480 Mr. W. Sutherland on the 



It has been found that the viscosity rj of mixtures is not 

 accurately expressible by a rnixture-by-weight formula such. 

 as v=PiVi+P2V2i but Thorpe and Rodger (Journ. Chem. Soc. 

 lxxi.) find the following type more successful • 



which may be written in the form 



yv=i} 1 p 1 v l + rj 2 p 2 v 2 . .... (32) 



As to variation of viscosity of a pure liquid with tem- 

 perature, it seems to me best to use the equation of Graetz 

 (Wied. Ann. xxxiv.), 



V = A(t c -t)/{t-t r ), .... (33)' 



w T here t c is the critical temperature and t r is a temperature 



some tens of degrees below the solidifying-point. Then for 



water we write 



. 538— f . 368-£ ,„.. 



VV^AspM— — +A lPl v 1TTI ^, . . (34> 



assuming that dihyclrol has t r 150 below zero ; because by 

 analogy the solidifying-point of dihydrol must be low, and at 

 any rate the second term on the right is of less importance 

 than the first. Using the viscosities of water given by 

 Hosking (Phil. Mag. [5] xlix.), we find 



A 2 =-002344, #=33-1, A^'001222, 

 which give the comparison : — 



Table XII. 



t 0° 20° 40° 60° 80° 100° 



10 5 »j 2 p 2 w 2 '.... 1624 837 520 353 257 194 



lO 5 ^^^ 172 159 144 130 116 103 



lOVcalc 1796 996 664 483 373 297 



10 5 jjvexp 1794 1011 662 480 370 297 



The discrepancy at} 20° illustrates the difficulty of getting' 

 mixture-formulae to represent the viscosity of mixtures. 



As to the effect of pressure on the viscosity of our mixture, 

 we can differentiate (32) with respect to pressure/, 





drj dv . .dpo dvi dv„ __ 



= v df +v ^' + {v ^- v ^f-^ v ^ "Wft^- (35)- 



