B. K. N. Wyllie & A. Scott — Plutonics of Garabal Hill. 537 



increase ia the amount of amphibole. The latter is the green horn- 

 blende common to the diorites, and is occasionally replaced by mosaics 

 of secondary actinolite. This substitution of felspar by hornblende, 

 accompanied by an increase in the coarseness of the texture, continues 

 till a hornblendite containing about 85 per cent of amphibole is 

 reached. This rock is exposed for a distance of 20 yards to the 

 south, and then the ground is obscured by peat, so that the other 

 margin of the diorites cannot be seen. The hornblendite consists 

 essentially of large crystals, more or less uniform in size and up to 

 3 cm. in length, of green hornblende which has wholly or partly 

 altered to brown, the alteration commencing invariably from the 

 centre. Occasional pyroxene cores are found, and these are generally 

 surrounded by a narrow band of green amphibole, with brown beyond 

 and finally green margins. The interstices are filled up with 

 aggregates of quartz, felspar, and small green hornblendes, with 

 occasional crystals of sphene. The finer-grained varieties of this 

 sequence probably resembled very much the hornblendite which 

 occurs along with scyelite in the Moine Gneiss regions.^ Occasional 

 schlieren, resembling those already described, occur among the 

 coarser rocks. These are generally much more felspathic than the 

 surrounding rocks and contain long prismatic hornblendes up to 6 cm. 

 in length. 



The brown hornVjlende of this series differs considerably from that of 

 the davainites. In the latter case the mineral has developed directly 

 from pvroxene and is invariably brown, while in the former case it 

 has formed from original green hornblende, difi'erent degrees of trans- 

 formation being visible, varying from incipient alteration at the centres 

 to completelj^ brown crystals. The mineral shows pleochroism from 

 deep brown to yellowish green and has strong absorption. It is 

 generally traversed by series of parallel dark bands arranged in lattice 

 fashion and with small outgrowths, the whole somewhat resembling 

 arborescent microlites. Teall- has described similar structures in 

 araphiboles from Cornwall and Anglesey, and Thomson^ also found 

 them in hornblendites from Wicklow. In these cases the dark bands 

 were assumed to be magnetite, formed by the magmatic resorption 

 . of pj'roxene and consequent replacement by hornblende. In the 

 Garabal Burn rock a careful examination, under a high power, of 

 the ends of the bands and outgrowths showed that they are composed 

 of aggregates of minute pale-green or colourless crystals, orientated 

 in irregular fashion with respect to the main direction of the bands. 

 It seems probable that the whole of the dark bands are made up of 

 these minute crystals, the black colour being analogous to that 

 observed in some opaque glasses, where the opacity is due to 

 innumerable colourless longulites, which act as prisms and disperse 

 and totally reflect the light, the degree of opacity varying with, the 

 amount of light totally reflected.* 



^ Geology of Ben Wyvis, etc. (Mem. Geol. Surv. Scotland), 1912, pp. 128-9. 

 2 British Petrography, 1908, pp. 475-94. 

 5 Quart. Journ. Geol. Soc, Ixii, pp. 475-94, 1908. 



■* Of. Pirsson, "Artificial Lava Plow and its Spherulitic Crystallisation": 

 Amer. Journ. Sci., ser. IV, xxx, pp. 97-114, 1910. 



