186 ME. L. J. WILLS ON LATE GLACIAL AND [June I912, 



through a drained peaty flat ^ past Tyndwr to the Dee near the 

 Golf Links. Such a drainage-system could not have cut out this 

 valley, the breadth of which corresponds to that of the Dee. 



At present the river flows through a narrow gorge at Llangollen 

 some 250 feet deep, in which there is no Drift, and which is clearly 

 post-Glacial. As further proof of this, we find spreading out from 

 its mouth a series of gravel-terraces that are cut in the Drift of 

 the pre-Glacial valley below, and mark the various stages in the 

 erosion of the gorge. 



The similarity of the Llantisilio and Llangollen diversions 

 suggests that they were both formed in the same way ; that is, 

 probably by the obstruction of the drainage by inert ice in the 

 valley-bottom, aided to some extent by a dam of drift. 



(4) The Argoed Diversion. 



The next diversion is not a ' short cut,' but the reverse. Near the 

 large house at Eron-Cysyllte known as Argoed, the river swings 

 northwards to near the Trevor Aqueduct, and then bends south- 

 eastwards again at a sharp angle (see Map, Pi. XI). This loop is 

 clearly post-Glacial, for the Drift comes nearly, if not quite, down 

 to the level of the Dee beneath Argoed, and also near the railway 

 viaduct ; while between these points the river flows through a fairly 

 deep gorge eroded in the solid rock near Trevor. Thus, whereas the 

 Dee formerly ran nearly straight, the present course is, as it were, 

 along the other two sides of the triangle. 



The explanation of this, as of the following diversions, must be 

 sought for in the post-Glacial surface-configuration of the Drift, 

 which determined the direction of the flow of the wat-^r when the 

 retreat of the ice occurred. At this particular point all that can 

 be said is that, even now, the surface of the Drift rises to about 

 100 feet above its old course at Argoed, and presumably it always 

 sloped towards the north and thus deflected the waters in that 

 direction. 



(5) The Chirk Diversion. 



The diversion next to be dealt with is a phenomenon of much 

 greater magnitude, and involves the river's desertion of its former 

 valley for many miles ; in fact, its present course from the railway 

 viaduct near Cefn to below Overton Bridge has been excavated in 

 post-Glacial times (see Map, PI. XI). Although above the viaduct 

 we can see Boulder Clay beneath the present alluvium, nowhere 

 along this stretch except near Erbistock does the surface of the 

 solid rock reach as low as the Dee. The river has dissected the 

 coating of Drift, and in most cases has cut through 100 feet or more 

 of the solid rock below. This is enough to demonstrate tiie post- 

 Glacial nature of the gorge, unless the Vale of Llangollen has 

 been overdeepened by Glacial erosion. 



^ This represents an old kettle-hole lake, and was a morass as late as 1850. 



