METHOD OF STUDY 265 



river system. By tracing the successive changes of these small tempo- 

 rary channels we may acquire much information as to what goes on in 

 the slower but otherwise approximately similar process which takes place 

 as the larger streams work downward into the rocks they traverse. 



Conditions of small Streams 



experiments with thin la yers of wa ter 



First, let us observe the conditions of the thin layer of water as it moves 

 over a slightly tilted uniform surface such as a plate of glass. As the water 

 is impelled by gravity to flow it shows an interesting tendency to gather 

 into streams rather than to move as a sheet. Even the slight irregular- 

 ities of the nearly perfect plane tends to create definite streamings. If 

 now a thin coating of any fine somewhat adhesive material be placed on 

 the plate, for instance, such as close grained clay, we note at once a dis- 

 tinct tendency to channeling — a drainage is in fact at once organized, 

 the troughs of which will be rapidly deepened until the conditions of 

 the rain-washed fields are essentially reproduced. It is thus made evi- 

 dent that the circumstances of movement of the thin layer of water which 

 the rain brings to the sloping surfaces of the earth are such that very 

 slight irregularities inevitably bring about the formation of distinct 

 streams. If the earth were in a state to feel the effect of these tiu}^ cur- 

 rents, it would, save in the very arid districts, be carved by channels so 

 small that several would be traceable in each square foot. That such is 

 not the case is due to the fact that almost everywhere the coating of vege- 

 tation is sufficiently dense to protect the soil from all except the consid- 

 erable rivulets. It is only where a very great number of small temporary 

 streams have combined to form a torrent that the energy of the moving 

 water is sufficient to cut through the matted vegetation and to attack the 

 soil. 



INFLUENCE OF VEGETATION 



While the existence of a vegetable coating in the natural field and its 

 absence in the bared areas we are now considering is a noteworthy dif- 

 ference, one in certain ways affecting the value of the experiments, it does 

 not really invalidate them. The effect of the plants is to limit the cut- 

 ting action to much fewer streams — those, as above remarked, where the 

 aggregated waters are able to brush away the mantle of vegetation. It 

 thus comes about that there is commonly a broad field between the head- 

 waters of adjacent rivers, which is protected from mechanical erosion by 

 the fact that the lesser brooks can not attack the subjacent earth. In 

 this way the formation of tablelands is favored, the destruction of the 



