264 Transactions. 



As regards alteration, the plagioclases vary from quite fresh individuals 

 showing good twinning to highly decomposed crystals in which the original 

 twinning has all but disappeared, and is indicated only by a faint banding 

 in portions of the crj^stals.* From these it would seem but a small step 

 to the highly altered feldspars without a trace of twinning that are com- 

 monly present. The alteration-products are largely calcite and quartz. 

 Some muscovite is occasionally present in small flakes, and in odd instances 

 haematite fills the cleavage-cracks. The most-altered plagioclases (those 

 with just a trace of repeated twinning) show a mosaic which is mainly 

 secondary quartz and feldspar. The exact nature of the latter is not 

 always apparent, but in some cases it can be determined as valencianite. 

 In one instance where a microscopic vein of quartz and calcite crosses a 

 section valencianite may be observed replacing a crystal of plagioclase 

 intersected by the vein. 



Nearly all the sections have feldsjDar crystals that by their optical pro- 

 perties appear to be orthoclase. These are in some instances fresh, and 

 show Carlsbad twinning. One such crystal, with extinction angles of 

 19 J° and 19|°, has one half slightly but decidedly zonary. Many of the 

 orthoclases, however, are much altered, and consist mainly of a mosaic of 

 quartz and calcite, with some muscovite, iron-oxide, &c. The apparently 

 residual feldspar in many cases suggests valencianite, but to prove decisively 

 that it is really secondary is no easy matter. A consideration, however, of 

 the manner in which transition forms occur between undoubted plagioclase 

 and the crystals in question leads to the conclusion that the latter must 

 represent highly altered lime-soda feldspars. In this connection the almost 

 invariable presence of calcite is strongly suggestive of an original lime- 

 content. 



Original quartz grains are present in all the sections. Some sections 

 have only one or two small quartz grains, others have several compara- 

 tively large phenocrysts. All the quartz grains have rounded outlines, 

 in many instances with deep bays, and thus show clearly the effects of 

 corrosion by a fluid magma. Very rarely can an approach to bi-pyramidal 

 outlines be seen. Occasionally the grains of quartz seem to include small 

 rounded patches of the groundmass, but these are probably all, or nearly 

 all, embayments cut across in the sectioning. 



Ferro-magnesian minerals are represented by fairly numerous pheno- 

 crysts, generally entirely altered to chlorite, serpentinous matter, iron- 

 oxides, &c. In a few sections augite is quite recognisable. Hypersthene 

 or similar rhombic pyroxene seems to have been originally present Ln every 

 section. Hornblende is probably represented in odd sections by dark 

 lozenge-shaped masses similar to those described by Park (6, p. 343). 



The other original minerals present call for little description. Small 

 crystals or grains of magnetite are always seen. Some of these, however, 

 are presumably due to the decomposition of ferro-magnesian minerals. The 

 magnetite is often partly altered to a leucoxenic-looking mineral, but this, 

 according to Finlayson, is probably siderite. Apatite can usuhUy be dis- 

 tinguished as small needles in the feldspars. Zircon is possibly present in 

 one or two sections. 



The groundmass in the various sections varies considerably in amount, 

 but may be said to form much more than half the rock, as a rule. In the 



* Sollas speaks of similar feldspars as showing traces of microperthitic structure 

 (11, vol. 2, pp. 18, 54). 



