academy of sconces] POWELL'S SURVEY 105 



stream. . . . There is one case which is specially noteworthy on account of its relation to the principles of 

 sculpture. ... If in the course of time one of the [two parallel] streams encounters a peculiarly hard mass of 

 rock while the other does not . . . the unobstructed stream will outstrip it [the obstructed stream], will encroach 

 upon its valley, and will at last abstract it. . . . Thus by abstraction as well as by monoclinal shifting, streams 

 are eliminated from hard rocks (140, 141). 



THE SUBSEQUENT ORIGIN OF WATERPOCKET CANYON 



Yet although both tendency and process were thus explicitly treated in general terms, 

 they appear to have reference chiefly to the case of equal streams in horizontal structures, 

 especially in bad-land areas. Nevertheless the principles involved apply very closely to the 

 case of parallel streams of unequal size that flow square across an inclined series of hard and 

 soft strata, such as is exposed in the great Waterpocket flexure; for here the Dirty Devil head- 

 waters represent the obstructed stream and the Colorado the unobstructed stream in Gilbert's 

 supposed example; the divide between the two will be locally homogeneous in each belt of 

 rocks and therefore subject to competitive erosion and shifting; but the shifting of the divide 

 between the two streams will be most rapid and therefore most conspicuous in the weakest 

 belts. Hence the streams of the Dirty Devil group and the Colorado may be regarded as 

 constituting an actual case of competitive parallel streams, in so far as the Waterpocket flexure 

 is concerned; and the fact that they cross it in opposite directions, the first with and the 

 second against the dijp of the flexure, only gives the greater advantage to the latter. 



When this great flexure was upwarped with a maximum displacement of 7,000 feet, a host of 

 small consequent streams must have run eastward down its slope ; but the Colorado was perhaps 

 strong enough to run against it then as it does now. In due time a part of the divide between the 

 deeply incised west-flowing Colorado and the next adjoining small east-flowing consequent stream 

 on the north, much less deeply incised, must have lain across the belt of weak shales along which 

 the present Waterpocket Canyon is excavated ; and as the divide was there in homogeneous ma- 

 terial but much steeper on the southern or Colorado side, it must have been rapidly carried north- 

 ward far enough to ' ' abstract " or divert the headwaters of the small consequent. Thereupon the 

 divide between the first and second consequent streams, which had previously been essentially 

 symmetrical, would become strongly unsymmetrical, by reason of the steep descent of the 

 diverted first consequent in its southward subsequent course along the weak shale belt to the Colo- 

 rado ; hence this divide would in its turn be carried northward until the second consequent was 

 diverted; and so on, as far as the asymmetry of the divide continued. The Colorado would con- 

 tinually gain, and the streams of the Dirty Devil group would as continually lose. The shale 

 valley is longitudinal in relation to the trends of the neighboring ridges, monoclinal in relation to 

 the great flexure that determines it, and subsequent in relation to its origin as compared with 

 that of the transverse consequent valleys. Preparatory to every capture this valley would 

 slowly increase in length by headward erosion. At the time of each capture, the area of the shale- 

 valley drainage would be suddenly enlarged by a lassolike looping of its divide around the head- 

 waters of the captured consequent, and the volume of the capturing stream would be at the same 

 time suddenly increased. Thus the Waterpocket Canyon, as the actual shale valley of the 

 flexure is called, and its stream, Hoxie Creek, must have been developed in true subsequent 

 fashion. It may be added, however, that substantially the whole of their present length would 

 appear to have been gained in an earlier cycle of erosion, preceding the present cycle of deep 

 canyon incision; for Hoxie Creek must have been well established before its horseshoe detour 

 could have been cut; yet the process of headward erosion is presumably still continued, for the 

 waters of the next-to-be-captured consequent stream must now make a circuit of about 100 

 miles, around the northern and eastern side of the Henry Mountains by the Dirty Devil and 

 along their southern side by the Colorado, before reaching the mouth of the direct 30-mile sub- 

 sequent valley which Hoxie Creek follows. 



When the problem is thus envisaged with the aid of hindsight, it is difficult to understand why 

 Gilbert did not have the foresight to see through it. Its essential features, as far as they involve 

 matters of fact, were all well known to him, for they are beautifully illustrated on the relief map 

 of the region, Plate 1 of his report. Over a score of small consequent headwaters are shown to be 



