888 MR JOHN S. FLETT ON 



magnetite in fine grains and little brown octrahedra of chromite or perofskite. Augite 

 phenocrysts are few and small, but it forms most of the groundmass in small crystals five 

 or six times as long as broad, and averaging *02 mm. in breadth. It is brownish-violet, 

 very feebly dichroic, with rather perfect idiomorphism, occasional twinning, and frequent 

 hour-glass structure. It lies scattered irregularly through the field, and grouped on the 

 faces of the olivines. Hornblende is practically absent, but a deep brown biotite is 

 present in small amount. It is hexagonal in transverse section, with sometimes a 

 darker rim and a pale centre, far more dichroic than hornblende (the a ray being 

 nearly colourless), and usually adherent to the iron ores and the olivine. In convergent 

 light itjs very nearly uniaxial. The rock is very fresh, but there is a little calcite and 

 chlorite. Between the minerals of the groundmass there is a very small quantity of a 

 transparent, colourless, almost isotropic groundmass, slightly turbid, and showing the 

 weakest irregular double refraction. It is moulded on the other minerals, and has no 

 definite outlines ; the other ingredients are apatite and magnetite. Here and there 

 occur small ocelli, with calcite in the centre, surrounded by a little felspar and a prism 

 or two of hornblende. 



The more northerly of two dykes at the house of Skaill, Rousay, on the shores of 

 Eynhallow Sound, is almost exactly similar to this rock. 



The Dykes of Tingwall [Rendall), of Quoynamuckle (Rendall), and of Kongie Geo, 

 Saviskail [Rousay). — They contain similar olivine and titaniferous augite, often in 

 radiating groups of six or seven. Biotite is absent, but hornblende is rather more 

 abundant, especially in the little ocelli, which may contain a small amount of felspar. 

 The rocks are much decomposed, and the numerous small vesicles are filled with calcite 

 and chlorite. In most of the slides there is a considerable quantity of a very turbid 

 brownish glass, which is filled with little needles of augite and skeleton growths of 

 magnetite ; it contains also microliths of hornblende, and possibly a little felspar. 

 Even where clearest and least altered, it has in polarised light the appearance of a 

 devitrified base, granular and speckled with decomposition products. 



On the shores of Kirkwall Bay, near the Skerry of Quanterness, I picked up in 1896 

 two blocks of trap, which in section turn out to be very interesting rocks. They are 

 coarsely porphyritic, and the phenocrysts resemble in every respect those described for 

 the camptonites of Hoxa. Olivine is absent from the sections (PI. II. figs. 5 and 6), and 

 the rock is thus a hornblende fourchite. The phenocrysts are augite and hornblende, 

 which show all the varieties and intergrowths described in the west dyke of Hoxa (see 

 photos, figs. 5 and 6, PI. II.). Both are very fresh, the augite showing only incipient- 

 change into chlorite. The groundmass contains little augite, and consists almost entirely 

 of hornblende similar to that of the camptonites, embedded in a clear glass. This glass 

 has a tendency to collect in certain parts of the slides, forming paler ocelli; it is 

 colourless or pale greenish, filled with innumerable microliths, which are often horn- 

 blende in beautiful dendritic growths very like those of the pitchstones of Arran. 

 These consist of spicules so thin as to have only a faint brownish or greenish colour ; 



