300 E. A. Walford — Terraced Hill Slopes of Oxfordshire. 



spinatus). The marls are porous, they are non-cohesive. The lime 

 ill them has either been removed, leaving them in a loose, incom- 

 pacted state, or it is in process of removal by the constant saturation 

 and passage of water. Oft-times we forget that the heated land- 

 surface of summer-time must draw again from below that which it 

 allowed to pass through. 



Although it is common enough to see on hill slopes small terrace- 

 like ridges caused by the downslide of surface soil, yet it is not 

 until we reach the position of the marls of the Middle Lias 

 that the downslide is made prominent, and that terraces, from an 

 incipient stage like the ordinary grass ridge to minor and major 

 terraced banks of varying regularity of form, can be traced. On 

 these slippery slopes the soil must creep. The rain and rainwash 

 loosens the light soil and marl below and about the roots of the 

 herbage, and urges its movement downward. So, just as the 

 children make ground-slides down the steep smooth banks, so the 

 Earth-Mother makes these pleasant folds along the sides of her 

 green vales as she strips away her waste for the use of the lower 

 lands. 



Below these slippery marls are depths of compact blue clay 

 (the zone of Ammonites margaritatus if near the marls). A little 

 below the point where the marls and clays meet is the line of water 

 outflow — its course can easily be traced around every hill slope. 

 Along this line there is constant removal of the marls by chemical 

 and mechanical solution. The effect is the loosening and sliding of 

 the beds downwards and outwards. The jumbled state of the land 

 along the line is often evident ; at Edge Hill it is very prominent. 

 If there be " hummocky " ground it is thereabouts. The pressure of 

 the overlying mass of rock (there twenty- five feet thick) with the 

 associated marls has no doubt aided in the squeezing out of the soft 

 beds along the water-line. Hence pressure appears as a factor in 

 the making of terraces. The terraced slopes of the chalk hills and 

 the lynchets of the sandy oolitic hill-caps of Dorsetshire frequently 

 begin near the crest. The water-line of the former is probably the 

 Gault, and of the latter the line is at the top of the clays of the 

 Upper Lias. 



Furthermore, it appears that the free passage of water through the 

 rock and marl is necessary, for the Upper Lias Clays have on the 

 Oxfordshire terraced hills either been wholly stripped from the hill- 

 tops or have been pushed far back by atmospheric denudation. 

 Eegularly terraced slopes are not found under clay-caps ; the clay 

 cover may roof the hilltop, but the appearance of terraces is coin- 

 cident with its wearing away from the top of the bank. 



I have noted that the flats of the terraces, especially of the lower 

 ones, are often water-soddened, but whether the flats have relation- 

 ship to minor lines of outflow is not clear and is scarcely probable. 



The amphitheatre form of terraced land is always a valley head. 

 The outflow of the stream — the valley-maker — marks ordinarily the 

 base of the amphitheatre. More frequently the terraces of the valley 

 head are small iu step, their curvature is broken, and the depth of 



