part 3] LOWER PALiEOZOIC of arthog— dolgelley. 303 



that the bulk-composition of each intrusion is different 

 from that of the others shows that the parent magma 

 was undergoing progressive deep-seated differentiation. 

 The differentiation followed the normal course, so that from an 

 ■essentially acid magma successive basic differentiates were drawn 

 •off, each less basic than those that preceded. In general, the later 

 or more acidic products were intruded into higher levels than 

 the earlier basic products, the Waen-Fechan intrusion being an 

 exception. 



We have still to consider the origin of the stratiform variation 

 within the limits of the individual intrusions. Here the variation 

 was evidently controlled by gravity, since the more basic rocks are 

 found in the lower parts of the intrusions. But it is not clear 

 whether each mass represents a single intrusion in which the varia- 

 tion is due solely to differentiation in place under the influence 

 of gravity, or whether each mass really represents a composite 

 intrusion formed while the progressive deep-seated differentiation 

 process was in active operation. The transition between the various 

 types in the field is so gradual, that it is difficult to see how multiple 

 intrusions can be present. On the other hand, the fact that gravity 

 exerted no control whatever over differentiation in place in the great 

 'Crogenen sill makes it difficult to understand why in neighbouring 

 intrusions gravity should exert a dominant effect. Thus, until the 

 rocks have received further study, we prefer to leave open the 

 question as to the origin of the stratiform variation within the limits 

 of each of the minor intrusions. This in no way vitiates our con- 

 clusion that the bulk variations in composition between 

 the minor intrusive masses must be due to a progres- 

 sive deep-seated differentiation, whereas the thin basic 

 selvages along the Crogenen sill must be due to differ- 

 entiation in place. 



The relative perfection in which the graphic intergrowths are 

 developed in the various rocks furnishes an interesting example for 

 study of the part played by a eutectic mixture. The more basic 

 rocks may not contain quartz at all, or only in small amount. So 

 soon as quartz appears, it is for the greater part in the form of 

 rnicrographic quartz-felspar intergrowths. In the basic rocks the 

 intergrowths are naturally small in' amount, interstitial in mode of 

 occurrence, and exceptionally fine-grained ; but, when they do occur, 

 they are exceedingly perfect. These interstitial patches represent 

 the last dregs of magma to consolidate, and these last dregs seem 

 to represent a eutectic mixture. With decrease in basicity the inter- 

 growth increases both in amount and in coarseness of grain, until 

 in the markfieldite type (of intermediate composition) the whole 

 rock is largely built up of granophyric quartz-felspar aggregates. 

 The intergrowths are here on a larger scale than in any other of 

 the granophyric rocks, and they are also very perfect (PI. XVIII, 



