262 PROF. C. LLOYD MORGAX OX THE TEEIDIAN 





I. 



II. 



III. 



Ignition . . 



1-G9 



1-31 



1-38 



SiO. 



. . 72-30 



74-09 



75-48 



Aip, .... 



. . 12-09 



14-98 



9-60 



FeO J •• 



2-1(3 



1-80 



J 1-73 

 t 4-26 



CaO 



. . trace. 



trace. 



•98 



MgO 



. . 2-02 



1-50 



1-32 



K.O 



4-80 



1-27 



•56 



Na^O 



. . 3-41 



6-07 



5^09 



FeS, 



•90 







99-37 101-02 100-40 



Allusion lias already been made to the very vesicular rock at Carn- 

 ar-wig. It is remarkably lavaform, but is more probably part of 

 a dyke. Two strong beds of soda-trachyte or highly silicated 

 andesite interstratified with the Pebidian may be lava, but perhaps 

 it is best to regard them as sills. Still I should not be surprised if 

 true lavas of similar character * to the dyke-material were to be 

 proved at this spot. In any case I wish to suggest the possibility 

 that the acidic dykes, such as those at Pen-y-foel, Carn-ar-wig, and 

 others at Ogofau-Dduon, Pen-maen-melyn, Perth Taflod, Ogof 

 Cadno, to the north-west of Porthlisky, and at Trefeithan, may be 

 volcanic dykes intruded during the last phase of Pebidian volcanic 

 activity. How far this view (merely advanced as a suggestion) 

 would apply to the much larger dykes near the St. David's schools 

 and in St. Non's Bay I am not at present jjrepared to say. There 

 is no doubt, however, that all these dykes have the granophyric or 

 spherulitic structure in common. 



g. The Porcellanites. — That these beds have undergone alteration 

 since their original deposition there can be no doubt. But whether 

 this is due to the influence of the quartz -porphyry dykes, as 

 suggested by Dr. Geikie, is, I think, open to question. They are 

 seen in Caerbwdy valley, in St. Non's Bay, and to a less extent 

 at Ogofau-Dduon ; while a porcellanitic bed is found in the Allan 

 valley near the boundary-wall beyond the vicarage. In Non's Bay 

 the weathered surface gives indication of the originally very line 

 sedimentary nature of at any rate some of the bauds. Sometimes 

 the beds are porcellanized in fine bands interleaved with softer 

 scarcely altered material. In the Allan valley it is a tuff-like bed 

 with largish included fragments which has been porcellanized. 



Dr. Geikie has drawn attention to the concretionary masses (the 

 so-called adinole) in St. IN'on's Bay : but Dr. Hicks regards them as 

 included fragments. " That they are true fragments," he says 

 (p. 548), " of some pre-existing rocks, and not concretions, is clear 

 beyond dispute from microscopical and chemical evidence." I am 

 surprised, however, that any one studying them on the spot in 

 St. Non's Bay could doubt their concretionary character. Both the 



* The undoiibtetl lavas seem to be somewhat more basic in composition. 



