BY HUGH MILLEE. 311 



contradistinction to the transverse streams/^ Let a stream, A, be 

 supposed to find its way along a homogeneous shale, in which 

 deepening proceeds straight downward. A channel, V-shaped 

 from the action of the elements upon the yielding sides, will 

 soon result. It is shewn by dotted lines. On reaching the 

 underlying sandstone the stream flows along a junction of softer 

 with harder beds. The sandstone resists while the shale yields. 

 Deepening proceeds obliquely, and by the shifting of the apex 

 of the Y sideways the section becomes unsymmetrical. The 

 process is now an undermining one. Overlying beds lose their 

 support and give way, and the scar, always travelling to the 

 left, a a' a", becomes in time crowned by a sandstone. 



If the stream flow, in the first place, along a sandstone, as at 

 B, the result is ultimately the same. The sandstone is first 

 trenched through into a gorge with upright walls, (hard rock 

 retaining its steepness,) and the stream on reaching shale pro- 

 ceeds as before. The shales thus tend to become in the long 

 run the conducting lines, and with the co-operation of atmos- 

 pheric waste, which attacks, among the rest, any outlying masses 

 such as ', the long-drawn hollows are flanked by sandstone 

 ridges, and a ^^ rig-and-fur^^ of escarpments is produced. I^or 

 must this be supposed mere theory. "Wherever streams run 

 along the strike of variously-bedded rocks of tolerably regular 

 dip, their channels tend to consist of a scarp and a dipslope. 



Further steps of development commensurate with the continued 

 deepening of the adjoining valley remain to be traced. Four of 

 our streamlets. A, C, D, E, are of different sizes, and must carry 

 on their work of uncovering dipslopes and making scarps at dif- 

 ferent rates. An advanced stage in the process of escarpment- 

 making is shewn by the inner shading in the diagram. A, being 

 the most powerful, has worked so rapidly as to overtake C, which 

 it has obliterated ; converting its at first distinct channel into a 

 mere rain-washed ledge or terrace on its own bank. The stream 

 D was feebler than C, but its position favouring it more, both its 

 channel and escarpment remain as yet distinct. E, little more 



* Tlicsc terms, if I remember right, are Jukes'. Thoy iiro now tcrmb of common ac- 

 ceptance. 



