VOLCANIC SERIES OF ST. DAVIDS. 263 



porcellaiiite bands and these adinole masses remind one of the 

 occurrence of Hint, or still more of chert in some of the older lime- 

 stones. That they are all products of some mode of the segregation 

 of silica is suggested by their appearance and mode of occurrence. 

 But the microcrystalline structure of the matrix is no doubt 

 puzzling on this view. It is probable that a fuller microscopical 

 study of these rocks in different stages of porcellauization would 

 yield results of considerable, and not merely local interest. 



4. The Eelation op Pebidian to Dim:etia:n'. 



a. Dr. Hichss Arvonian. — It is right that I should here state why 

 I omit any consideration of the relation of Pebidian to Arvonian. 

 The reason is that I have been unable to satisfy myself of the 

 existence of a separate and distinct Arvonian system in the area within 

 which I restricted my observations. I do not deny the existence of 

 such a system. But I saw no satisfactory evidence of it in tho 

 immediate neighbourhood of St. Davids. 



b. The Dimetian Question. — Between the City of St. Davids and 

 the Crow and Bishop Islands, off Porthlisky Point, there occurs a 

 granitoid rock, well exposed at Bryn-y-garn and elsewhere. The 

 question at issue concerning this rock is, whether, in relation to 

 Pebidian, it is a subsequent granite-intrusion or a pre-existent 

 Archsean formation. There are three kinds of evidence bearing on 

 the question : — (1) that derived from the lithological character of the 

 rock, (2) that based on the petrological and stratigraphical relation- 

 ship with the surrouuding rocks, and (3) that afforded by the presence 

 or a,bsence of Dimetian fragments or pebbles in overlying deposits. 



Concerning the first point, the lithological character of the rock, 

 I have no new facts to communicate. Had I been ignorant of the 

 controversy on the subject, I should unhesitatingly have pronounced 

 it a granite or an intrusive rock allied thereto. But since others of 

 far greater knowledge and experience than myself have held and 

 hold a different view, I do not presume to suppose that any opinion 

 of mine "will carry weight. 



On the petrological and stratigraphical evidence I may be allowed 

 to offer a few remarks. 



c. In Portlilislcij Baij. — In this locality there is no very satisfactory 

 evidence of either faulting or intrusion. As before noticed, however, 

 the strata in this bay have been much altered in some way, and if the 

 Dimetian be intrusive, one naturally surmises that it may have a 

 (possibly indirect) connexion with this alteration. It is therefore 

 worth noting that the Cambrian conglomerate and overlying beds 

 show a somewhat similar alteration in Perth Melyn, west of 

 St. Non's Bay, and here the Dimetian is close at hand. Still, though 

 this is naturally suggested if the Dimetian be otherwise proved to 

 be granite, I do not think that by itself it can be taken as satisfactory 

 evidence of intrusion. 



The fault-line, if it exist, is obscured. I found, however, un- 

 Q. J. G. S. No. 182. T 



