Yol. 56.] SILURIAN SEQUENCE OF RHAYADER. 115 



difficulty to the edge of the last division and over it, and in the 

 same manner as the basement-bed will eventually be found reposing 

 on the Dyffryn Flags at Dol-Ifor Farm. Unfortunately little rock is 

 exposed, but the relief of the ground so persistently bears out the 

 suggestion of overlap that a survey of the area renders the true 

 condition of aiFairs absolutely conclusive. 



(3) Faulting and overlap in combination. — This maybe 

 dismissed ; for, if once overlap be conceded, it is unnecessary to 

 complicate the problem by the addition of thrust-faulting. 



What, then, must have been the physical conditions at the time of 

 deposition of the Rhayader and Caban Groups necessary to bring 

 about these remarkable phenomena ? In the first place, it is evident 

 that this Caban Group was deposited subsequent to the Gwastaden 

 Group, and before the Rhayader Pale Shales; for we have seen that 

 it is overlain by the Rhayader Group, and underlain by the Dyffryn 

 Flags of the Gwastaden Group. Consequently it occupies a position 

 somewhere between the Dyffryn Flags and the Rhayader Group. 

 It cannot rest between the Dyffryn Flags and the uppermost beds 

 of the Gwastaden Group, for I have proved that there is no such 

 break in the succession of the higher Gwastaden sediments as 

 might represent the time of the deposition of the Caban rocks. The 

 succession is consistent, and has been checked from point to point, 

 in order to show the conformable relationships of the various divisions 

 throughout. The only sign of irregularity is between the Gigrin 

 Mudstones and the Rhayader Pale Shales. ]Nor can this Caban 

 Group be a representative of any one of the Gwastaden divisions, for 

 its graptolites are not comparable with those of the Wye Valley. 

 The Gafallt Grits, it is true, yield several of the species common to 

 the uppermost beds of the Gwastaden Group, but with many new 

 forms in addition. One would suppose, therefore, from palseonto- 

 logical evidence alone, that in the Caban area a set of beds younger 

 than those of the Gwastaden Group occurs. The Caban rocks, being- 

 newer than those of the Gwastaden Series, and being overlain by 

 the Rhayader Pale Shales, must occupy, then, a position between 

 these two groups. 



Having now shown that the Caban Group represents the great 

 break between the Gwastaden and the Rhayader Groups, and 

 bearing in mind the overlaps that occur within these upper 

 groups, it becomes an easy matter to trace the ancient history of 

 the rocks. The simplest method, perhaps, of showing the relation- 

 ship of the various groups is by means of a longitudinal section 

 taken along the base of the Rhayader Pale Shales, restoring the 

 ancient land- surface to its original form by eliminating all faults 

 crossed by the section (fig. 17, p. 116). At the western end of the 

 district the Pale Shales rest immediately on the zone of Mono- 

 yraptus tenuis, and proceeding eastward a greater thickness of 

 the Gwastaden Group becomes visible, until, at the Tannery, its 



i2 



