Mr. 0. J. Lodge. Experiments on 



[June 4, 



The copper is commercial quality and evidently of miserable con- 

 ductivity. I afterwards got some real copper from Messrs. Thos. 

 Bolton and Sons, and with it the phenomena are still better marked. 



EAKLV EXPERIMENTS. 



3. The large glass condenser (0'028 mfd.) was charged through 

 one or other of the long wires, and a choice was offered the discharge, 

 so that it might go either round the wire or leap an air-gap, as it 

 chose ; as shown in fig. 1. 



FIG. 1 . 



A are the ordinary terminal knobs of the Yoss or Wimshurst 

 machine where the spark occurs ; B is the discharge interval acting 

 as a shunt to the wire or other resistance. MQN represents diagram- 

 metrically one of the wires round the room. The spark-length B was 

 adjusted so that it was an off chance whether the discharge chose it 

 or the wire. It was noticed that when the discharge chose B the A 

 spark was strong, but when the discharge choselihe wire the A spark 

 was weak. The difference appeared to be only in the noise or sud- 

 denness of the spark, for when a Kiess's electro-thermometer was 

 inserted in the circuit it indicated about the same in either case. 



A capillary tube was filled with very dilute acid so that its resis- 

 tance was about ^ megohm, and was connected across the B knobs 

 instead of the long wire. When this acid tube was thus made the 

 alternative path, and the B knobs placed so far apart that the dis- 

 charge was obliged to choose it, the A spark was very weak, being 

 reduced to a quiet spit, which could be analysed by a slowly rotating 

 mirror into several detached sparks. 



