BEGINNING OF THE GORGE 39 



island, and the recession of that one to a point above the other doubtless 

 caused the abandoning of this, the shorter gorge. The bench or island 

 did not merely divide the streams, the falls, and the gorges, but it, too, 

 gradually uncovered and extended as the gorges were making. For ex- 

 ample, on the east side of the island there is the remnant of the wall of 

 a cataract and gorge. The crest of this cataract is level with the top of 

 the island (figure 2, plate 2), and shows that the stream flowed across 

 the north part of the bench, descended by a short rapids 10 feet, plunged 

 then 25 feet, and then fell about 5 feet more in the next 50 yards. At 

 that point the terrace is cut entirely off by the steep cliff made by lateral 

 erosion by the Mississippi. Along the cliff a person can see the nearly 

 longitudinal section of the entire rapids and fall. The front of this fall 

 extends in straight line northeast to southwest and the stream ran south- 

 easterly. Above the fall the river at that stage covered the top of the 

 limestone bench, there being notably no deepened channel for the stream. 



The east fall just described must represent a stage when Soldier gorge 

 was half made, since the latter extends so much in front of the other. 

 At the southwest corner of the island there is similar evidence of a cata- 

 ract which relates to the terrace at that point, though it is not so clearly 

 marked off as that of the east fall. It may, however, be mentioned as the 

 west fall, contemporaneous with the east fall, indicating the stage to 

 which Soldier gorge had then cut. The approach of the stream to this 

 west fall is over the top of the limestone bench which later became the 

 greater part of the island. 



The island appears, in short, to have been produced by the dividing of 

 the falls, and not the dividing of the falls due to an already existing 

 island. The island at first was small and located at the brink of the 

 falls. As the gorges on either side grew longer, the intervening bed, of 

 the stream, was abandoned, thus extending the island. Finally this be- 

 came the west bank of the river after Soldier gorge was abandoned. The 

 island may have once extended some 500 feet farther north than the 

 bench now does, that much of the cliff having since been cut away by the 

 j\Iississippi at its present stage. If the island was so extended, the point 

 at which the main fall and gorge intercepted the stream which was flow- 

 ing to Soldier gorge might be some 750 feet north of the last named. 



What I have called the east fall may in fact have been either a middle 

 fall or for a time part of the main east fall. In either case the main east 

 fall appears to have receded in a course which was well toward the present 

 left wall of the gorge of the Mississippi here. The Soldier gorge was the 

 shorter one and therefore had a more rapid descent to the junction than 

 the other had. Exact comparison is, however, not practicable. 



