THE FLOW OF WATER IN CONCRETE PIPE. 15 
be excluded. Herein lies one difficulty in making tests on concrete 
pipes in commercial service. 
There are two general ways in which access to the stream within 
a pipe may be secured. A small tube like A or B, figure 4, may be 
thrust directly into the stream, or a connection set in the shell of 
the pipe, like C. Tubes of type A or B were smoothly tapered, 
with the holes passing through the opposite sides of the tube. When 
a connection was to be made far from the end of the concrete pipe, 
a hole was tapped in the latter and the tube inserted through a 
stuffing box. . According to White 1 a tube of type A not only 
registers the proper pressure head, without velocity-head influence, 
in an open channel, but also within a closed pipe. His third con- 
clusion is that Pitot tubes whose constants are unity in open canal 
ratings will remain unity, whatever the pressure of the liquid. His 
tubes M and N are of this type. Our tube of type A was, in essen- 
tials, the same as White's static tubes of types M and N. Although 
our tubes of this type were tapered with some care, Lawrence and 
Braunworth 2 showed that the taper was of little importance. 
It is to be regretted that Mills 3 did not continue his experiments 
on piezometer connections to cover tubes like A and B. He did, 
however, establish the fact that a connection of type C registers the 
proper pressure head. 4 
The use of an orifice in the shell of the pipe, normal to the curve — 
that is, of type C — is justified by F. P. Stearns 5 and later by Desmond 
Fitzgerald. 6 
Carrying the idea still farther, Marx, Wing, and Hoskins proved 
that — 
"When the pressure in the given cross section of the pipe everywhere exceeded 
that of the atmosphere an open piezometer will stand at the same height at whatever 
point of the cross section it be attached, and whether it communicates with the pipe 
at one point or several." 7 
In each of the experiments described in this paper, regardless of 
the auspices under *which it was conducted, but one connection was 
made at each end of the reach tested. • 
i The Pitot Tube: Its Formula. W. M. White, Journal Assoc. Engin. Socs., 27 (1901), p. 35. 
» Trans. Amer. Soc. Civ. Engin., vol. 57, p. 273. 
3 Experiments upon Piezometers Used in Hydraulic Investigations. H. F. Mills, Pro. Amer. Acad. 
Arts and Sci., vol. 14 (Whole Series, 1879), p. 26. 
* Mills concludes "that with an orifice whose edges are in the plane of the side, and passage normal thereto, 
the piezometric column will stand neither above nor below the surface of the stream, but will indicate the 
true height of the surface . " He also notes that a very slight variation of the passage from the normal causes 
an erroneous reading of the piezometer column; that where the orifice projects into the stream the column 
rises above the surface of the stream when the orifice faces upstream or is at right angles to the stream and 
the column does not rise to the elevation of the surface in the stream when the orifice faces down stream. 
6 Trans. Amer. Soc. Civ. Engin., 14 (1885), p. 3. 
• Id., 35 (1898), p. 241. 
7 Trans. Amer. Soc. Civ. Engin., 40 (1898), p. 526. 
