i66 THE BOOK OF NATURE STUDY 



force were in the ascendant, the arch would not emerge at all ; 

 if the latter, the arch would emerge, but its upper layers would 

 be removed during emergence. In spite of this the position of the 

 axis of the arch would be the highest primitive ground on which 

 transverse drainage would be initiated (Fig. 54). 



These streams would flow across hard and soft strata alike, 

 and would cut into both (Fig. 50). But the rate of erosion would 

 be governed by the resistance afforded by the harder beds, as 

 any slackening of slope across the softer beds would diminish 

 the erosive power of that part of the stream crossing them. In 

 this way hard beds would begin to be trenched, and the walls of 

 this part of the valley would stand up steeply because of the 

 resistance of the hard bed to weathering. Where softer beds 

 were traversed, the valleys would widen out (Fig. 56) by the effect 

 of rain, frost, wind, and soil-making, and rills and eventually 

 streamlets would begin to flow into the transverse valleys at these 

 points (Fig. 57). Falling into the cross streams with considerable 

 velocity, these streams would cut back their head waters along 

 the strike of the softer beds, and thus lower the level of that 

 part of the country, that is, along the outcrop of the softer rock. 

 As the transverse streams deepened their valleys, the longitudinal 

 ones would also both deepen and lengthen theirs, removing soft 

 rock from both sides of the hard bands and giving them the wold 

 outline, with dip-slope and scarp. Further denudation by them 

 would clear off most of the soft material from the dip-slopes, and 

 undercut the hard bed in their other side, causing retreat of the 

 scarps and gradual movement of the stream beds in the same direc- 

 tion (Fig. 58). All this denuding work would be kept in activity 

 by the deepening of the cross-cut gorge of the transverse stream. 



Thus the relief of the wold, resulting from the resistance of the 

 hard bed when its softer neighbours have been in part removed 

 by denudation of strike streams, would be entirely dependent on 

 the deepening and keeping open of the breach through it by the 

 transverse streams into which longitudinal streams are draining 

 and carrying their denuded material (Fig. 59). 



If anything were to happen to the strike stream, such as 

 damming up or loss of erosive power, the fall of the longitudinal 

 streams into it would diminish, the velocity of its water would 



