416 peteeson. PROBLEM OF GULLYING [Ch. 23 



struction of land, on the one hand, and constantly mounting diffi- 

 culties relating to silt disposal, on the other. 



MECHANICS OF GULLY CUTTING 



Gullies follow a simple pattern of development which is character- 

 istic in all localities. Almost invariably they are marked by vertical 

 or nearly vertical headcuts and banks which give them a typical 

 rectangular cross section. Once established, the headcuts advance, 

 and the channels widen by sapping and undercutting the banks. Abra- 

 sion by flowing water, either at the falls or above, is a relatively minor 

 cause of enlargement. The depth of cutting and the gradient of the 

 downstream channel vary widely, apparently dependent on a function 

 of the flow, the character of the eroding sediments, and the slope of 

 the valley floor. Depths ranging from a few inches to 60 feet have 

 been observed. 



Advancing in this manner, gullies may cut for an indefinite distance 

 without any disturbance of the surface, either laterally or upstream, be- 

 yond the confines of the channel (Fig. la) . Also, because cutting gen- 

 erally takes place at depths below the reach of plant roots, the presence 

 of even dense vegetation above headcuts has little if any influence on 

 the rate of advancement (Fig. 26) . 



The upstream progress of headcuts occurs at highly variable rates 

 which depend on the volume and duration of flow and on the character 

 of the eroding sediment. A headcut can remain inactive for a period 

 of years, then under proper conditions of flow it may progress hundreds 

 or even thousands of feet within a few hours or days. Table 1 shows 

 the measured progress of a number of headcuts during the past few 

 years. The relatively insignificant progress made by the gullies during 

 the past few years (which in most of the localities shown have been 

 abnormally dry) strongly suggests that most of our long gully systems 

 have been developed mainly during the few years that produced ex- 

 traordinary floods. Unfortunately no records on growth are available 

 for these years, but how else can one account for the San Simon mov- 

 ing only a few hundred feet in 6 years when it has cut nearly 70 miles 

 in the 44 years since its beginning in 1905? The condition is one that 

 should be considered in making an estimate, on the basis of known 

 records, of long-term silt carried by any particular stream, for, unless 

 the period of record contains its quota of these wet seasons, the esti- 

 mate can be highly erroneous. 



