Considering both the above tables, A always receives more water than its share, while B and C on the whole receive too little. Considering table V. by itself, matters might, perhaps, be set right by altering the total number of days from 14 to 13 or 12, but this, besides being somewhat objectionable for the reason already given, might not improve matters when table VI. came into operation. It is desirable to avoid frequent changes or complicated rules. It is objectionable to make any turn consist of other than a whole number of days. The shifting of the regulator gates is begun at sunrise, a time when officials are about and can see what is happening. All gauges are read early in the morning, and those at regulators are read after the regulation has been done and the flow has become steady. If any regulation were done in the evening, the entry in the gauge register of that day would convey a wrong impression, and the discharge would be incorrectly booked. Moreover, any system of regularly booking evening as well as morning gauges leads to swelling of the already voluminous gauge register. The best method of adjusting matters is to make slight alterations in the full supply gauges. Suppose the normal full supplies in all three branches to be 6 feet. When table VI. is in operation the full supply of A can be reduced to about 5·8 feet. This would give, during the first 5 days, less water to A and more to B, and there is the further advantage that a very small supply, 200 c. ft. per second, is not run in any branch. As regards table V., branch A never receives full supply. This is a rare case.[31] If it were safe, as it might be, to run slightly more than full supply in C, this could be done, and it would increase the supply in C during the last four days and reduce that in A. Otherwise a certain gauge would have to be fixed for A which would give it less than 2,200 c. ft. per second during the first 5 days, and the balance would go to branch B. Similarly, the gauge of B could be slightly reduced, and this would increase the balance going to C. The orders given to the gauge reader are, as before, to send the full supply down one channel, and the balance to the next. The only additional procedure necessary is to inform the gauge reader from time to time what the full supply gauges are. In any case such information has probably to be conveyed to him at times because the channels undergo changes, and the discharge corresponding to a given gauge also changes. [31] The total discharge, 2,200 c. ft. per second, assumed, is very low compared with the full supply of 6,000 c. ft. per second.