Dr. L. T. More on Dielectric Strain. 529' 



the limits were obtained. It was easily done by reducing 

 mechanical pressure on the tube until the deflexion could 

 just be observed. Later it will be seen Prof. Cantone noticed 

 that when his tubes were charged the deflexion was greater 

 than when discharged, leaving a permanent displacement of 

 the bands. For example, in one case a charge caused a dis- 

 placement of 0*95 fringe and the discharge 0*58 fringe. 

 Dr. Sacerdote does not assume in his criticism that this dis- 

 placement of the zero of 0'37 fringe is due to inaccuracy in 

 reading ; but I should suppose imagined that the two changes 

 of length were different. He must then either accept the 

 limits of reading which I state or else question the reliability 

 of all my experiments, and he does just the reverse. 



It is proper here to introduce the limits of the apparatus 

 used as a standard. In the paper of Professors Cantone and 

 Sozzani, cited above, are given the number of interference- 

 bands displaced by the glass tubes when charged. These 

 vary from 0*18 to 2*61 bands, and by far the greater number 

 are less than one band. I do not believe the writers would 

 wish to claim accuracy below a tenth of a band, and it seems 

 as if these standard results were theoretically as near the 

 limits of observation as mine were. 



The limit at which a change of length could be seen was, 

 for my apparatus, a quarter of one division, and when the 

 tube was charged the zero showed no deflexion either on charge 

 or discharge ; theoretically it should have moved 3*5 divisions. 



But Dr. Sacerdote has not calculated these results from the 

 highest potentials used. The value taken was for a spark- 

 length of 12 mm. between balls 2 cm. in diameter. Now on 

 p. 201 of my paper may be found this statement: "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 this 

 potential / still got no defle,vion. ,> It should be understood 

 by this that the distance between the electrodes was, as before, 

 2*5 mm., the electrodes being still the tinfoil pasted on the 

 glass. The kerosene formed no part of the dielectric, but 

 merely acted as an insulator to keep the electricity from, 

 leaking from the tinfoil down the glass. The potential equi- 

 valent to a 2-centimetre spark is about 140 C.G.S. The 

 square of this value used instead of the square of 110 should 

 have caused a deflexion of the micrometer of 5' 7 divisions 



