2 BULLETIN 126, U. S. DEPARTMENT OF AGEICULTUEE. 
UNLINED CANALS. 
The census of 1910 showed for that year 81,837 main and lateral 
ditches, aggregating 125,591 miles in length and having a maximum 
capacity of 618,097 second-feet. Assuming that not over 4 per cent 
of the total volume of water used in irrigation was carried in pipes, 
flumes, and lined canals and deducting all channels having imper- 
vious linings, there remained over 120,000 miles of unlined irrigation 
canals in the West. 
Transmission losses in the channels considered herein may be 
grouped under the headings leakage, evaporation, aiid seepage. 
Cheap and faulty structures are a common cause of leakage, but such 
losses are only a small percentage of the total loss. 
A large percentage of the water used to moisten the top layer of 
soils is evaporated. 1 Water flowing in an open conduit evaporates 
from the surface an amount dependent upon the temperature of 
both air and water, the velocity of the wind, and upon other factors. 
This loss, as in the case of leakage from faulty construction, is so 
small that it may be neglected without causing appreciable error. 
Evaporation data obtained from 37 different stations throughout 
the various arid States and covering the months of June, July, August, 
and September give average daily rates in inches for this period as 
follows: Maximum, 0.34; minimum, 0.18; and mean, 0.26. For prac- 
tical purposes the loss of water from a canal through evaporation 
is a negligible quantity. 
SEEPAGE LOSSES. 
The results of measurements of seepage show that this is the most 
important source of loss from canals. 
As Table I indicates, in many cases full data are not available, 
particularly as to the character of the materials through which the 
canals are excavated. The canals measured vary widely in capacity 
and with one exception are situated in various parts of the arid regions 
of the United States. About 1,500 miles of separate canal sections are 
represented in the data collected and considering the character of mate- 
rials, erosion, age, etc., they cover a wide range of conditions. It must 
of course be recognized that with measurements taken under such 
diverse conditions and by the different methods used in collecting these 
data allowance must be made for probable inaccuracies in the results. 
It may, and probably does, happen that in cases where the amount 
of loss or gain is small, the variation may really amount to a gain 
where it is given as a loss and vice versa depending on the accuracy of 
the methods used and the care taken in making the measurements 
for a given canal. However, in collecting these data an effort was 
i U. S. Dept. Agr., Office Expt. Stas. Bui. 248. 
