530 Dr. L. T. More on Dielectric St 



rain. 



instead of 3*5, a displacement certainly within the limits of 

 observation. With somewhat higher potentials the tubes 

 always ruptured but gave no signs of an elongation. These 

 facts make it extremely difficult for me to acquiesce in Dr. 

 Sacerdote' s explanation why I should not have expected to 

 observe a deflexion. 



I agree with Dr. Sacerdote that an optical arrangement is 

 much preferable to any mechanical process of amplification. 

 My own apparatus was certainly an optical one, although he 

 implies that it was not. As far as accuracy and stability are 

 concerned, there is probably little difference between reading 

 by a microscope the deflexions of a beam of light reflected 

 from a mirror and reading the deflexion of a beam of light 

 by its interference-bands. Whether the mirror rests on three 

 legs on the tube or is cemented to it, both must record any 

 irregular and vibratory motions which the tube may have. 

 The use of a tilting mirror has been employed so often and 

 with such success that I hardly thought its use would be 

 questioned. And I believe I have shown that there was 

 about the same ratio between the phenomenon to be observed 

 and the magnifying powers in the two cases. But my method 

 has this distinct advantage: lateral motions of the glass tube 

 are magnified to the same degree and as clearly as an elon- 

 gation, and so this most serious source of error can be at once 

 detected and eliminated. I have not found that Prof. Can- 

 tone noticed this source of error, or took precautions to avoid 

 it further than to choose glass tubes as uniform as possible. 

 But very small differences in thickness and lack of centre in 

 position cause such large deflexions as to invalidate all 

 results, as my observations and those of Prof. Quincke clearly 

 show. 



As evidence of the existence of dielectric expansion and 

 of its proportionality to V 2 /d 2 , both Prof. Quincke and Prof. 

 Cantone place much weight on the fact that the linear elon- 

 gation, as found by them, is approximately one-third the 

 volume expansion of the liquid in the thermometer-condenser. 

 The value of this evidence seems doubtful. According to 

 the recent theory of Dr. Sacerdote this relation does exist. 

 But it should be remembered that this theory is based on 

 coefficients wdiose existence depends on the experiments 

 under discussion. On the other hand, Maxwell's theory of 

 stresses in the aether certainly does not point to the above 

 relation ; a tension along the lines of force and a pressure at 

 right angles to them will produce a differential effect on the 

 volume, and the volume change will be less than the elonga- 

 tion at right angles to the lines of force, instead of greater. 



