B. K. JSf. Wyllie & A. Scott — Plutonics of Garabal Hill. 501 



between the diorite and the pyroxene-olivine rocks. It is some- 

 times penetrated by the diorite in such a way ' that the two are 

 hard to distinguish. In thin section this rock is found to be 

 practically a hornblende rock. The hornblende is brown, with 

 pleochroism from dark brown to light yellow.^ The crystals are 

 uniform in size and packed closely together with few interstices. 

 The cores of the crystals are very commonly of diallage, with flecks 

 of hornblende arranged symmetrically throughout them. These 

 pj'roxene cores, again, show a sprinkling of parallel-orientated 

 Schiller rods. Even in crystals which have no colourless core these 

 schiller patches occur, and the explanation is obvious that all the 

 hornblende was formerly pyroxene. 



Grains of rhombic pyroxene (enstatite) occur, sometimes in clusters, 

 more often separately. Generally they are seen to be an earlier 

 formation, the hornblende being moulded round them, or enclosing 

 them. Interstitial felspar is scanty or absent. But veins occur 

 which are more felspathic, the felspar then amounting to about 

 10 per cent of the whole. Where determinable, this felspar presents 

 the characters of andesine. Grains of apatite are also an important 

 accessory, and a few scattered flakes of biotite can always be 

 distinguished. 



This rock seems to be unique in Scotland, and as we have not 

 been able to parallel it elsewhere, we propose to refer to it for 

 the present as Davainite, which we therefore define as a rock 

 consisting essentially of brown hornblende, which is paramorphic 

 after pyroxene, the total amount of other minerals such as 

 hypersthene and felspar being small. The name hornblendite we 

 reserve for an end-member of the diorite series, to be mentioned 

 hereafter. 



The other ultrabasic rocks show, under the microscope, diallage, 

 enstatite, olivine, brown hornblende, biotite, and black oxides, with 

 a constant but small amount of apatite. Olivine is generally 

 serpentinized where it is present in co»siderable amount. Biotite 

 is almost always present, though never abundant. Enstatite is 

 generally subordinate to diallage, but usually fairly abundant. 

 Hornblende is the same brown variety described in Davainite, 

 Here also it seems to be new-formed from pyroxene, and all stages 

 of the replacement are found. Some specimens from the centre 

 of the Loch Garabal mass show no hornblende. It seems to increase 

 in amount towards the margins. Locally, a small amount of basic 

 plagioclase (bytownite) appears. 



Diallage is by far the most abundant constituent. It occurs in 

 large crystals, giving the rock a porphyritic aspect. In one case this 

 porphyritic structure is quite real, as the interstitial matter is 

 granular olivine, enstatite, and smaller diallage crystals. The rock 

 in which this is found is a vein, intrusive into a finer-grained 

 pyroxenite, so that the phenocryst explanation is reasonable. The 

 diallage crystals are commonly rendered dusty or dense by parallel- 

 orientated schiller inclusions. 



^ The pleochroism of the hornblende is: X, light yellovi^ ; Y, dark brown ; 

 Z, reddish brown. 



