Viscosity of Solids and its Physical Verification. 215 



case of hard-drawn wire (Ag, Fe, and gernian silver) minimum 

 viscosity is found associated with maximum susceptibility to 

 change of temperature. 



18. Following the suggestion of § 14, it may be inferred 

 that in the case of very complex molecular structure, insta- 

 bility of configuration will be a more probable occurrence 

 than in the case of simple bodies. Conformably with this 

 view the complex organic solids * like silk will, cait. par., 

 show more pronounced viscous deformation than metals or 

 mineral solids. These known facts are thus in general ac- 

 cordance with Maxwell's theory. Nor is it remarkable that 

 a complex substance like glass should lie somewhere between 

 hard steel and soft steel in the scale of viscosity, showing 

 therefore greater viscosity than hard steel and less viscosity 

 than the soft non-carburized metal. 



1 9. Maxwell's theory lends itself at once for the explanation 

 of superposition of viscous deformations, inasmuch as the in- 

 terpretation given is independent of the special peculiarity of 

 the strain to be discussed. I will adduce a few magnetic 

 results which bear upon this point. 



In considering the permanent effects of temperature on the 

 residual magnetic induction in hard saturated steel, Dr. 

 Strouhal and I f found it necessary to discriminate sharply 

 between two species of magnetic loss : — 



(1) The direct effect due simply to thermal action on the 

 magnetic configuration; 



(2) The indirect effect due to the action of temperature in 

 producing mechanical annealing. 



These two kinds of loss of magnetic induction often occur 

 together. Considered separately the latter, cwt. par., is very 

 decidedly the greater in amount, and its character typified by 

 the concomitant phenomenon of mechanical annealing. The 

 former is not only smaller in relative magnitude, but subsides 

 completely within a much smaller interval of time. In 

 general, the occurrence of permanent magnetism in hard steel, 

 in its thermal relations is subject to nearly the same laws of 

 variation as those adduced in §§ 9 to 13 for ordinary mecha- 

 nical strains. Instability of the carbon configuration is more 

 seriously detrimental to magnetic permanence than is insta- 

 bility of thermal configuration. 



If the unstable carbon configuration be removed by thorough 

 annealing at 100°, then the cold, hard, resaturated magnet 

 must show exceptionally good magnetic stability, as regards 



* Cf. fur instance, Kohlvausch, Pogg. Ann. cxxviii. p. 414, 1866, and 

 many others, 

 f S. and B., Wied. Ann. xx. p. 662, 1883. 



