1880.] Strained Material, Leyden Jars, and Voltameters. 419 



and r' should be both very small, so that the exponentials in the equa- 

 tions both of the charge and discharge curves should be the same. 



When a beam or wire is suddenly bent or twisted, the forces pro- 

 ducing the strain being maintained constant, it is known that the 

 strain increases with the time, and if the forces suddenly cease to act, 

 that the strain does not altogether disappear at once, but diminishes 

 gradually. This strain w, we have found can, like differences of 

 potential and current, be most easily expressed as the sum of 

 certain exponentials of the time. There are other analogies with 

 condensers. Rapid tremors given to a strained body, our experi- 

 ments show, cause it to attain its ultimate state more rapidly ; 

 one example of this is the more rapid soaking out of the residual 

 charge in a Leyden jar, as noticed by Dr. Hopkinson; our curves 

 connecting w and t for strained beams, like those connecting v and 

 t in condensers, or u and t in voltameters, we find to be of a 

 slightly wavy nature, and to be all similarly affected by change of 

 temperature. Like several other experimenters, we have found that 

 the strain in a body at any time depends not only on the forces acting 

 upon the body at that time, but also on all the forces which have 

 acted on the body during all previous time ; so that, for instance, if 

 a beam fixed at the ends has been loaded negatively for some time, 

 if this load be removed and a positive load applied and soon taken off, 

 the beam may be found to have a negative deflection ; or again, if a 

 current be sent for some time through a voltameter in one direction, 

 and then reversed for some time, and finally the battery removed, the 

 discharge from the voltameter, through a galvanometer, will, in cer- 

 tain cases, first be found opposite to the latter charging current, but 

 subsequently it will become nought, and then become the opposite of 

 the first charging current, this experiment being best seen with 

 electrolytes having a certain amount of rigidity. Similar phenomena 

 are also evidenced with iron under magnetisation, and Dr. Hopkinson's 

 experiments on Leyden jars also lead to similar results. 



We have consequently, while investigating the results we have 

 experimentally obtained during the last few years with strained sub- 

 stances, been led to construct the following theory, which is analogous 

 with that of the late Professor Clerk Maxwell for Leyden jars. 



A perfect fluid is such that the only stress which can exist at any 

 small interface which may be drawn in the fluid is a stress normal to 

 the interface, and hence if the fluid is contained between two parallel 

 planes which have a motion in their planes relatively to one another 

 there is no force required to cause this motion, or to maintain it, in 

 the case of a perfect fluid. If the fluid is a gas or a vapour, we can 

 calculate the force which would be required to maintain a given relative 



