452 Mr. Owen Glynne Jones on 



extent of only *03° C. in an Lour and a half, and good results 

 were then obtained for the viscosity of glycerine. Within 

 the last fortnight the main temperature has not been nearly 

 so constant, and more has been learnt concerning the varia- 

 tion of viscosity with temperature than the absolute value of 

 viscosity for any definite liquid. 



The mass of a sphere of mercury suitable for gtycerine 

 would be between '003 and *010 grm. This is too small to 

 be determined with sufficient accuracy by direct weighing, 

 with a balance estimating to one tenth of a milligramme. 

 But there is no necessity to determine such a mass directly. 



Let a mass of mercury weighing about *1 grm. be taken 

 and weighed carefully. Let it be divided into about 12 parts, 

 approximately equal to each other. Let the speed of each 

 through a column of very viscous liquid, such as glycerine or 

 clear castor-oil, be determined. Then, using Stokes's result, 



a 2 



v 



2 



■■■ /*«-;• 



and for a given liquid with constant viscosity 



m oc v%, 

 or m = Jcv%. 



Hence 2m = k% v%. 



But Sm = M, the whole mass of mercury, and 2^2 = Vs the 

 speed that the whole mass would have if it fell as a sphere 

 through the infinite liquid. Perhaps taken as a whole the 

 mercury would not fall as a sphere, and the liquid in the 

 burette would not be as of infinite extent ; but by this sub- 

 division such errors are avoided. 



We thus get the speed V corresponding to the mass M, 

 and these values can be applied at once in the equation 



2 2 (°— p) 



or 2 / 3M \ig-p 



~~9 ff \4<ir<r) ~T~' 3 



far ivr 4 3 12 / 3M \§ 

 M= o7r<rcr, and a 2 = 7r[- ) 3 . 



3 \47T<7/ 



