The regulation should usually be so effected that rushes of water in any portion of the channel are avoided, but if scour occurs in a particular part of the channel it may be necessary to try and obtain slack water there. Until it is proved by experience that they are unnecessary, soundings should be taken periodically downstream of large works. When a branch or escape is closed the leakage should be carefully stopped. The necessary materials should be always kept ready in sufficient quantity. 3. =Gauge Readings and Discharges.= For the head gauge of each distributary and for certain gauges in the canals, discharge tables, based on actual observations, are prepared. If changes occur in the upper part of a channel, the discharge corresponding to a given gauge reading is altered. One remedy for this is to have a second gauge downstream of the “silt wedge” or scoured or narrowed reach. The indents are then made out with reference to the second gauge, but any slight adjustments due to fluctuation in the water level of the canal, are effected by means of the head gauge. Unless the zilladar and Subdivisional Officer are on the alert, the gauge reader is likely to evade going to the lower gauge every morning, and to enter fictitious readings for it, inferring them from the readings of the head gauge. If there are any outlets between the two gauges, their discharge has to be observed or estimated and added to the discharge of the distributary as entered in the table corresponding to the readings on the second gauge. The above system can be worked with advantage in cases where the distributary bifurcates two or three miles from its off-take. The men in charge of the two regulators can work together, one of them or an assistant, going daily from one regulator to the other and back. Usually, however, the vitiating of the discharge table at the head gauge has to be faced, and the table to be constantly corrected. It is impossible to frame beforehand any rule or formula which would give a certain correction for a certain depth of silt deposit. Moreover, there might or might not be a contraction of the channel due to deposit on the sides. The usual plan is to observe a discharge some time during each month. If the result is in excess of the tabular discharge, all the discharges for that month are increased in the same proportion. They can be booked according to the table and totalled, and the correction applied to the total.