THE FLOW OF WATER IN CONCRETE PIPE. 99 
It is suggested that concrete pipe, especially where it is to be laid 
on or near the sui'face of the ground, should where possible be built 
in cool weather rather than hot weather, both because of the tem- 
perature of the air and also on account of the temperature of the 
water. If the pipe is built in warm weather it is apt to shrink in 
cold weather and when cold water is run through it in such a way 
as to develop objectionable circular cracks. 
In 1907 or 1908, at the time the Los Angeles Aqueduct was being 
designed, investigation was made to determine the value of the 
coefficient n in the Kutter formula in a number of the conduits of 
southern California. The result of these investigations was pub- 
lished about that time in the Engineering News over my signature. 
An interesting feature that developed as a result of these investiga- 
tions was that the carrying capacity of the ditch and the value of 
the coefficient varied immensely with the question of whether the 
particular ditch or conduit was exposed to the rays of the sun or 
whether it was covered. If it were covered, the lining kept clean 
and smooth, but when it was exposed to the action of the sun a 
vegetable growth immediately set in. The growth resembled both 
the lichens on stone and long streaming grass. The result was that 
we found values of n of about 0.012 for ordinary covered conduits, 
and anywhere from 0.014 to 0.018 for these uncovered conduits. 1 
It was also observed that in. ditches where the sand-box arrange- 
ments were poor that small bars of sand would follow down the 
conduit or even through the pipe lines and very materially, in fact 
totally, change the carrving capacities of the ditch. In other words, 
the capacity of the ditch js affected by the provisions that are made 
to remove sand and gravel that might otherwise enter. 
My experience is that plastered conduits of any kind should if 
possible be covered. It has been found by experiments made, I 
think at the University of Michigan, that the expansion and con- 
traction of concrete varies greatly with the richness of the mix, the 
richer mixes expanding more than the leaner mixes. It is customary 
to make the plaster richer than the main bulk of the concrete in 
order to get a smoother surface. The different rates of expansion 
and contraction of the body of the concrete and the plaster is largely 
responsible for the scaling of the plaster. If a plaster is desired it 
should have about the same ratio of sand to cement as is used in the 
main mix between sand and cement, and much more labor should 
be put on than ordinarily in producing the smooth surface. By 
careful forming in conduits and by troweling it is found to be possible 
in many cases to obtain a perfectly smooth surface without the use 
of plaster. 
I believe that it is entirely possible to build a jointed concrete 
pipe that roots will not enter. A number of years ago I made an 
examination of the irrigation system at Rialto. This was all built 
of cement pipe which had been in service for a term of years, perhaps 
15 or 20. The pipe ran under trees of all classes and kinds — cypress, 
eucalyptus, orange, and lemon trees and deciduous fruits. We cut 
into the pipe in many places and by means of mirrors threw the 
light back through the pipe so that it could be carefully examined. 
The pipe was clean, in every instance showing evidence of course of 
1 The above experiments are described as Nos. 58, 59, and 60. 
