204 Dr. L. T. More on the Supposed Elongation of 



but then I observed no up or down motion with potentials as 

 great as twelve millimetres, or about 33,000 volts. With a 

 perfectly straight tube, 55 centimetres long, a lateral motion 

 at the top of. 0'013 centimetre causes a vertical displacement 

 of one division *. 



Supposing that the potential was not great enough, it was 

 resolved to confirm the work of former investigators. The 

 outer cylinder was removed, the space between the glass and 

 inner brass cylinders filled with acidulated water, and the 

 middle of the outer surface covered with tinfoil 45 centi- 

 metres long. In no case could I observe any elongation or 

 contraction. A comparison with the results of Righi and 

 Quincke shows that an elongation of at least ten divisions 

 should have occurred. 



With glass tubes Righi obtained an elongation 1/500 milli- 

 metre under the following conditions : — 



Length of tube 1 m. (length of part charged not stated). 



Wall thickness 1mm. 



Spark-length ... 1 cm. (balls 15 mm. in diameter). 



Change of length 0002 nun. 



Length of tinfoil on niy tube ... 45"0 cm. 



Wall thickness 2 - 5 mm. 



Spark-length 12 mm. (balls 2 cm. in diameter). 



Change of length 0. 



According to Righi, Si <x -^ j-^- ; according to Quincke, 



l(v t — v Y 

 81 oc 72 ■ From the former my tube should have 



elongated 0*002 x 0"45x l/2 - 5 (considering the potential as 

 one centimetre spark) =3*6 x 10~ 4 millim. ; and from the 

 latter, 0-002 x 0*45 x l/2'5 2 = 1*4 x 10" 4 millim. One division 

 of my micrometer equals 1*5 x 10 -5 millim., so that the former 

 elongation is equivalent to 24 divisions, and the latter to 9' 6 

 divisions. 



With this arrangement higher potentials were not used as 

 the electricity leaked from the tinfoil to the base of the 

 apparatus. For this reason, still leaving the tinfoil on the 

 glass, the outer brass cylinder was put in place and a metallic 

 connexion made between the two so that they should be at 

 the same potential. On filling the intermediate space with 

 kerosene a spark-length of two centimetres was obtained. At 



* By bending the glass tube slightly by mechanical means, it was easy 

 to confirm the statement, that if the tube bent the image of the fibres 

 moved laterally and at the same time vertically. When the force was 

 removed the image returned accurately to its original position whether 

 one or two legs of the tilting mirror were on the glass tube. 



