96 Mr. F. Simeon on the 



Theory. — For the flow o£ liquid through a horizontal tube 

 we have Poiseuille's equation 



„_ (Pi— P*)irt .a 



v WQ7" 



where 77 = coefficient of viscosity of liquid, 

 Pi ~~p2 = driving pressure, 



Q = quantity transmitted in time t, 

 r = radius o£ tube, 

 and I = length of tube. 

 In the experimental arrangement there is a constant 

 driving pressure II, and a variable pressure due to the 

 difference in level of the liquid in the vertical tubes. The 

 resultant pressure varies from (H+pgJi) to (IL—pgh) during 

 a reading, where p = density of liquid, and h = change of 

 level in either tube. The average pressure is II, but there 

 is a small influence of the variation of pressure which can be 

 calculated. 



On integration, the expression for r\ becomes 



7j = k.t. 



n U-pc/h 



A 



where fc= ^ A7 =const. for the particular apparatus, A being 



"the area of cross-section of each vertical tube. 



There is another constant of the apparatus, viz. ~ = a. (say), 

 and so U 



v = k.t. 



1 + a/o 



^T^Tp 



If the suffixes s and w denote values for solution and water 

 respectively, we have 



Vs_ts ps^ log ( 1 + «/P«0 - log (1 — ap ir ) 

 Vv, t w ' p w ' log (1 + «/>,) - log ( 1 - ap s ) ' 



As a test of the accuracy of the method, the average times 

 of passage of distilled water at 15° C. and at 20° C. were 

 found. Assuming Thorpe and Rodger's value at 15° C, the 

 value at 20° C. was calculated from the ratio of these times, 

 and it was found to differ from their value at 20° 0. by not 

 more than ^ per cent. 



A series of measurements of the viscosities of calcium- 

 chloride solutions of varying concentrations up to saturation 

 was made at 15° C. The following results were obtained. 



