302 PROF. A. H. COX AND ME. A. K. WELLS ON THE [vol. lxxvi y 



100 feet thick), and at the same time too regular over many miles, 

 for one to ascribe their origin to the invasion of an early basic 

 intrusion by a later acid intrusion. Further, the transition from 

 basic to acid material is too complete for their periods of formation 

 to have been discontinuous. As already stated, intrusion-breccias 

 are here the exception, the single case observed on Pared-y-Cefn Hir 

 being due to some local movements during consolidation (p. 287 J. 

 The basic rocks in these selvages, therefore, owe their 

 existence to differentiation in place within the mass of 

 acid material. A noteworthy fact is that the selvages at both the 

 upper and the lower contacts are of approximately equal thickness, 

 and both show the same rock-types, which proves that the differ- 

 entiation in place was not controlled by gravity in this 

 the largest intrusion of the district. In the same way the larger 

 of the two Tyn-y-llwyn intrusions shows along parts of the upper 

 contact a selvage of basic material that must have resulted from 

 differentiation in place. This absence of control by gravity has an 

 important bearing on the arguments put forward below. 



We have next to consider the evidence for deep-seated 

 differentiation. The intrusions may be likened to one-half of 

 a cedar-tree laccolite. The lower intrusions contain, with one 

 exception, a more basic material than those above, and in each 

 single intrusion the material at the bottom is more basic than that 

 at the top. The question then arises whether (1) the intrusions 

 are simply offshoots of a single large mass — a cedar-tree laccolite — 

 in which the basic rocks have resulted from differentiation in situ, 

 or whether (2) each intrusion was derived separately from a magma 

 which was in process of undergoing deep-seated differentiation. 

 The latter supposition seems to be the true one, as is apparent 

 from the following considerations. 



First, the tongues are of such length, and are separated by 

 such considerable masses of sedimentary rocks, that it is difficult to 

 postulate the convection -currents that would be required to collect 

 together the basic material, supposing the present differences to be 

 solely due to differentiation in place. In this connexion it should 

 be kept in mind that the present southward dip of the tongues is 

 entirely due to tilting subsequent to intrusion : originally the sills 

 must have been horizontal, or neaily so (p. 304), therefore it is not 

 a case of basic material having gravitated into downward-tending 

 offshoots. 



Secondly, the position of the highly basic Waen-Fechan in- 

 trusion, coming as it does between two acid masses (see fig. 4, 

 p. 289), would be quite anomalous, on the assumption that it resulted 

 from differentiation in place. 



"We can only suppose, therefore, that the basic rocks in the 

 minor intrusions do not owe their existence to differ- 

 entiation in place, but that they represent distinct 

 intrusions. Their similarity shows that they were all 

 derived from a common parent magma; but the fact 



