212 ME. A. HARKEE ON THE OVEETHRTTST [May I903, 



in most parts of the complex, the process has, of course, gone much 

 farther. Thus the flowing-movement, while bringing out more 

 prominently the heterogenous composition of the rock-body, has 

 effaced the evidence as to how the heterogeneity arose. The missing 

 links are supplied by a comparative study of different localities. 

 Given a granitic magma enclosing debris of more basic rocks, an 

 irregular distribution of the debris such as is seen where the 

 xenoliths are still traceable, reactions between the basic rocks and 

 the acid magma of a kind familiar in many other districts, and that 

 drawing-out of the whole by flowing-movement which is proved 

 by the banded structure, we have a complete explanation of the 

 principal part of the Rum gneisses. 



There remain certain thoroughly-basic rocks which form bands in 

 some parts of the complex, but make up only a small fraction of the 

 whole. These illustrate another principle elsewhere abundantly 

 exemplified in the British Tertiary suite, namely, the tendency of an 

 acid intrusion to follow closely the line of an earlier basic intrusion, 

 often accompanied by noteworthy reactions between the two rocks. 

 I have not thus far expressed any explicit opinion as regards the 

 source of the ultrabasic and basic xenoliths discussed above. It is 

 possible that they were in part brought up by the acid intrusion, 

 having been derived from underlying rock-masses ; but of such 

 concealed masses we have no other clear indication, the known 

 intrusions of gabbro, peridotite, etc., being situated at higher 

 horizons. We may perhaps suppose with more probability that the 

 xenoliths are relics of ultrabasic and basic intrusions which occupied 

 in part the present position of the gneiss, prior to the more 

 voluminous intrusion of acid magma which destroyed them. This 

 hypothesis has the advantage of accounting at the same time for the 

 continuous or lenticular bands of basic rocks which are in places 

 associated with the more acid and hybrid rocks as integral parts of 

 the complex. 



These basic rocks appear to have been of the nature of gabbros, 

 now transformed by metamorphism, and in some measure by inter- 

 change of material with the acid magma. A dark hornblendic rock 

 of this kind, with more or less evident banding and foliation, is 

 a prominent part of the gneissic complex at a locality north-east 

 of Priomh-loch Mor. It has the general aspect of a medium-grained 

 diorite [10733]. In a thin slice it is seen that the deep green 

 hornblende, which makes up more than half of the rock, presents in 

 places the crystal-outline proper to that mineral, proving that it is 

 not merely pseudomorphic but has crystallized as such. The rest of 

 the rock consists chiefly of a finely -striated plagioclase ; but there is 

 also some unstriated felspar, which may be orthoclase, and a few 

 little interstitial grains of quartz are seen. These last two minerals 

 probably point to a certain impregnation of the recrystallized basic 

 rock by the granitic magma. 



For comparison I take another example from the small patch of 

 gneiss enclosed in the porphyritic quartz-felsite of the hill east of 



