30 



Trans. Acad. Sci. of St. Louis. 



These values are the means of 20 observations made on a 

 day that was peculiarly favorable by reason of absence of 

 disturbing wind. A polar curve representing the observed 



values is shown in Fig. 2. 



The values may be represented by a series of cosine har- 

 monics, but it requires more than ten terms to represent the 

 values, and an equation of this kind is of no particular value. 



It is evident from these results that when the Pitot tube is 

 used for determining the velocity of flow in pipes it may be 

 combined with two other independent tubes on either side, 

 making an angle of 60° with it and an angle of 120° with each 

 other. The Pitot tube responds to the combined effect of 

 statical pressure, and that due to velocity. 



It should be so placed that its reading is a maximum. The 

 other two tubes should show equal pressures, and represent 

 actual pressures, independent of velocity effects. These tubes 

 must be so placed that the middle tube does not appreciably dis- 

 turb the stream lines, passing over and into the mouths of the 

 lateral tubes. For small pipes this is not easily done. It is 

 better in such cases and in fact in all cases to use a sinerle 

 tube as the Pitot tube, and to place at a near point a disk 

 collector, like that described in this and a former paper pre- 

 viously referred to. 



