Mat 10, 1918] 



SCIENCE 



451 



gorge with a rush of turgid mud-laden water 

 and even when the rain has passed there is 

 in the air the continual dull crash of the 

 caving banks. At places the arroyo fills all 

 the canon bottoms, at others it is a mere 

 crack in a wide expanse of alluviiun, but it is 

 continually encroaching on the bottom land. 

 The depth of the erosion varies greatly and is 

 controlled apparently by the distance of bed 

 rock from the alluvium surface (which is gov- 

 erned by the amount of alluvial filling that 

 had taken place) and by a fixed minimum 

 grade which is determined by the amount of 

 overloading of the stream and the grade of 

 bed rock. This minimum will become smaller 

 therefore as the alluvium is gradually removed 

 from the canon bottoms. The maximum 

 depth of erosion observed by the writer is 

 twenty-five feet, the average is probably about 

 ten feet. The former is reached in exceed- 

 ingly narrow canons such as the upper Cha- 

 quaqua Canon, and that of the Purgatoire in 

 southeastern Colorado and Yellowjacket and 

 Sandstone Canons in southwestern Colorado. 

 The arroyos are formed only along intermit- 

 tent streams. The canofls of McElmo Creek 

 and the Purgatoire Eiver seem to have been 

 dry at least part of the summer in the early 

 days (although they now flow water all tlie 

 year around) and for this reason they exhibit 

 the arroyo at the canon bottom. 



The steepness of the alluvial banks testifies 

 to the recent origin of these arroyos. What 

 caused them to apx>ear so suddenly? Rarely 

 is it that the processes of erosion are dis- 

 turbed yet it appears that the disturbance 

 which caused these arroyos has taken place 

 during the last sixty years. The settlers who 

 first entered these caiions found the bottom 

 lands low and rounded with no suggestion of 

 an arroyo at the center. The writer has 

 talked with pioneer ranchers both in south- 

 western and southeastern Colorado and on this 

 point they are vmanimous. The arroyos have 

 developed since their advent. To this may be 

 added this further physical evidence: 



1. Along the bottoms of Yellowjacket, Sand- 

 stone and Hovenweep Canons in southwestern 

 Colorado the arroyos are cutting into the 



ruins of Indian houses (stone) which are ex- 

 tremely old as they represent a civilization 

 much like that of the Zimi while the Utea 

 have occupied this region since the time of 

 historic record. The houses were built on 

 alluvial flats and it is only recently that the 

 streams have cut into them. 



2. Old roads and trails frequently cut 

 straight across gullies which it is now im- 

 possible to cross. (Southeastern and south- 

 western Colorado). 



3. Along the sides of the canon wall where 

 the alluvium has been completely removed 

 from the cliff sides the imprint of roots still 

 remains (Chaquaqua Canon — Southeastern 

 Colorado.) 



4. The fact that water is more abundant in 

 the canon bottoms now than previously seems 

 to have a bearing on this subject In the early 

 days (1860-1865 in eastern Colorado) (1870- 

 1880 in western Colorado) water appears to 

 have been very scarce in these caiions. This 

 would seem to be due to the water flowing 

 mider a heavy alluvium cover as the precipita- 

 tion records indicate no perceptible climatic 

 variation. The formation of these arroyos 

 seems to have uncovered a number of these 

 hidden flows of water. 



5. No alluvial terraces are found. The 

 canon floor is usually very nearly flat. If 

 these arroyos were cyclic, we should expect 

 to find a series of terraces representing a 

 series of stages in the erosion of this alluvium. 

 Such is not the case, even in comparatively 

 wide canons. The usual canon rock terraces 

 represent cycles exceedingly remote when 

 compared with the one imder discussion. 



Comparisons of the drawings and photo- 

 graplis of government reports of 1860 to 1870 

 with recent photographs confirm this hypothe- 

 sis, as the older reports do not seem to show 

 any arroyos like those now developed. 



The development of these arroyos seems to 

 have been, therefore, contemporaneous with 

 development of ranching. To what must we 

 ascribe them then? The \vriter believes they 

 are caused by cattle. Cattle influence erosion 

 in two ways: first by the wearing of trails; 

 second by the destruction of vegetation. 



