386 W. A. E. TJSSHEE, ON THE TRIASSIC 



rally in the southern part of the area sandstone seems to be the 

 upper variety, faulted out at Exmouth*, but appearing immediately 

 north of it, and probably occupying the bed of the Exe as far as 

 Topsham, where it is developed. Between Topsham and Silverton 

 it covers a considerable area with an average breadth of about two 

 miles. It probably conceals other varieties of the division at Polti- 

 more. Between Heavitree and Honitons Clist (east of Exeter) these 

 sandstones appear to dip with the ground, and to contain traces of 

 clay. Owing to the sudden changes the members of this division 

 present, it is almost impossible to say whether the absence of Lower 

 Sandstones about Bradninch is due to replacement by breccia, or 

 whether their development north of Collumpton is due to the con- 

 cealment of lower beds of different character, or to change from 

 breccias. 



Due east of Burlescombe the whole division is either cencealed by 

 Lower Marls, faulted out, or represented by clay-beds. Traces of 

 the sandstones (Lower) occur at Eord Place and near Whipcoats. 

 West of Thorn St. Margaret the Lower Sandstones either conceal 

 or represent all the lower varieties of the lower division ; for their 

 junction with the Lower Marls is a faulted one, as also north and 

 east of Thorn St. Margaret, where they are let in by faults. Be- 

 tween Thorn St. Margaret and Wiveliscombe the visible part of the 

 lower division is chiefly represented by sandstones, faulted against 

 the Devonian at Horridge Down, south of Wiveliscombe (fault shown 

 in the railway- cutting). 



It is very likely that north of Burlescombe these sandstones repre- 

 sent the middle or lower beds of the division, and not the upper part, 

 as at Honiton's Clist and Broadclist. A large east-and-west fault from 

 Grant's Farm, north of Wiveliscombe, shifts the position of all the 

 divisions on the north side of it eastwards ; so that we find sand- 

 stones and breccias of the Lower Series faulted against Upper Sand- 

 stones, and Conglomerates and Upper Sandstones against Upper Marls. 

 From this fault to Washford (near Williton) the lower division is 

 well developed. It is much faulted north of Stogumber. In this 

 district sandstones generally occur intercalated with breccias ; but, 

 where separable, are found to underlie them. 



From a point between Bilbrook and Eoadwater, to Dunster, marls 

 generally conceal the sandstone, which is shown at Withycombe and 

 Dunster Park. I am inclined to consider these last belonging to 

 the Upper divisions, and the absence of the rest of the series due 

 both to numerous faults and overlap ; but this is only a supposition, 

 based on lithological affinities which are not always trustworthy. 



From Dunster to Minehead sandstones occur in the Lower division, 

 where it is thrown up by faults. Between Bratton and Selworthy, 

 west of Minehead, sandstones with intercalated beds of breccia, 

 which Mr. H. B. Woodward concurs with me in referring to the 

 Lower series, occur. South-west of Grabbist Hill a thin fringe of 

 sandstones and intercalated breccias occurs, probably overlapped 



* At Dawlish red sandstones overlie the breccia, and are apparently brought 

 down by successive step-like faults between Langstone Point and Dawlish. 



