Vol. 67.] THE CARBONIFEROUS SUCCESSION IN GOWER. 529 



of their relations to the phases (see Table Y, 1, p. 526, and 'Varia- 

 tions in Depth, etc.' pp. 532 et seqq.) leads to the further con- 

 clusions : — (ii) The periods of isostasy recorded in the phases have 

 followed either a movement of elevation (in the case of the phase at 

 the base of C 2 -r-S.) ; a continental period, during which depression 

 has not been exactly counterbalanced by sedimentation (in the 

 case of Km) ; or a shallowing, marine period (in the case of the 

 phase at the top of SJ. (iii) Each period of isostasy has been ended 

 by a sudden movement of depression. 



4. Taking into account, also, the fact that the lagoon-areas have 

 been coastal features, we may generalize the two last conclusions 

 (par. 3) thus, — lagoon-areas have been developed during periods of 

 steady depression, more or less prolonged, which have either 

 marked the first stage of the submergence of a foundering coast, or 

 followed intervals of upward movement interrupting a general 

 downward movement of a sea-floor. (The one condition is pro- 

 bably a particular case of the other.) These periods of steady 

 depression have preceded more rapid subsidences. 



It follows, as a corollary, that lagoon-phases may be expected to 

 accompany transgressions, as do Km and the phase at the base 

 of C 2 + Sj, but also that they will be subject to overlap by the 

 deeper-water deposits that have succeeded them. 



5. The phase at the base of C 2 + Sj, of which our knowledge is 

 most complete, persisted far longer in the northern than in the 

 southern part of the Bristol area : for in the north it is succeeded 

 by Sj beds, whereas at Weston it is confined to the base of C 2 . 

 The significance of this fact lies in the greater proximity of the 

 former region to the Avonian coast. In the western area (Pem- 

 brokeshire and Pendine) also, the phase apparently persisted longer 

 near the Avonian land than elsewhere (fig. 7, p. 527), although there 

 the matter is complicated by overlap of higher horizons over lower. 

 It appears, therefore, that the depression that closed the phase 

 affected the seaward margin of the lagoon- area before the more 

 landward parts. The depression was greatest in the south, both 

 in the Bristol area and in the west. 



No such landward lag is known in the case of the depression that 

 ended the lagoon-phase at the top of S 2 . 



The radiolarian phase. — The conclusions regarding earth- 

 movements, detailed above, afford a means of testing the hypothesis 

 that the radiolarian cherts at the base of P are similar in origin 

 to the Modiola phases on which those conclusions are based. 



The formation of the cherts was preceded in Gower by a shallow- 

 marine period (D l to D 2 _ 3 ). The following evidence suggests 

 that this period was closed by elevation. At Ifton, 1 near Newport 

 (Mon.), on the east, as well as at Haverfordwest 2 on the west, 

 Millstone Grit, similar lithologically to the band that succeeds the 



i Eep. Brit. Assoc. (Winnipeg) 1909, p. 478. 



- O. T. Jones, 'Summary of Progress for 1906' Mem. Geol. Surv. 1907, 

 p. 53. 



