PE15IDIAN ROCKS OF PEMBROKESHIRE. 163 



opposing strike and become transgressive in the manner already in- 

 dicated. In my other paper (ante p. 147) I show that the evidence 

 points to their holding similar positions to one another also in parts 

 of Caernarvonshire. Neither of these formations, so far as they are 

 at present known, can be closely allied, petrologically, to the so- 

 called Lanrentian gneiss (Lewisian of Murchison) of the Western 

 Highlands and Hebrides, or even perhaps to that in Anglesey and 

 South-west Caernarvonshire ; hence it seems more than probable 

 that the latter belongs to a still older formation. They will, how- 

 ever, I believe, be found to have a wide distribution over the British 

 area when the metamorphic rocks of the various districts have been 

 carefully worked out ; and even now evidence is gradually accumu- 

 lating which points unmistakably to other areas than those above 

 mentioned, such as Charnwood, the Wrekin, &c* At present the 

 indications, therefore, are that we have three British pre-Cambrian 

 formations with rather marked petrological characters ; and they 

 probably occur in the following descending order : — 



Cambrian, 

 Sediments the result of denudation chiefly. 



I. Pebidian (Huronian? of Canada). 

 Deposits chiefly derived from subacrial and submarine volcanoes. 



II. Dimetian (Upper Laurentian ? of Canada). 

 Sediments chiefly the results of denudation, but partially volcanic. 



III. Lewisian (Lower Laurentian ? of Canada). 

 Deposits largely the results of denudation, but probably in part organic. 



It would undoubtedly be unsafe to lay down a rule that formations 

 may generally be recognized, and hence correlated, by petrological 

 characters only ; and yet there can be no doubt that most formations 

 or groups of rocks show some peculiarities dependent on age and 

 special physical changes at the time of deposition, which may fre- 

 quently act as guides to correlation over considerable areas. In 

 these pre-Cambrian rocks we have a persistent metamorphism, with 

 no local cause to account for it ; it is also somewhat peculiar to itself, 

 duo probably to great and repeated changes, from an alternation 

 frequently from deep depression with great superincumbent pressure, 

 to elevation. The above guides, therefore, if kept in view, may at 

 least aid in a recognition in other areas in the absence of the fossil 

 evidence which at present is wanting. 



* See descriptions of rocks in these districts by Allport, Quart. Journ. Geol. 

 Soc. vol. xxxiii. p. 449 ; Callaway, Quart. Journ. Geol. Soc. vol. xxxiii. p. 652 ; 

 and Hill and Bonney, Quart. Journ. Geol. Soc. vol. xxxiii. p. 754. 



