182 0. Barus — Viscosity of Solids. 



the result of partial rotations of molecules of a form other than 

 spherical. 



Mssen's* theory is unique. He cpnsiders the ether in the 

 space surrounding the body, in its relations to the ether within 

 the intermolecular spaces of the body ; and he bases his theory 

 on the conditions under which the external ether enters the 

 said intermolecular spaces, when the body is deformed by stress. 

 He thus obtains both a time and a thermal effect In many re- 

 spects this curious theory seems to me to anticipate Osborne 

 Reynolds'f in recognizing the importance of the " dilitancy " of 

 a granular medium. 



3. Maxwell's J theory would require more extended com- 

 ment ; but the terms in which his views are expressed are so 

 precise, that it is impossible to abbreviate them. cf. §9, 14. 

 Aside from the remarks of the next paragraph, the ideas 

 underlying Maxwell's theory have been given by many others, 

 indeed by almost all the chief writers on solid viscosity ; but 

 Maxwell carries them through consistently to a complete theory. 



I have stated that Maxwell's theory is the analogon of Clau- 

 sius's theory of electrolysis. Where the latter uses " Tlieil 

 molecule" and electromotive force to effect chemical decom- 

 position, Maxwell has unstable configurations and stress avail- 

 able to produce permanent deformation. In Clausius's case the 

 number of decomposable molecule (i. e. unstable configurations 

 as regards the action of electromotive force), in any given case 

 of actual electrolysis, is practically infinite. This corresponds 

 to Maxwell's viscous fluid, hard or soft. In a viscous solid, 

 molecular configurations are present in all degrees of stability, 

 with a sufficient preponderance of stable configuration to con- 

 stitute a solid framework. The relative number of unstable 

 configurations varies with the viscosity of the material. If, 

 therefore, I conceive the case of an electrolyte exhausting itself 

 with respect to electrical conductivity, by the chemical decom- 

 position induced by current, until conduction cease, I have the 

 analogon of a solid which is reaching the limit of viscous de- 

 formation, cf. §21. 



From this analogy it follows that a solid (?) electrolyte is 

 necessarily viscous ; whereas a viscous solid is only an electro- 

 lyte when the molecules break up into parts oppositely charged. 

 Again, a viscous solid (?) is probably more viscous when under- 

 going electrolytic decomposition, than when no current passes 

 through it. § Experiments to the same effect can however be 



* Xissen: Inaug. Dissert. Bonn. 1880. (Not accessible to me.) 



f Reynolds: Phil. Mag., V, xx. p. 469 : 18S5. 



t Maxwell: 1. c. p. 311. 



§ T have since been at some pains to verify this surmise, working with glass at 

 360°. But the experiments thus far are not decisive because the amount of cur- 

 rent passing through glass is not only very small, but soon ceases entirely even 

 in the case of thin-walled tubes (Warburg. Wied. Ann., xxi, p. 622). 



