number of gigantic electric-light cables have been in- 

 serted in the circuit at different times, one on a bobbin like 

 two cart-wheels being kindly lent by the Engineers of the 

 Liverpool Electric Supply Company ; but the easiest way of 

 getting very large self-induction is to use No. 20 or No. 22 

 G. P. or I. R. wire not too thickly covered. I have recently ob- 

 tained a large coil wound so as to give maximum self-induc- 

 tion which I estimate as 7 or 8 secohms ; I have not yet 

 used this, but I frequently use a number of coils of No. 21 

 wire packed together as close as is easily possible, and making 

 a total self-induction of 1 secohm or ten thousand kilometres. 



Without going into further details, I may say that the ob- 

 served frequency of the oscillations, as estimated from the 

 appearance of the serrations in the revolving mirror, or by 

 the pitch of the musical note accompanying the spark, agrees 



