1893.] NATURAL SCIENCES OF PHILADELPHIA. 335 



have carved out a channel in this flood plain nearly twice as wide 

 as the natural channel. The greater spring floods extend even 

 ahove this channel and cover the neighboring flood plain. Thus 

 the valley of the Colorado is terraced, the first terrace being about 

 thirty-five feet above the bed of the low-water channel, and the sec- 

 ond, or flood plain proper, fifteen feet or thereabouts above the first. 

 The first or high water terrace is sometimes on both sides of the 

 river, though generally on one side better developed than on the 

 other. The second or flood plain terrace is usually well developed 

 on both sides. This form of terrace is not without its significance 

 in connection with some of the river terraces of the glacial belt. 25 



The Colorado which, as suggested above, is a young stream prob- 

 ably because of the rejuvenating influence of the Tertiary uplift, 

 shows all the signs of youth. In the hard Silurian not only is the 

 valley a canon, but the channel is marked by numerous rapids. On 

 the softer Carboniferous the canon is disappearing, particularly 

 where the retarding influence of the hard Silurian barriers is felt. 

 The tributaries, and the river itself above the temporarily base- 

 levelled parts, are rapidly at work, and deep and narrow canon val- 

 leys with rapid slope are the result. 



(u) Loss of Drainage Territory. — The influence of the Silurian 

 barrier has made itself felt in another respect. As has been stated, 

 the eastern border of the denuded area is an escarpment of Creta- 

 ceous of irregular form — a region of buttes and mesas. This 

 escarpment is a part of the valley of the Colorado and the minor 

 details of sculpturing are the result of the erosive work of the tribu- 

 taries to the Colorado. Only a few miles from the main river (six 

 to ten miles) there is a flat-topped divide separating the waters of 

 the Brazos from those of the Colorado. The former has tributaries 

 here that have to flow seventy-five miles before emptying into the 

 main river. In other words, the Brazos has pushed its conquest, by 

 headwater erosion, into the territory of the Colorado and has placed 

 the divide only a few miles from the latter. 



The reason for this seems quite certainly to be the influence of 

 the Silurian bai'rier. This explanation will be found in more detail 

 elsewhere. 26 It is, briefly, that the two rivers, the Brazos and the 

 Colorado, are superimposed through the Cretaceous on the Palaeo- 

 zoic, but the latter, encountering the hard Silurian, has had its down 

 cutting retarded, whereas the Brazos has had only the much softer 



25 Tarr, Am. J. Sci., XLIV, 1892, pp. 59-61. 



