THE KIVEK F1SHEK1ES OF MAINE. G75 



river beds iu the western part have a slightly steeper incline, and some in the eastern a more 

 moderate one. The declivity is in general well distributed through the course of the river, but is 

 in nearly all cases greater at a distance from the sea, and its uniformity is much broken by the 

 occurrence of abrupt falls and lakes. The falls are not often precipitous or sufficient to prevent 

 the ascent of fish, though this sometimes occurs, as at Hiram on the Saco, East Rumford on the 

 Androscoggin, and on several tributaries of the Keunebec. 



LAKES. The great number of lakes that dot the surface of the State forms a striking feature 

 in its topography. On the published maps there are represented 1,620, varying in size from 100 

 acres to 120 square miles, and it is probable that several hundred more above the minimum in 

 size are unrepresented. Their depths are in no known case very great, but few of them exceed 

 100 feet, and many with an area of several square miles have in no place a greater depth than 

 50 feet. One of the deepest is Lake Sebago, which has in one spot 410 feet of water. The 

 lakes exert a favorable influence on the rivers as homes of migratory fishes in various ways. They 

 serve as regulators of volume, preserving from extreme low stages, and as clarifying basins in 

 which the sediment borne by the tributary streams is deposited. They afford breeding grounds 

 for alewives and possibly retreats for salmon during the summer preceding and the winter follow- 

 ing their spawning season. 



RIVER BEDS. The stony character of most of the soil of Maine and the rapid flow of the 

 rivers combine to give the latter in general a clear hard bed of rock in situ, or of bowlders and 

 gravel, with, however, occasional stretches of muddy bottoms and banks. 



In the western part of the State the lowland adjoining the rivers is often composed of sand, 

 which in that case forms the banks and sometimes part of the river bed. As a whole these rivers 

 abound in the gravelly rapids to which salmon resort for spawning purposes. 



PUEITY OF WATER. There is in the soil of the State comparatively little material that is 

 readily taken up and held long in suspension by the streams and rivers. The occasional turbidity 

 following a sudden and excessive fall of rain in the disforested districts soon disappears through 

 the precipitation of the earth in suspension. The lakes also are immense clarifying basins, and 

 discharge waters of great purity. The waters are, in most cases, strongly tinged with brown from 

 peaty or earthen solutions, and the general color of the water, both of rivers and lakes, when seen 

 in great depth, is an inky blackness, though when examined in a small receptacle, as iu a goblet 

 or pail, they appear of crystal purity. 



VOLUME. Of the total rainfall it may be assumed that 65 per cent, is evaporated and the re- 

 maining 35 per cent, is discharged by the rivers into the ocean. This calls for a mean delivery of 

 67.44 cubic feet of water per minute (35,452,000 cubic feet per annum) for each square mile of area 

 drained. Applied to the Penobscot, this ratio indicates a mean discharge of 498,000 cubic feet per 

 minute. 



The fluctuations in the volume of a river are governed by several circumstances, of which we 

 may mention, first, the irregularities of precipitation ; second, the variation of the evaporation with 

 the changes of weather and the season ; third, the area of storage basins in the shape of lakes, and 

 the extent to which they are brought into use. Many rivers of the second class have such exten- 

 sive storage basins that the volume is very constant, the variations of level rarely exceeding 2 or 3 

 feet. But on the Kennebec and Penobscot a variation of 8 feet within four months is not very 

 rare, and a difference of 6 or 7 feet between flood and drought is the rule. For instance,* on the 



'These data are furnished me by Mr. W. W. Fellows, eiigineer in charge of the water works, from pronl 

 observations. 



