LOSSES FROM CANAL.S BY SEEPAGE. 3 1 
of the water holds the gravel in place, the effect observed 
is a real one. 
In consequence a canal with rising water will have more 
and with falling water less than the normal loss, or more 
than the normal gain. This is shown in numerous cases 
with the records on the North Poudre canal. The length 
of time during which this will affect the conditions depends 
on the area and extent of the gravel beds near the channel. 
The principal effect passes off in a short time, for as the line 
of saturated soil becomes further removed from the chan- 
nel, the movement of water is much more slow. 
One consequence often realized in practice is that if 
water is to be run through a long canal, the division can be 
made better and fairer if the water is run completely 
through the canal before opening the lateral gates. The 
whole of a small stream of water may be required to satisfy 
the thirsty sand. A large stream may accomplish the same 
purpose in a shorter time and with less loss. Hence often 
it is better to use the whole stream if necessary to wet the 
bottom of the canal for its whole length, before beginning 
the division of water, and if the canal is run in sections, to 
begin the distribution at the lower end of the canal is the 
better way. If a small stream only is used, nearly all may 
be taken to wet up the channel and leave little for the lower 
users. 
('ONCLUSIONS. 
1. The losses from evaporation are relatively insignifi- 
cant compared with the seepage losses from most canals. In 
the cases most favorable to evaporation and least favorable 
to seepage the evaporation is not over 15 per cent, of the 
seepage. 
2. In the case of reservoirs it was concluded in bulletin 
43 that the seepage was less important than the evaporation. 
This is different from the results found in ditches, not be- 
cause the evaporation is less, but because the seepage is 
much more. 
3. The losses are sometimes enough to cover the whole 
canal 20 feet deep per day. 
4. The loss in clay soils is less than in sandy or gravelly 
soils, but rarely as small as 3 inches daily. 
5. The loss is greater when water is first turned in 
than after the bed has become saturated. 
6. Sometimes the canals are found to gain for the 
whole or part of their length, or the canals may act as 
drains. This is more likely to be the case when the canal 
