[satterly] 



VISCOSITY EXPERIMENT 



121 



TABLE II 



The best way of getting the average from the flow results is to 

 plot a graph between the pressure head (in cms. of water) and the 

 volume of air passing per second. The graphs for the three tubes 

 are shown in Fig. 2. It is noticed that at the lower end of the graphs 

 the curves are straight showing the direct proportionality between 

 p and V, but at the higher end the graphs turn towards the axis of 

 pressure showing that an increase in pressure at this stage does not 

 send through a proportional volume of air. 



Therefore to get the coefficient of viscosity the best straight line 

 is drawn through the first part of the graph and the slope of this 

 line obtained. 

 From Fig. 2 we see that for a tube A 



p/V= .190 in appropriate units. 

 Substituting in equation (4) we get 



TT (.078)4 



r?- o X(. 190X980) X-TTT =00019 C.G.S. units, 

 o 14.1 



For tubes b and c we get in the same way 



Tube (B) p/V= .173 



and V = .00019 C.G.S. units 



Tube (C) p/V= .158 



and ri = .00020 C.G.S. units. 



The results are in close agreement and with the correct value of 

 r? given in books of tables. This value is about .00018 + at 20°C. 



Osborne Reynolds^ showed that the formula quoted above in 

 equation (1) was true as long as the motion was "stream-line" 

 motion but that once a critical velocity was exceeded eddies were 

 set up and a more than proportionate increase of pressure was required 

 to send through a given volume of fluid. He and others have shown 



^Osborne Rej'^nolds, Phil. Trans. 1883, Poj'nting and Thomson, Properties of 

 Matter, ch. XVIII. 



