122 



to feed the springs, and thus account for the observation, that mostly — but by no 

 means always — forest areas produce more constant flow of springs. But, first of all, 

 the forest floor is not like a sponge'; and secondly, a sponge would not act as these 

 writers conceive it. The experiment is easily made. To be sure, the sponge will 

 take up large amounts of water until it is fully saturated, but not a drop will it give 

 up below, unless it is pressed or more water is added. As soon as the water supply 

 stops it begins to evaporate at the top, and supplies itself by means of its capillary 

 forces with moisture from below. It would, therefore, be apt to drain off any 

 moisture offered to it from below, instead of giving up its own moisture. Fortunately 

 the forest floor has only to a slight degree this water capacity of a sponge, a moss- 

 cover coming probably the nearest to it, and, therefore, being the most objectionable. 



In so far as it does have the capillary forces of a sponge, it withholds the water 

 from the soil and from the springs, and yet this very withholding capacity, this very 

 loss of moisture to the soil, does produce a beneficial influence on the flow of springs 

 by checking again the most avaricious robber of moisture — evaporation. 



A forest litter, according to depth and kind, will allow from 60 to 75 per cent, 

 of the rain falling upon it to filter through to the soil, while the moss gives up 

 hardly more than 40 to 50 per cent. But the naked soil loses, by evaporation, its 

 water so quickly, that only 20 to 40 per cent, will find its way into the drainage. 



This protection of the water supply is equal to that of a mulching or to the 

 cultivation in the field, namely, simply as a check aga,inst loss of water by evapora- 

 tion. Besides, there is in this cover a protection of the granular structure of the 

 soil. Such granular structure makes the subterranean drainage easier than on a 

 naked soil, which is beaten together and compacted by the raindrops, and at last 

 makes the water run off' superficially. The same effect, of breaking the force of the 

 raindrops, is due to the intei'ception of the rainfall by the foliage, which,, to be sure, 

 withholds from 10 to 15 per cent, from the soil altogether, and thus this amount is 

 lost to the soil. 



Another loss of water for the springs and runoff is represented in the amount 

 that is utilized by the trees in growing, if, indeed, this may be called a loss; for it 

 produces useful material, and the amount of water which is returned to the atmos- 

 phere in the transpiration from the leaves aids in producing those very conditions 

 above and in the forest which may make it a climatic factor for its surroundings, 

 namely, an area of cooler and at the same time moister air strata, with a tendency 

 to condensation. 



It is interesting here to state that the amount of water needed by the plant is 

 somewhat in the proportion offered to it: that is to say, in dry seasons and in dry 

 regions the same species of tree will transpire proportionately less water than if the 

 supply is ample. Conifers require from six to ten times less water than deciduous trees. 



I have made a calculation upon the basis of some rough figures which give the 

 amount of transpiration by various trees, and find that a coniferous forest in the west, 

 taking the rainfall at 2J inches, or 4,500,000 ppunds of water to the acre, would 

 hardly transpire more than 330,000 pounds, or 8 per cent, of the total precipitation 

 during the period of vegetation. So that the loss to the soil waters from this 

 source would not appear as great as some are inclined to believe. 



Lastly, beyond any dispute and easily understood is the influence of the forest 

 on the runoff, and especially on the superficial runoff — from mountain slopes. 

 Here again, the simple mechanical obstruction which the inequalities of the forest 

 floor and the trunks offer is the principal factor of influence. But also by distribut- 

 ing the water differently into subterranean and superficial runoff from what it would 

 have been on a naked slope, an influence upon the system of springs and brooks is 

 exerted. Yet when it comes to extremes of rainfall, even the best kept forest and 

 forest floor— and we know but few such in the United States — would not be able to 

 prevent the floods. In such cases the topography is decisive, and on the steep slopes 

 of a West "Virginia river-bed, for instance, no forest cover is capable of retarding the 

 runoff of a heavy rain and the freshet that is bound to come. "When it comes to 

 discuss the influence upon the water conditions of a large stream with many affluents, 



