18 Prof. L. T. More on Elect rostriction. 



brass tube 55 cms. long. 4*7 cms. inner diameter, and 2 mm. 

 thick. To insulate it, two glass caps, slightly larger in diameter, 

 were added to the ends, as shown in fig. 1. When the glass 

 wedges were in place, I succeeded in overcoming the lateral 

 motion entirely. 



With a spark-gap o£ 12 mm., the image remained quite 

 at rest until the spark passed, when there was a sudden 

 motion o£ 25 divisions. Keeping the potential constant, 

 a number of charges were given, the jerk as the spark 

 passed grew successively smaller, and finally ceased alto- 

 gether. The spark-gap w-as increased to 2 cms., but produced 

 no deflexion. Higher potentials could not be used as the 

 upper cap broke. 



When the silvered glass armature was used, the electricity 

 leaked rapidly from the sharp edges of the silvering, causing 

 a violent motion of the oil and a glow-discharge between the 

 surface of the fluid and the wire cage covering the mirror. 

 The edges of the brass-tube armature were well-rounded and 

 the discharge materially lessened ; so much so that the 

 generator reached its highest potential in far less time, and 

 the tube retained a high charge for many minutes. The 

 heating was correspondingly less and slower. 



Heating Due to 10 SparJcs of 12 Millimetres. 



Deflexion 2 minutes after discharge , 22 div. 



„ 2 ,, later... 20 „ 



?9 ^ ?? ?? ^"i ?J 



10 „ :. 76 „ 



The following experiments were made when the room 

 temperature was constant, and sufficient time was allowed 

 between two . chargings for the tube to cool off and the image 

 to come to rest again. 



With the spark-gap set for 12 mm. a spark passed through 

 the oil from the upper edge of the armature to the wire 

 cage. Just before and at the time of the discharge, the 

 tube expanded 27 divisions. The generator was short-circuited, 

 and the heat effect for one minute was 11 divisions. 



The discharge was a second time through the oil, the 

 accompanpng expansion, 13 divisions, and the motion during 

 one minute after discharge, 11 divisions. 



Spark-gap 10 mm., the discharge was across the spark- 

 micrometer. There was no motion whatever of the image. 

 Heating effect for one minute after discharge, 9 divisions. 



Spark-gap 11 mm. Xo effect except a slight and rapid 



