460 W. A. E. USSHER ON THE CHRONOLOGICAL VALUE OF THE 



(b) These fringing sandstones, being prolonged either way, are 

 found to be the upper portions of the subdivision extending from 

 Sidmouth and Otter ton Point to Williton, and on the other hand to 

 pass downwards into the Wembdon breccia in the Bridgewater 

 district. 



(c) Prom the fact that no beds of sandstone, not distinctly refer- 

 able to the Upper Keuper beds, have been hitherto exposed in 

 sections in Taunton Yale except those fringing the Quantocks. 



(d) It is difficult to say whether the Bridgewater sandstones are 

 the base of the Trias in that district : if not, there is no possibility 

 of correlating the beds subjacent with the regular sequence of de- 

 posits in the West-Somerset and South-Devon areas. 



Third proposition. — That a subsidence progressing from south to 

 north led to earlier deposition in South Devon and to a consequent 

 attenuation of the Lower beds towards Watchet and Porlock, be- 

 tween which places their continuity cannot be shown. 



Fourth proposition. — As a consequence of the preceding the 

 lowermost beds of the south-coast Trias far exceed their more 

 northerly equivalents in thickness, and afford a strong probability 

 that a reconstruction of the English-Channel valley would exhibit 

 a still greater development of beds, dating perhaps as far back as 

 late Permian times. 



Proof. — (a) The gradual attenuation of the area occupied by 

 beds below the Upper (Lower Keuper) Sandstones proceeding north- 

 wards from Tiverton junction. 



(b) The less perfect representation of the bottom beds as we 

 recede from the south coast, their comparative homogeneity in 

 many places, and the absence of any strata in the subdivision, even 

 in the Crediton and Tiverton valleys, equivalent in thickness to the 

 breccias and breccio-conglomerates between Dawlish and Babba- 

 combe. 



(c) The more strictly local origin of the breccias (to judge from 

 their contained fragments) in the northern and central parts of the 

 West-Somerset and Devon area. 



(d) Between Watchet, Williton, and Porlock, the probability of 

 large faults bringing together the Upper and Lower Marls ; the high 

 dip of certain sands locally flanking the Palseozoic rocks at their 

 junction with the ever-present marls ; the presence of beds at 

 Alcombe and Minehead which might be referred to the conglomerate 

 subdivision on lithological grounds and by position as far as ascer- 

 tainable ; the occurrence of beds bearing strong lithological affinities 

 to the Lowest Triassic division, at Minehead and in the valleys be- 

 tween Dunster and Porlock, notably at Luckham; rendering the 

 identification of beds subjacent to the conglomerate division very 

 uncertain, and affording strong probability that the Keuper is alone 

 represented, this area having been dry land during the deposition of 

 the lower beds ; and, owing to local causes, the basement beds of the 

 Keuper in this locality betray a strong similarity to the basement 

 breccias (presumably of Bunter age). 



(e) The thickness of the Lower beds of the Trias (i. e. the series 



