312 PBOF. A. H. COX AND ME. A. K. WELLS ON" THE [vol. lxxvi, 



fault-hollow is seen to be aligned with the deep and straight 

 rift-like Ganllwyd Valley (valley of the middle Mawddach) which 

 debouches on the opposite side of the estuary valley. This rift-like 

 valley evidently represents the north-eastward prolongation of the 

 Derwas Fault. But, while the fault south of the estuary extends 

 in a direction north-east and east-north-east, the fault north of 

 the estuary runs more nearly north-north-east. The fault there- 

 fore swings round in the same manner as, but more rapidly than, 

 the strike of the sedimentary rocks as the} r follow the trend of the 

 Harlech Dome. The fault is thus a true strike-fault, completed, 

 so far as concerns its eastern portion, during the movements that 

 gave rise to the main dome, but initiated, at least along its 

 western portion, during a much earlier period, before the igneous 

 rocks (of Bala or pre-Bala age) were intruded. 



(ii) The Dolgelley Fault. — This fault runs parallel to, and 

 about half-a-mile south of, the Derwas Fault. Its course deter- 

 mines the direction of a deep valley which (south of the Wnion 

 Valley) is bigger and more persistent than the hollow caused by 

 the Derwas Fault. Like the latter, it acts as a normal fault, 

 throwing down to the south-east. But, unlike the Derwas Fault, 

 it evidently cuts and displaces the intrusions. Mr. Lake & Prof. 

 Reynolds 1 considered, however, that along this fault also the main 

 movement took place prior to the intrusion, and that movement 

 was subsequently reversed. We shall produce below (p. 313) 

 further evidence bearing on this point. 



(iii) The Mynydd-y-Gader North Fault. — One other 

 fault which appears to be earlier than the intrusions is that along 

 the northern front of Mynycld-y-Gader. It cuts out the Basement 

 Series and also more and more of the Upper Cambrian rocks as it is 

 followed eastwards, until eventually some 1500 feet of rock are 

 missing. Along much of its course it is concealed by diabase 

 material, often perhaps only in the form of screes, and it is not 

 possible to state in this case whether or no the fault affects the 

 intrusion itself. As with the Derwas and Dolgelley Faults, the 

 downthrow is to the south. 



(iv) The Ceunant Fault. — The remaining fault — the Ceu- 

 nant Fault 3 — is in a quite different category. It acts as a 

 repeating fault, throwing up the rocks on its south side, and in this 

 way differs from the faults so far mentioned, which act as normal 

 faults throwing down to the south. The extent of the throw is 

 apparently 3000 or 4000 feet in places. This semblance of a large 

 vertical throAv is, however, probably due to a lateral shift, rather 

 than to a true vertical throw, just as in the case of the Llyfnant 

 Fault, 20 miles to the south, described by Prof. 0. T. Jones & 

 Mr. W. J. Pugh.3 



1 Q. J. G-. S. vol. Iii (1896; p. 521. 2 Ibid. p. 512. 



3 Ibid. vol. lxxi (1915) p. 373. 



