130 KANSAS UNIVERSITY QUARTERLY. 



may have produced long, continuous ones. The natural seams in the 

 rocks in other cases probably have been followed by percolating 

 waters which would soon dissolve out the adjacent walls and make an 

 opening to the shales below. 



After an opening was once made only a short time would be 

 required to cut a deep channel into the soft substratum, and the 

 formation of a river would be commenced. Should the surface be 

 inclined from both sides toward the channel the area of the water- 

 shed would be greatly increased and the mechanical action of the 

 water intensified. The downward wearing would progress rapidly 

 until the ne.xt limestone stratum below was reached, provided the 

 inclination of the surface was considerable and the rainfall abundant. 

 But with this there would be a widening of the channel due to lateral 

 decay. The limestone mantle would protect the shale at the upper 

 surface, but the face of the wall of from 50 to 250 feet would gradu- 

 ally crumble away under the combine<l action of the different meteoric 

 agents until the undermining process would cause portions of the 

 limestone to fall. In this way the j^rocess of widening the valley 

 continues. 



Under such circumstances a river valley from one to five miles wide 

 per se by no means indicates a greater flow in the stream during past 

 time than at the present. For the widening process will not cease as 

 long as the bluffs remain higher than the valleys. Neither is the 

 amount of water carried by a stream an important factor, provided it 

 is sufficient to remove the relatively small amount of solids carried 

 into it by the lateral tributaries. After a stream has almost reached 

 its base level, and its valley considerably widened,, the process is 

 slow, the solids being removed principally in the form of solutions, 

 e.xcepting at times of freshets, so that a small stream can readily carry 

 them away. The valley of the Wakarusa near Lawrence is as wide as 

 that of the Kansas river which is ten times as long and carries more 

 than twenty times as much water, while different small tributaries 

 from the south not more than six or eight miles long have valleys 

 almost as wide. Still further, many wide valleys exist in Kansas, 

 valleys of erosion as truly as are river valleys, which have been 

 produced by general surface erosion almost independent of any 

 particular stream, the countless little rivulets being sufficient to carry 

 away the dissolved parts by the slow process of sipage, while the 

 solids are removed during heavy rains and freshets. 



Another interesting feature of the case is that every lateral tribu 

 tary greatly increases the widening process. Lateral fissures in the 

 mantle of limestone produce lateral tributaries to the principal 

 stream. Each of these cuts its channel to a level with the main 



