302 II. E. Gregory— Note on the Shape of Pebbles. 



blocks on Carrizo Mountain, from which it was derived. It 

 would appear that such pebbles and bowlders had made their 

 journey without the companionship of fragments of similar 

 size. The slowness with which pebbles are rounded where 

 they constitute a minor portion of the transported materials is 

 a matter of everyday observation in the Colorado Plateau 

 Province. In humid regions characterized by continuous 

 stream flow, and where the sediment is largely supplied by 

 surface run-off, the finer materials are strained out by running 

 water. In regions where dry and wet seasons alternate and 

 where wind is effective, the proportionate amount of finer stuff 

 is greatly increased. Dunes strewn along valley bottoms, and 

 formed of material carried from inter-stream areas, are abun- 

 dant in arid regions. Following sudden showers, resulting in 

 concentrated run-off, this material is swept into streams in 

 enormous quantities. In 1914 Black Falls, 10 feet high, on 

 the Little Colorado River, became obliterated, between two 

 periods of stream flow, by a deposit of 36,000 cubic yards of 

 wind-blown sand. So much sand is supplied along the middle 

 portions of the Little Colorado that pebbles and bowlders, 

 though large in number, have little chance to be abraded on 

 rock bottoms or by grinding against each other. Such blocks 

 may exist without change throughout an entire physiographic 

 cycle, for during the time of the stream's greatest erosive power 

 they may remain embedded in finer materials. Alternate wet 

 and dry seasons would appear to favor subangularity of pebbles 

 regardless of distance. 



In regions characterized by intermittent stream flow it may 

 be observed that the abrasive power of sand streaming 

 past and over bowlders on a river bed is an important factor in 

 determining form. For stretches of hundreds of feet in cer- 

 tain streams on the Navajo Reservation the upstream sides of 

 bowlders are polished and worn and rounded, even faceted and 

 etched, while the downstream sides are essentially unmodified. 

 In places working up- valley across huge bowlders involves 

 climbing precipitous faces and descending slopes, like travel 

 over miniature cuestas. To obtain a quantitative estimate of 

 erosion by this process, holes one inch deep were drilled in 

 opposite faces of sandstone bowlders at eight localities. On 

 visiting two of these localities a year later (in 1910) it was 

 observed that in the case of three bowlders abrasion had 

 reduced the depth of the upstream holes '03, *06 and *06 

 inches respectively. Five localities, including the two men- 

 tioned, were re-visited in 1911, at which time the deepest hole 

 was 0*4 inch and at one locality no trace of the drilling remained. 

 A hammer scar, made by clipping sandstone rock, exhibited 

 fairly well-rounded edges after the passage of a single heavy 

 flood. On the downstream side of the same bowlder, a blue 



