34 — 
H as the depth to be taken. This form of correction is due to Fteley and Stearns * 
It gives a much larger correction for velocity than is furnished by the Francis cor¬ 
rection, but it agrees much better with measurements I have made for high veloc¬ 
ities. Ihe experiments on which it was based were limited to velocities of 2.5 feet 
per second, lable II. gives the per cent increase in discharge caused by different 
velocities. It will be seen how great this correction becomes, sometimes causing 
an increase of several hundred per cent., and, consequently, shows the importance 
of keeping the velocity within low limits. F 
lo aid in the practical allowance for velocity of approach, two tables have 
been prepared and are printed as tables I and II of the appendix. Table II shows 
the increase in per cent, over the quantities given in tables III and IV by various 
velocities of approach. Ihe increase with a given velocity varies with the depth of 
water over the weir, being greater for small depths. A velocity of one foot per sec¬ 
ond increases the discharge over a weir when the water is flowing 3 inches deep 
over 14 per cent; if flowing over the weir 1 foot deep, only 3.5 per cent. The table 
I shows what the average velocity is as the water passes through the weir or it 
&hows what is the velocity in the channel if the section is the same as that of the 
weir as it frequently is. A comparison of the two tables will show the proper sec- 
fimhi 0 glVe the Channe lnorder that the resulting error shall be within reasonable 
A.s the effect of the causes which tend to increase the dis¬ 
charge cannot be entirely eliminated, the tendency is to in¬ 
crease the discharge. In consequence of this, Cippoletti pro¬ 
posed to increase the amount as computed by the Francis 
formula by i per cent, making the formula, 
p . Q- 3 . 36 %LHF 
but as it seems probable that the coefficient of discharge as¬ 
sumed by him was too large, the ordinary formula will be 
etter to use in the dearth of sufficiency of experimental 
knowledge. 
I he Cippoletti form of weir because of the reasons 
already given has the most advantages of any module 
known to the writer for measurement of water for irrigation 
purposes. It is reliable to within 2 per cent, with the Francis 
formula, it placed according to the conditions given, and 
probably within 1 per cent. The ordinary methods of meas¬ 
uring or guessing at the discharge of water vary from 40 to 
400 per cent., as usually used. All that may be said of its ad¬ 
vantages, save the one of having the effective length of the 
sill in proportion to the actual length of the sill is true of the 
rectangular weir also It meets most of the conditions for a 
good module. It lacks means of self-adjustment, or of pre¬ 
serving constant heights of water. Where adapted, the spill- 
box may be used in connection with it, when that condition 
would be very nearly met. Several canals have introduced 
essentially this combination, and so far as reports have 
reached the writer they have been satisfactory. 
SUBMERGED WEIRS. 
_ In SQm e cases the water in the lower canal may be higher 
* Trans-Am. Soc. C. Eng, 1884. 
