640 DR. J. A. XHOMSON ON THE [Dec. I9I3, 



Owing to the large dimensions of the hornblende-crystals, only a 

 few of these appear in a thin section of average size. They are 

 made up of variously coloured amphiboles, ranging from a deep 

 brown through a vivid blue-green to pale-green and almost colour- 

 less varieties, the last two being in excess. The lustre-mottling is 

 the outward expression of a pcecilitic structure — due in part to the 

 enclosure of narrow prismatic masses of saussurite within the 

 hornblende, and in part to the presence of ellipsoids of tremolite, 

 surrounded by the coloured hornblende, but in crystallographic 

 continuity with it. The saussurite-pseudomorphs are nearly 

 opaque, and consist predominant^ of clinozoisite and epidote, 

 with relatively little albite ; they still retain the crystal form of the 

 original felspars very faithfully. Iron-ores are not abundant, and 

 are altered to leucoxene or granular sphene. In chloritized specimens 

 pyrite is also present. 



The brown hornblende must certainly be regarded as original. 

 Like that of the rocks described in the last section, it is often 

 rendered nearly opaque by rows of ferruginous inclusions obliquely 

 transverse to the cleavage-planes. In 1908, when studying the 

 hornblende-peridotite and lustre-mottled amphibolite of Glenda- 

 lough, I ascribed this structure to the magmatic resorption of 

 augite 1 ; but subsequent study of many similar rocks which exhibit? 

 it from Cornwall, Scotland, New Zealand, and Western Aiistralia, 

 has led me to the view that it is the result of an accommodation to 

 increasing pressure according to the volume law. As it is known 

 to develop only in the magmatically-formed brown hornblende of 

 igneous rocks, it forms a valuable means of verifying the original 

 nature of the brown hornblende of anrphibolites ; and, further, when 

 the original hornblende shows this structure, the distinction of the 

 secondary hornblende is rendered more certain. 



The deeply coloured blue-green borders of the brown hornblende 

 recall those so common in hornblende-peridotites, in which rocks 

 they are usually explained as a bleaching of the brown variety ; 

 although, possibly, they arise by an interchange of material with the 

 pilitic hornblende formed from the olivine. The bluish hornblende 

 is not arfvedssonite, as supposed by Mr. Simpson. 2 It is to be dis- 

 tinguished from the blue-green hornblende of contact-altered 

 amphibolites by its inconstant character, for it varies in colour 

 gradually and irregularly from place to place ; whereas in the latter 

 the colour remains constant in any one crystal, or is suddenly 

 replaced along a crystallographic plane by a different colour. 



The pale and often fibrous hornblende occurring in the interstices 

 of the larger crystals as outgrowths on the brown and green 

 hornblende, or as patches within the latter, must be regarded 

 as secondary, and of uralitic or (as suggested below) of partly pilitic 



Q. J. G. S. vol. lxiv (1908) p. 481, figs. 2 & 7. 

 Bull. Geol. Snrv. W. Austral. No. 6 (1902) p. 64. 



