"Vol. 65.] THE CAtTLDEOlSr-SUBSIDEXCB OF GLEN COE. 665 



with which, the fault-surface is found smooth and unbroken in 

 contact with the fault-intrusion. This relation must be considered 

 in connexion with the extraordinary number of xenoliths character- 

 istic of the fault-intrusion. It seems probable that the magma 

 advanced upwards largely by the process known as 

 stoping. 1 Doubtless the solid schists were heavier than the 

 ascending magma, 2 and accordingly tended to break off along 

 their numerous structural planes of weakness and sink as frag- 

 ments into the intrusion. But of all the planes of weakness, the 

 fault-plane would be predominant, since, as has been seen, it was 

 occupied by crushed material. Such a plane, it can readily be 

 imagined, would tend to be stripped clean and bare when exposed 

 to the operation of stoping, regulated by conditions strongly 

 favouring asymmetry. 



The argument so far employed relies upon the concordant 

 testimony of several miles of section. It is, moreover, corro- 

 borated by the evidence which has been afforded by a detailed 

 examination of the flinty crush-rock exposed at the foot of Stob 

 Mhic Mhartuin (fig. 8, p. 650). 



The flinty crush-rock here consists of finely ground-up quartzite 

 and quartz-schist with a certain proportion of indeterminable base. 

 It also contains, and this is the significant point, a few xenocrysts 

 of felspar (13402, 13403) which are sometimes broken and have 

 clearly been derived from the adjacent fault-intrusion. These 

 xenocrysts are of about the same size as the smaller phenocrysts in 

 the chilled margin of the intrusion, and some of them show well 

 developed crystal-faces. Many are clear ; but others, lying near 

 the edge, are decomposed to the same extent as the phenocrysts in 

 the intrusion. They seem, however, to have suffered more from 

 fracturing and rounding than these phenocrysts, and often present 

 a fragmental appearance. They may occasionally be found even 

 half an inch in from the contact-zone, and separated from the 

 latter by the uninterrupted flow-banding of their host. The 

 position and fragmental nature of many of these xenocrysts proves 

 that they have been involved in the viscous flow of the flinty 

 crush-rock, and since they have certainly been derived from the 

 fault-intrusion, it is manifest that viscous flow still con- 

 tinued in the flinty crush-rock after the arrival upon 

 the scene of some portion of the fault-intrusion. 



Doubtless the phenomenon just described was of strictly limited 



1 Cf. especially R. A. Daly, ' The Mechanics of Igneous Intrusion ' 3rd 

 paper, Am. Journ. Sci. ser. 4, vol. xxvi (1908) p. 17. 



2 Density of quartzite, the lightest schist in the district, = 2*60 : that of the 

 fault-intrusion, Stob Mhic Mhartuin, =2"66. The latter figure gives 2"47 as 

 the density of the same fault-intrusion under ordinary temperatures, but in the 

 glassy condition ; cf. J. A. Douglas, Quart. Journ. Geol. Soc. vol. lxiii (1907) 

 p. 145. With rising temperatures the difference between the density of the 

 uncrystallized fault-intrusion, merging into true liquid, and that of the solid 

 crystalline quartzite would certainly increase. 



