REVIEWS 639 



that, expressed in round numbers, the proportionate amount of the 

 yearly precipitation in forested and non-forested areas which consti- 

 tute runoff and evaporation, respectively, are as follows : 



In Forest ^ In Open 



Per cent of precipitation Per cent of precipitation 

 Runofif GO 60 



Evaporation (vegetation) 15 10 



Evaporation (earth surface) 5 24 



Transpiration 20 6 



100 100 



The results of the Emmental experiments are, of course, strictlv 

 applicable only to regions with similar climatic and soil conditions. 

 The analysis, however, of the factors which determine the run-oft' on 

 forested and open soils, as well as the manner in which the experi- 

 ments were conducted, place the results on a general scientific basis. 

 Many of the conclusions drawn by Dr. Engler may be regarded as 

 generally applicable, while other deductions must not be generalized. 

 There are so many factors working together in various ways in aft'ect- 

 ing the regimen of a stream that it is essential to guard against careless 

 application of the Swiss results to all cases. Careful consideration of 

 the factors of climate and soil and their relation to the vegetation of 

 forest and open land will, however, be the best safeguard against false 

 conclusions. The climatic, orographic and soil conditions on the two 

 parallel watersheds in the Rocky Mountains at Wagon Wheel Gap are 

 very different, the actual comparison of a lightly forested and denuded 

 watershed has barely begun. We must wait for the facts to develop 

 before any conclusion can be drawn. We may state here merely the 

 deductions based upon the observations of the Swiss watersheds which 

 seem to be of a general application and wait for our own results to 

 confirm or modify them before we safely adopt them for our guidance. 



One general fact that has l^een confirmed by the Emmental experi- 

 ment is that a good forest cover has a very beneficial effect upon the 

 regime of streams in mountainous and hilly regions. To be sure, it is 

 not always possible for the forest cover to prevent floods. Under cer- 

 tam conditions, the forest soil may lose its retaining power entirely, as 

 for instance, during heavy general rains, so that as much water will 

 run off from forest soil as from open land. But even in such cases 

 the favorable effect of a soil under forest cover on the run-off is still 

 clearly noticeable. Thus while landslides occurred almost regularly in 

 general rains on the steep meadow or pasture land or on bare land, the 

 damage from this source on the forested watershed was very rare. 



