666 MESSES. CLOTTGH, MATTFE, AND BAILEY ON [Nov. I909, 



occurrence, for the intrusion, when once it had made its way 

 along the fault, must have acted as a far more effective lubricant 

 than the flinty crush-rock which lay alongside of it. It is clear, in 

 fact, that the intrusion retained its fluidity until movement had 

 ceased, for it has crystallized without any marked appearance of 

 fluxion-structure. It can be shown, too, in this same Stob Mhic 

 Mhartuin section, that the flinty crush-rock became rigid, no doubt 

 owing to complete reconsolidation, while the intrusion was still 

 liquid : for, at one point, minute angular fragments of the flinty 

 crush-rock are found embedded, with trifling displacement, in the 

 margin of the fault-intrusion. 



There is no difficulty now in understanding why it is that the 

 fault-intrusion invariably shows a chilled margin to the fault- 

 plane, although in the external zone, away from the fault, its 

 margins are sometimes chilled and sometimes not. The cheeks of 

 the fault had in some cases been raised to a very high temperature 

 by friction, before the fault-intrusion had risen into position. But, 

 bearing in mind the small thickness of the fault-rock produced, 

 and that, at the most, we cannot claim more than incipient fusion 

 of even the extreme product (flinty crush-rock), it appears highly 

 probable that the quantity of heat mechanically produced was 

 quite inconsiderable. Once the intrusion had arrived alongside 

 the adherent film of flinty crush-rock, the latter, relieved now of 

 friction and cooled by conduction, would be free to consolidate. In 

 the next stage its temperature would drop to that of the intrusion, 

 which in turn began to solidify and crystallize out. Then both 

 rocks would cool together, yielding, as it were, a composite chilled 

 edge in obedience to the dominant factor of the situation, namely, 

 the conduction of heat away into the cold interior mass 

 downthrown by the fault. 



(b) Glen Coe and Askja compared. — The walls of the Glen 

 Coe cauldron have long since been planed down by erosion, but we 

 can well imagine that the subsidence was originally marked at the 

 surface by an irregular hollow in the midst of a mountainous 

 country. To-day we can do no more than recognize a few steep 

 slopes of hill and valley, the work of an ancient subaerial. erosion, 

 sheltered for ages beneath the vast accumulation of the volcanic- 

 pile. 



It is probable that the lavas gathered most deeply over the 

 cauldron, but also spread out for some distance over the surrounding 

 country. It is certain, at any rate, from the local variations of 

 the volcanic sequence and from the occurrence of a local assemblage 

 of early irregular dykes, that the sources of supply were near 

 at hand. 



When an enquiry is made in Glen Coe as to the possibility of 

 marginal craters having poured forth lavas into the cauldron, the 

 evidence is found to be inconclusive. The early fault-intrusions 

 may have fed many of the hornblende-andesites preserved in the 



