Vol. 67.] THE CAKBONIFEROUS SUCCESSION IN GOWER. 495 



discovery in them of Glyphioccras spirale (Phill.) 1 and Posido- 

 nomya becheri Bronn (p. 552). The importance of distinguishing 

 between the ' rottenstones ' and the overlying beds has been 

 recognized by the officers of the Geological Survey, who have 

 included the latter in the Millstone Grit, the ' rottenstones ' (as 

 previously mentioned) being ' Upper Limestone Shales.' 



Lithological sequence. — In descending order: — 

 2. Dark, well-laminated, non-calcareous shales, with some iron- 

 stone-nodules and thin sandstones ; in the lower part with 

 black cherts 2 ; according to De la Beche, 3 there is a thin 

 limestone (4 inches) in the middle. Very thick. 

 1. Radiolarian cherts 4 with interbedded shales, the cherts show- 

 ing the fine dark and light stripes, due to thin lamination, 

 frequent in radiolariau deposits, and the shales, also, well- 

 laminated and non-calcareous, but soft and in part weathering 

 to a chocolate colour. Under the microscope it is seen that 

 the cherts are exceedingly fine-grained, and that the dark and 

 light laminae differ one from the other in grain and in the 

 relative proportions of their organic and inorganic consti- 

 tuents, the latter chiefly the finest quartz-silt. Many of the 

 laminae are lenticular, some of them sharply so. Wavellite 

 occurs along joints. The total thickness is slight, the group 

 forming, as it were, basal beds to the shales above. 



Apparently, the series includes no recurrence of the fossili- 

 ferous limestones characteristic of the underlying Avonian. 

 Eauna. — 2. Largely unfossiliferous, but some beds are crowded 

 with thin-shelled goniatites 5 preserved as casts. 



1. Of the abundant radiolaria and subordinate sponge-spicules 

 in the cherts, the former correspond closely with Culm forms. 6 

 The shales are largely unfossiliferous, but some have yielded a 

 few lamellibranchs (p. 551) and indeterminate plant-remains. 

 Thickness. — Unknown, though probably at least 1500 feet, as 

 Glyphioceras spirale (Phill.) has been found in abundance at 

 apparently that distance above the base. Of this amount 

 probably much less than 50 feet belongs to the radiolarian-chert 

 group. 

 Limits. — The radiolarian cherts, doubtless, pass up into Group 2; 

 but the junction does not appear to be satisfactorily exposed, 

 and no repetition of similar cherts above is known. The upper 



1 By Mr. J. Pringle of the Geological Survey, Swansea Memoir, pp. 25 & 30 ; 

 the specimens, 'J. P. 1260, 1265-1276' in the Survey collection, were found 

 in a stream-bank on Barland Common, 150 yards west of Old Castle. 



2 Swansea Memoir, p. 25. 



3 Mem. Geoi. Surv. vol. i (1846) p. 133. 



4 Dr. Strahan has described in the Swansea Memoir, pp. 22-25, how the 

 discovery by Dr. G. J. Hinde of the radiolarian origin of these cherts resulted 

 from De la Beche's comparison of the beds with the cherts of Codden Hill. 

 He appends Dr. Hinde's description of the microscopic characters of the 

 rocks and their organisms. 



5 Swansea Memoir, pp. 25 & 30. 



e Dr. G. J. Hinde in the Swansea Memoir, p. 25. 



