ENDLICH ON EEOSION IN COLORADO. 859 



its widest part, is located in a steeply sloping, smooth wall of conglome- 

 rate. In a narrow crack, water slowly trickles down to the toi) of the 

 opening. Although, in the course of ages, even this slightly wearing 

 movement can produce a visible effect upon the rock, it is not — as a 

 movement — the cause which led to the formation of this wide opening. 

 At that point the conglomerate varies considerably in composition. In 

 the immediate vicinity of the cave, it is softer, contains fewer large 

 bowlders, and these are but loosely cemented by clay and feldspathic ma- 

 terial. Saturation of this rock and subsequent expansion of the water 

 upon freezing cause portions of the roof and walls to "scale off". 

 After some of the cementing material has been removed, the bowlders, 

 no longer held in place, drop out, thus gradually enlarging the excava- 

 tion. As soon as such portions of the rock are reached that are suflQ- 

 ciently cohesive to resist this action, the growth of the cave will come 

 to an end. Ample evidence was found at that point to demonstrate that 

 this was really the method of formation. Masses of debris on the floor 

 of the cave and dangerously loose bowlders overhead corroborated 

 other evidence. Indians and wild animals have not uufrequently uti- 

 lized these and other caves as places of shelter. Eemnants of charcoal 

 indicate the places where at one time fires had been built. One of the 

 most striking examples where such caves or excavations produced by 

 fluviatile erosion have been utilized by man may be found in the 

 ruins of the old cliff-dwellings in Southwestern Colorado. Several 

 of the streams there have worn long and deep recesses into the 

 readily yielding sandstones and shales prevailing in that region. Into 

 these, single houses and entire settlements have been placed by the 

 shrewd aboriginal inhabitants. Although often removed a considerable 

 distance from water, the architects of those times preferred to take ad- 

 vantage of the places which nature had prepared for them. Both shel- 

 ter and protection from enemies were afforded them, and tbey adapted 

 their style of building to the places which they chose for the purpose. 

 In the various publications of the Survey, full accounts of these dwell- 

 ings will be found. 



ARCHES. 



Arches, or " natural bridges", as they are frequently termed, can be 

 formed wherever the rock containing them is sufficiently thin to be per- 

 forated by erosive action. We have here to consider mainly such arches 

 the genesis of which is directly referable to agents of erosion. Viewing 

 them from this standpoint, we may say that an arch is the most com- 

 plete form of a cave. If the material containing the latter should be 

 sufficiently thin to allow erosion to progress throughout its entire extent, 

 then we will have the former as the result. It is evident that definite 

 conditions, perhaps not often met with, must exist before we can 

 expect an arch to be completed. Necessarily such products will show 

 much variation in form and size, dependent upon the material through 

 Bull. iv. No. 4 8 



