426 



Profs. John Perry and W. E. Ayrton. [May 27, 



There cannot be much doubt of the fact, that electric induction 

 through dielectrics, and magnetisation of iron, are really phenomena of 

 stress and strain, and it is probable that the correctness of Ohm's law 

 when applied to dielectrics is dependent on the theory of continued 

 strain, which we have just given, and which we are attempting to 

 prove. In addition to experiments such as we have described, we 

 have been making others to find whether beams which are kept in a 

 state of strain rise in temperature through internal friction, which is 

 of course analogous with electric resistance, and in addition to see 

 whether a soft steel magnet acted upon by a constant current also 

 rises in temperature ; but hitherto these heating effects, if they exist, 

 have been too small for our instruments to detect. In speaking of 

 beams we refer especially to glass "beams ; a beam of sealing-wax would 

 probably under the same circumstances perceptibly rise in tempera- 

 ture. 



As a further example of the reductions which we are describing in 

 this paper, we give a table of observed and calculated values of the 

 diminishing deflection of a glass fibre, which had been twisted for some 

 days, and then released. The fibre was 85 centims. long, of a mean 

 diameter of 0*034 centim. suspended vertically, so that its upper end 

 could not turn in the support, and carrying a weight of 211*6 grms. 

 Attached to this weight was a vane dipping into water to diminish the 

 rapidity of the swings. 



The untwisting of the vertical glass wire was observed by the re- 

 flection of a ray of light, coming from a fixed lamp, by a mirror rigidly 

 attached to the wire at its lower extremity. The spot of light, or 

 rather the image of the cross-wires in the lamp slit, moved over a 

 horizontal semicircular scale graduated in millimetres, and one metre 

 in radius, with the mirror turning at its centre. An arrangement 

 similar to that employed in our deflected beam experiments was used 

 to obviate errors arising from difference of position of the flame of the 

 lamp on different days. 



The mirror and the weight having been kept twisted through a 

 right angle for about a week, the deflecting couple was suddenly re- 

 moved, and time readings of the diminishing deflections taken. As at 

 first the deflection diminished with considerable rapidity, and as in 

 addition the spot of light described a series of oscillations about the 

 varying mean position, the system of registration we employed was as 

 follows : — The position of the spot of light on the scale at the end of 

 every swing was observed as accurately as possible, and the time re- 

 corded by simultaneously pressing down the key of a " break circuit 

 chronograph." From these successive time observations of the swings 



