414 Messrs. Kei lokibe and Sukeaki Sakai : Effect of 



is due to the friction of the surrounding air and to the 

 viscosity of the wire; but if the oscillation is performed in a 

 vacuum, the damping is wholly due to the latter quantity. 

 No method of calculating the coefficient o£ viscosity of solid 

 metals from the oscillation experiments alone is yet known. 

 From his statical method of experiments, C. Barus •* obtained 

 for the absolute value of viscosity of steel an enormously large 

 value of 6 to 60 x 10 17 . "W. Voigt t proposed a theory of the 

 viscosity of solid substances, and obtained expressions for cal- 

 culating the coefficient of viscosity. Using his experimental 

 data, he obtained for the coefficients of copper and nickel the 

 values 5*48 xlO 7 and 473 x 10 8 respectively. Thus we 

 cannot at present say anything about the correct order of 

 magnitude of the coefficient of viscosity. 



The following method of calculating the coefficient of 

 viscosity of solid metals, which is due to Prof. K. Honda, is 

 very simple, and gives corrected values of the coefficient in 

 question. 



In the equation of motion of the oscillating system before 

 referred to, the second term is obviously the resisting 

 couple, arising from the viscosity of the wire and acting at 

 its lower end. This couple, acting in an elementary ring- 

 shaped area of the section of wire bounded by r and r + dr, 

 is 



dv 

 dz* 



d()=2irr 2 drr } 



where r) is the coefficient of viscosity and z the direction of 

 the wire. Since v is the tangential velocity of the ring and 



equal to k 



dO 



' dt 



varying linearly with z, we have 



dv _ r dO 

 dz~ldt ] 



•'• 



0: 



.(•*-•?*]>-■*-..*: 



Hence 





■ '='#• 



* Phil. Mag. (5) vol. xxix. p. 337 (1890). 

 t Wied. Ann. vol. xlvii. p. 671 (1892). 



