— 8 — 
• u SS ti , }e . statu 1 t ? establishing the ratio between the statute 
1 ch and the cubic foot per second in such terms as to make 
i f, a , tute lnc .hes equivalent to one cubic foot per second 
should be considered as defining the inch. This ratio may be 
very widely variant if the inch is measured according to stat¬ 
ute terms. The cubic foot per second is an absolute unit 
whose quantity cannot be subject to dispute, though the accu¬ 
racy of. measure may be. The state laws provide that in 
appropriating water to ditches the quantity shall be estimated 
as Second 6 fee 1 ? 6r SeCOnd ’ or as frequently shorter expressed, 
divisors. 
As oidinarily constructed, the division can rarely be 
exact, but, frequently, the convenience of an approximate di- 
vision more than counterbalances any inaccuracy there may 
f ' e • rare ly have occasion to use divisors, 
:°-K J Cn lf dl £ ch has to P ro ra te the water, a better dis¬ 
tribution can be effected by means of modules. If the water 
is to be divided into two equal portions, by placing the two 
lateral ditches in identical relations to the main dftch, in a 
tught and uniform channel, the division is exact. Empha¬ 
sis should be laid on the identical relation, for many divisions 
are seen where the conditions are not the same, as, e. one 
ranch continues straight, the other may make an abrupt 
turn, one may pass through a covered box, etc. In these 
cases some advantage is given to the ditch having the freer 
discharge. The effect of these differences is greater than is 
fhpQp 1 ” 3 ^ll y supposed. It is, however, generally easy to meet 
these conditions if the parties desire. In the same way the 
water may be subdivided into four, eight or sixteen equal 
parts. But where it is required to divide the water into two 
unequal, or into three or more portions, equal or not, the di¬ 
vision becomes one of approximation only. The difficulty 
arises from the fact that the water has not uniform velocity 
across the whole channel, the center has greater velocity than 
lat near the banks. If, therefore, equal openings be made 
across the channel those near the center have the greater dis¬ 
charge. Making the central openings smaller only partially 
evades the difficulty, for as the relative velocities of the cen 
ter and sides differ with different depths, this arrangement 
would still be inexact for any one depth except that for which 
In its most common form the divisor consists of a parti¬ 
tion dividing the channel into two portions in proportion to 
the respective claims. This, in effect, assumes that the veIo ° 
