350 Messrs. J. D. Fry and A. M. Tyndall on 



pressure P — p indicated by the gauge is therefore 



pv 2 pv 2 



From this K may be calculated. It will be noticed that if 

 its value is exactly unity there will be no gauge deflexion, 

 and for this reason this method possesses an advantage over 

 the second method. 



In the second method the pitot was placed parallel to the 

 axis of a pipe of radius R through which air was flowing at 

 a constant rate, and a series of pressure readings obtained 

 by moving the pitot along a diameter of the pipe. 



If p r is the pitot pressure and v r the velocity of flow at a 

 distance r from the centre of the pipe, then as before 



+ = W\ 



P ' 



The quantity of air flowing per second through a ring of 

 thickness dr at this radius is therefore given by 



2ir rdr 



P 



d q =K^l 



Integrating this across the section of the pipe, we have 



'.it ra 



where Q is the total quantity of air flowing through the 

 pipe per second. 



Unless K has a constant value at all velocities it cannot 

 be taken outside the integral, since the velocity of the air in 

 the pipe varies at different points along a diameter. If K is 

 not constant and this is done, the values of K obtained are 

 purely artificial and depend not only on the quantity of 

 flow through the pipe but also on the distribution of that 

 flow across the section. 



The symbol K x will be used in what follows for the 

 artificial values obtained in this way. 



The only justification for using this method to measure K 

 itself lay in the fact that previous experimenters had found 

 that K had a constant value over a limited range, and there 

 appeared to be no theoretical reason why any marked 

 change should occur when the range was extended. 



The observed values of K obtained by the centrifugal 

 method and of Ki by the pipe method are shown in 



