[kekfer] PIïKSIDENTIAL ADTRrsS 9 



and assume the proportion of this precipitation ( of rain and snow ) 

 which, after deductions for evaporation, the demands of vegetation, or 

 infiltration, wonld reach the Avheels. An allowance must also be made 

 for that portion of the rainfall which may be carried off in floods. 



The area over which this precipitation would be in reach for water 

 power purposes, would embrace all the main land of Canada south of the 

 St. Lawrence, as well as all north of it in the St. Lawrence valley, and 

 so much of the Hudson Bay and Mackenzie Eiver watersheds as can be 

 utilized, or imported by transmission. 



As regards the power of the water thus estimated, we must embark 

 in a much more speculative estimate as to the average fall which should 

 1)6 assigned to it for the whole region. We have in the undeveloped 

 districts some scattered meteorological observations to assist us in esti- 

 mating probable rainfall, and we have also a few barometrical observa- 

 tions giving the height above sea level of summit waters. On lower 

 levels we have more numerous rain gauges, and summit levels ascertain- 

 ed by railway surveys. 



For the whole river the total fall may be less than 100 feet, as in the 

 case of the French river which has Lake Nipissing for a mill pond, or 

 rise to 1,500 feet or more as at the rivers below Anticosti. In the case 

 of the French river (which is the lower part of a longer stream) we have 

 surveys, and know that its whole fall can be utilized, as would be 

 done if it is made navigable by locks and dams. In the 

 others (where no surveys have been made) some will be more or less like 

 French river, while at others only a portion of the total fall upon them 

 may be profitably utilized. The most valuable will be those which, like 

 jlilontmorency, bring all their water with sufficient head to the point 

 where it is worth most. The upper sections of the rivers will be the 

 least valuable, as having less water and being more remote until reached 

 by a nev/ railway, or a transmission wire. 



The chief difficulty with respect to the quantity of water is the want 

 of rain gauges over so great an extent of unoccupied territory. Where 

 the rainfall is known, the proportions which reach the streams have been 

 ascertained in connection with reservoirs for water supply and other 

 purposes. We can therefore only state a hypothetical case especially 

 as to the power to be assigned to the available water. 



Assuming, however, an average annual precipitation of twenty-four 

 inches and taking one-half of this as available for water power, every 

 ten square miles would yield an average of nearly one horse power for 

 every foot of fall. A million square miles (and there is much more) 

 would give nearly 100,000 horse power for every foot of fall. As there 

 would be several hundred feet of fall which could be utilized, our water 



