Dr. L. T. More on Dielectric Strain. 531 



In the somewhat analogous case where iron and nickel are 

 strained by magnetization, changes in volume are found to 

 be much less than changes in linear dimensions, as shown by 

 Barrett, Bidwell, and Cantone. The value of this evidence 

 may also be criticised from the fact that the proof of the law 

 is claimed by both these investigators. Yet the numerical 

 results of Prof. Quincke are far larger than those of Prof. 

 Cantone, so much so that Dr. Sacerdote considers them of 

 no quantitative value. Doubt is thrown not only on the law, 

 but on the effect itself where proof is obtained on such other- 

 wise discordant results. The same may be said of their law 

 of proportionality to V 2 /d 2 . 



I have written these notes with no intention of attempting 

 to prove that because I found no elongation when certain 

 o-lass tubes were electrically charged, the phenomenon does 

 not exist, but because I wished to show that my apparatus 

 was capable of measuring elongations of the size found by all 

 other investigators in this field. Also because the evidence 

 of the existence of the effect is very conflicting and, by reason 

 of probable errors, doubtful. 



These errors, which may have influenced the results of 

 Prof. Cantone, are : — 



1. No adequate precautions seem to have been taken 

 against that most troublesome source of error, the bending of 

 the tube by electrostatic attraction. On the contrary, a long- 

 fine tube of glass, with thin walls, suspended from one end is 

 peculiarly liable to this error. 



2. The elongations are too close to the limits of the apparatus 

 to establish a law when the existence of the effect is in doubt. 



3. The change of length on charging is quite different 

 from that on discharging. This difference, or change of the 

 zero, varies from nothing to forty per cent, of the total elon- 

 gation. The writers assume this difference to be due to heat. 

 The amount of the effect seems to be erratic and not propor- 

 tional to the total deflexion, though in a general way it is 

 said to be dependent on the time of charge. An extraneous 

 effect of such a magnitude must cause doubt as to the amount 

 of the effect due to electrostriction, and much more doubt as 

 to the establishment of any law in regard to it. 



4. As the delicacy of the measuring apparatus increases 

 and greater precautions are taken against errors and extra- 

 neous effects, the amount of the elongation constantly de- 

 creases. This, I think, makes one question its existence. 



To conclude, these reasons make me still believe that part, 

 if not all, of the effect observed may be due to other causes 

 than the one sought. 



