the outlet of Lake Huron had also been lowered. This was the starting point of the investigation into the spoliation of Niagara. The channel of the river had been deepened just after 1890, owing to natural scour by the currents, the effectiveness of which was increased by the powerful jamming of vast quantities of ice against the barriers at the Upper Rapids, immediately above the falls, and to a small extent by the shifting of the boulders on the river bed just below the outlet of Lake Erie.
8. Corrections of Discharge Measurements.—The discharge measurements had been made by the U. S. Lake Survey after the lowering of the lake outlets had occurred, that of the Erie outlet causing the subsidence of the lake level to nearly one foot. As the daily records of the lake fluctuations have been kept for fifty years or more, it was possible to determine the discharges of the rivers throughout that time. Failing to take into consideration this lowering of the lake outlets, the calculations of the river discharges prior to 1891 were excessive, and those of Lakes Huron and Erie showed inadmissible differences; as also found by Mr. Russel. These discoveries reduce the calculated discharge of Niagara River, prior to 1891, by 22,000 cubic feet per second, which would bring its volume from 1860 to 1890 into agreement with the mean measured discharge for the fifteen years from 1891 to 1905, inclusive, or 204,000 cubic feet per second; and the low water discharge has fallen to 160,000 cubic feet. There have been years of high water and others of low, yet by taking groups of years, the mean values are remarkably uniform, but the latter period must not be regarded as one of low water, a fact which I can not too strongly emphasize, although the lake levels have actually been much lower than during the preceding period, due, as just stated,-to the lowering of the outlets. Such years, however, as 1901, showed very low water in Lake Erie, and reduced discharge of Niagara River.
9. Present very High Water.—In contrast with this, the lake-levels during 1907 were extraordinarily high, increasing even till the present month (June, 1908). Fragmentary information, preserved, indicates that Lakes Erie and Ontario were unusually low in 1819. The fuller record of subsequent years shows that the highest water occurred in 1838, and nothing has been comparable to it until the present high stages. Although these do not make the lakes appear to be now so high as seventy years ago, this is because of the lowering of their outlets and the further diversion of water for power purposes, both of which, if allowed for, would bring the lakes to higher levels than at any time since records have been kept. The 1838 period of high water began in that year and continued until 1840, after which the lake levels subsided to normal conditions. There is no reason to suppose that the present extreme high-water conditions will continue longer than