Engineer's reports and the U. S. Water Supply papers (for example, in the State Engineer's Report of 1902 and 1905 and the Water Supply paper No. 97), that zero of gage was set at bottom of the five-foot discharge tunnels.

Our measurements show that the bottom of the gate well is 39.58 feet below the present thirty-six foot mark of the gage, and since the caretaker states that the bottom of well is one foot below bottom of gate, it appears certain that, as the gage now stands, the published statement is wrong.

It is difficult to compute precisely the discharge of these sluice gates from the given head in the lake and the given height of gate opening, because of the great loss of head through the two guard gates that supply the gate chamber leading to each one of the two sluice gates and their liability to become obstructed by snags. We found this loss to be 13.3 feet when the two guard gates to one sluice chamber were wide open, and the one sluice gate was wide open, the reservoir being at gage height 35.0.

From what could be learned, it appears that any estimate of discharge through sluice gates prior to 1907 will be of uncertain precision because of the clogging of the guard gates or their screens with drift wood. Moreover, the hoist on one was defective and one guard gate on well B was kept closed for a long period, of which we find no record. The operation of the iron sluice gates was so difficult that both were replaced at low water in winter of 1906-7.

To determine the loss of head through the upstream gates or guard gates, of which there are two, each about 3 ft. x 5 ft., to each of the separate wells for the two 5-foot sluice gates, measurements were made by opening sluice gate B to various heights and noting gage reading in lake and the height in the well. The results are given in table following.

Because of the inability at this time to close one of the guard gates, it was not possible to drain the well and enter it for precise measurements of aperture of either guard gate or sluice gates, nor was it practicable to make precise measurements of elevation of sluice ways relative to scale height. The standard bench mark was inaccessible under snow and ice.

The values in the last column are about what might be expected for the coefficient of a rectangular gate opening. Therefore, the screens were probably clear.