Manchester Memoirs, Vol. Iviii. (1914), No. 7- 3 



shaped so as to disturb the stream lines as Httle as 

 possible. These are conditions easy of fulfilment. 



It very often happens, however, that /t is of the same 

 order as v'l2g, whence great inaccuracies result in deducing 

 V unless h can be measured very accurately, or eliminated 

 by using the facing gauge at practically atmospheric 

 pressure. The latter, however, is often impossible, and 

 it is thus necessary to measure h accurately. So the 

 accurate measurement of velocity is seen to depend upon 

 the accurate measurement of static head or pressure. 



II. Measurement of Air Pressure {h). 



Measurement of pressure is usually attempted by pro- 

 viding an orifice whose plane is as nearly as possible 

 parallel to the stream lines at that point, this orifice being 

 in communication by means of a small tube with a bour- 

 don gauge, manometer or other instrument of a similar 

 nature. Since the insertion of such a tube in the stream 

 itself tends to disturb the stream lines, it is usual to make 

 both tube and orifice as small as possible. So long as 

 the effect of any disturbance due to the presence of the 

 tube is small compared with the pressure measured, such 

 a method is satisfactory and practical. But there are 

 very many circuits in which this is not the case, particu- 

 larly when the velocity of the air stream far exceeds the 

 limit of steady flow. Then is observed that phenomenon 

 known as " induction effect," which is in reality a lower- 

 ing of the pressure to be recorded by a counter pressure 

 due to the rush of the fluid past the orifice. This effect 

 may be so large as to mask altogether, or even to reverse 

 the true pressure reading, and no change in the form or 

 sensitiveness of the gauge will help matters. Thus, with 

 air circuits for ventilation and dust-collecting purposes, 

 where the total pressure above or below atmosphere does 



