52 SECTIONAL ADDRESSES. 



acute form, of the great tensional stresses that preceded and accompanied 

 the extrusion of the plateau-lavas. 



The crowding together of the dykes in a region of central subsi- 

 dence can only be explained by assuming that such regions fractured more 

 readily and were possibly subjected to greater tensile stresses than the 

 country on either side. Further, this intense fracturing could only occur 

 if the tensional stresses were acting on a relatively thin crust and tangential 

 to the upper surface of a reservoir filled with liquid or plastic material. 

 We are, therefore, led to the conclusion that the dyke-swarms must be 

 due to a general rise of the primary basalt-magma, and that this rise was 

 accentuated in the regions of central subsidence. The magnitude of the 

 tensional stretch in these regions can be gauged by noting the aggregate 

 thickness of dykes in a given distance. In Mull it was found to be about 

 one mile in twenty-four, while in Arran the stretch is even greater and 

 amounts to as much as one mile in eight or ten. 



I have only touched upon a few features of the dominant and most 

 interesting types of Tertiary intrusions, but before leaving this part of the 

 subject I would like to refer to the mutual relations of some of these bodies 

 to each other and to the rocks into which they have been intruded. 



In unravelling the structure of an igneous complex two essentials con- 

 front us : firstly, the proof that any rock-body is a unit and not composite, 

 and secondly, its absolute or relative age. In the case of most intrusions 

 we apply the criteria of chilled margins and contact metamorphism, but 

 this necessitates a relatively low temperature for the rocks into which 

 the intrusion penetrated. With the ring-dyke complexes, however, the 

 intrusions, other than the dykes, sills and cone-sheets, are of considerable 

 magnitude and fairly deep-seated ; while as regards time they followed 

 closely upon each other. Thus, frequently, the cooling effect of an earlier 

 intrusion upon one of later date is not well marked. 



In Mull, and also in Ardnamurchan, it is not usual for the ring-dykes, 

 especially those of more basic composition, to show markedly the ordinary 

 effects of rapid chilling at their margins. Instead, there is usually some 

 slight assimilation of the country-rock and formation of narrow transitional 

 belts between contiguous intrusions. 



Extreme cases of the interaction of an intruded magma with an igneous 

 rock of related but different composition have been described by Dr. Harker, 

 first from Carrock Fell and later from Skye and elsewhere. To rocks due 

 to such a process he gave the now well-established name of hybrids, and 

 he called the process of interaction hybridization. 



When the rocks have interacted or hybridized at their contact with 

 each other it is not always easy to discover whether we are dealing with 

 a single intrusion locally differentiated, or with an intrusion of composite 

 nature ; for the results of partial differentiation and the effects of 

 hybridization have much in common. It is, however, in such cases that 

 what the officers of the Geological Survey have termed screens come to 

 our assis tance. 



A sc reen is a narrow mass of country-rock separating two neighbouring 

 steeply bounded intrusions, and older than either of them. Such masses 

 separati ng ring-dykes will have curved outcrops, and are the logical sequel 

 to the a rcuate Assuring of the walls of a centrally subsiding area. 



