624 MESSES. CLOUGH, MAUFE, AND BAILET OX [Nov. I909, 



of an orderly sequence in the eruptions. The constitution of the 

 pile of lavas seems to indicate that the district was supplied from, 

 more than one centre, under such conditions that individual foci 

 ■were independent of one another in regard to the type of material 

 erupted, even though their periods of activity overlapped to a very 

 considerable extent. On this hypothesis alone it seems possible to 

 understand the variations of group 2 (p. 618). 



(3) The great rarity of pumiceous varieties of lava, and the corre- 

 lative absence of tuff's possessing typical ash-structure, lead to the 

 conclusion that the eruptions were not as a rule accompanied by 

 explosion, but belonged rather to the quietly welling, effusive type. 

 Such fragmental rocks as may have a-pyroclastic origin form but a 

 small fraction of the volcanic succession, and are of a kind not 

 easily distinguished from the true sedimentary rocks with which 

 they are associated. 



(4) The volcanic rocks accumulated on a denuded and irregular 

 land-surface of the Highland Schists, which was relatively smooth 

 where formed of easily disintegrated phyllites, but carved into crag 

 and ravine where constituted of the more resistant quartzites and 

 quartzose schists. The deposits commonly intervening between the 

 lavas and the old floor comprise breccias formed of the shattered local 

 rock, angular scree-like breccias, and also boulder-conglomerates 

 such as may have been accumulated by torrents. 



Thin lenticular beds of shale and grit and coarse breccias recur 

 at intervals between the lavas, some of which have their cracks 

 and crevices filled up by horizontally bedded silt. Such deposits 

 point to the presence of streams, and probably also of temporary lakes 

 occupying the floor of the cauldron. From a consideration of the 

 whole evidence it is thought that the eruptions were subaerial. 

 The suggestion is made that some anomalously situated breccias 

 and patches of lava may be due to the infilling of landslip-cracks 

 and earthquake-rents. 



(5) Away from the faulted margin of the subsidence, the volcanic 

 rocks possess a general, but variable dip in a southerly direction. 



III. The Botjndaky-Fatjxt and the Fault-Intktjsions. 



(a) General Description. 



In view of the important theoretical issues involved in the 

 interpretation of the cauldron-subsidence of Glen Coe and its atten- 

 dant igneous phenomena, detailed descriptions are necessary to show 

 the nature of the evidence which is encountered from point to point 

 along the course of the boundary-fault. Before entering upon a 

 piecemeal consideration of this evidence, however, a general de- 

 scription of the more important features has been compiled under 

 several headings, without regard to geographical arrangement. In 

 the next section local details will be considered in the order in 

 which they present themselves to anyone following the circuit of the 

 fault, but repetition will be avoided so far as is possible. 



