FISHERIES, GAME AND FORESTS. 517 



snow falling during the winter months to so far waste away under the influence of 

 unobstructed winds as to be, when the spring break-up comes, considerably less in 

 quantity than it would be if the area had been protected by heavy forests. The result 

 is that while the snow will melt quicker under the influence of higher spring 

 temperatures than it would if in the forest, still the quantity to be melted is, by reason 

 of the said gradual wasting, considerably less than where the opposite condition 

 obtains. On the other hand, with the snowfall protected by a forest covering, the 

 wasting away at time of high spring temperature is slower than on an exposed area 

 subject to the same meteorological conditions. The forest meteorological observations 

 show that this conclusion must be essentially true. We have then a set of conditions 

 in which the effect of the forest in increasing the tendency to extreme flood at the 

 time of the spring break-up by the storage of large quantities of snow, is modified by 

 the greater loss from evaporation taking place on a deforested area during the entire 

 winter season. Hence, as regards floods caused by the sudden melting of winter 

 snows, there is an essential balance, the net result being, in the case of deforested areas, 

 a shortening of the time of extreme spring run-off,, with, by reason of gradual wasting 

 away during the winter, ordinarily no special increase in height of floods. 



As regards floods due to heavy rains either in the spring or fall, and which occur 

 without reference to the spring break-up, a similar line of reasoning applies. 

 Exposure of the ground of the deforested area leads to a more rapid evaporation from 

 the surface, with quicker exhaustion of the surface moisture. Hence, usually, the 

 ground in the open is in good condition at the beginning of an extreme rainfall to 

 absorb considerably more water, before large quantities begin to run off, than it would 

 be if covered with forest. In this way a balance of spring, summer, and fall flood 

 conditions is likewise attained on the deforested area the same as during the winter. 

 The rational conclusion appears to be easily drawn therefore (i) that while deforesta- 

 tion does actually decrease the net annual run-off of streams, still (2) it does noi per se 

 necessarily materially increase the height of floods. As meeting, therefore, the 

 objection of Mr. Roberts it remains to point out that the mere fact of non-increase 

 of floods is absolutely no proof that there is not a material decrease in total run-off 

 from deforested areas. 



In considering the foregoing views we must not overlook that thorough under 

 drainage of catchment area may so decrease the time of maximum run-off as to 

 increase flood heights. This is especially true in any region where large open 

 drainage ditches have been cut. It is within my experience that the regimen of a 

 stream has been entirely changed by drainage alone. On this point a large amount 

 of interesting and useful data is at hand, but which cannot for lack of space be 

 introduced here. 



