14 The Colorado Experiment Station. 
water uniform across the entire section of the ditch. Great pains 
are taken in some of the old Moorish boxes in Spain, the methods 
running back for centuries. Special care is taken to make the ditch 
straight for a distance of 50 or 100 yards above the point of divi¬ 
sion, paving the sides and bottom and increasing the cross section 
so as to lessen the velocity. In such cases the division is made quite 
exact. The principal error is then most likely from the difference 
in the freedom with which the water passes down the main ditch 
and enters into the side channel. If there be a drop at B-C, the 
effect from this cause may be neutralized or made insignificant. If 
there is no drop at B-C, the retardation in the lateral channel may 
have considerable effect and the change in favor of the main channel 
may be quite noticeable. 
In some cases, however, this is just, for the users who receive 
water farther down the channel are subject to greater losses. If 
equal owners are to receive equal amounts then the user nearest 
the head of the canal should be given less than the amount pro¬ 
portional to his interest to allow for loss in carriage for the greater 
distance. Whether this is the proper basis of division depends 
upon the understanding of the users, and very possibly would be 
different in different ditches. The users only can determine 
whether the individual is to bear the loss of carriage to his head- 
gate, thus giving advantage to the upper user, or whether this loss 
is to be shared by all. The special circumstances of each case may 
determine which is the just principle for that ditch, depending on 
the method of construction, the history of the development and the 
understanding. 
<0 
Fig. 1.—A Common Form of Divisor. 
Figure 1 represents one of the most common forms 0+' divisors. 
The partition board A is movable, and may be placed at different 
distances from the side C, so that the user can vary the proportion 
