Vol. 59.] TOKKIDONIAN", ETC. OF THE ISLE OF BUM. 211 



great part of the latter are susceptible of explanation on these lines. 

 A minute description of the rocks would be merely an expansion of 

 this general statement, and for our immediate purpose a few brief 

 remarks will suffice. 



Where recognizable xenoliths occur in rocks forming part of the 

 gneissic complex, they are, if not in course of dissolution, at least 

 highly metamorphosed, and their original nature is not always 

 patent to observation. In some cases they are of a compact black 

 rock, which looks like an indurated shale, and in thin slices shows a 

 laminated structure marked by abundant brown mica and granular 

 magnetite. There are, however, none of the special aluminous 

 minerals, such as sillimauite, which usually characterize highly- 

 metamorphosed argillaceous sediments, and the presence of porphy- 

 ritic felspars is proof of an igneous origin. The majority of the xeno- 

 liths clearly come from ultrabasic and basic igneous rocks, though 

 often apparently of finer texture than the common peridotites and 

 gabbros of the island. They usually illustrate an advanced stage of 

 replacement, a term which I use to denote the transformations 

 consequent upon interchange of substance with the enveloping acid 

 medium. It is a common observation in other districts examined 

 by me that such replacement may proceed almost without limit 

 while the sharp boundary of the original enclosed fragment remains 

 intact, and of this the Rum gneisses afford abundant illustration. 

 The xenolifch thus comes to be replaced by a cast or pseudomorph, 

 preserving the former outline but consisting of a granular aggregate 

 wholly of new formation, to which the acid magma has contributed. 

 Granules of pyroxene are the principal elements of the replaced 

 xenoliths. They are partly of augite, but more commonly of hyper- 

 sthene, a mineral which might confidently be expected from the 

 reaction of olivine with a magma rich in silica. The reciprocal 

 modification of the acid matrix, essentially an enrichment in the 

 dioxide-bases, shows itself in a somewhat diminished proportion of 

 quartz and an increased prominence of the ferromagnesian silicates, 

 lime-bearing felspars, and magnetite. Pyroxene-granules are often 

 abundant, but these are in great part mechanically detached from the 

 borders of the transformed xenoliths. The micrographic structures, 

 which rule in the purely-acid portion of the complex, are usually 

 lost in these partly-basified products, a feature constantly observed 

 in similar xenolithic and hybrid rocks in Skye. 



Rocks of the general type described, sometimes graduating into 

 pure granophyres, show little or no gneissic banding ; and in the 

 well-banded gneisses, which make up the principal part of the 

 complex, xenoliths have usually been obliterated. This is easily 

 understood. We have seen how the enclosed fragments are trans- 

 formed to mere aggregates of granules, and it is evident that the 

 preservation of their original outlines is contingent on a condition 

 of tranquillity in the surrounding magma. Movement would 

 quickly resolve the aggregates into detached granules, and indeed 

 the thin slices enable us to verify this breaking-down process in 

 Where strong fluxion has supervened, as is the case 



