A CONTRIBUTION TO THE PETROGRAPHY OF BENGUELLA. 557 



texture, but occasionally becomes massive and then shows aggregate polarisation. 

 This mineral retains the cleavage of a mica, although it is not so perfect as in the 

 biotite. The double refraction is rather lower than that of biotite, but its pleochroism 

 is usually very feeble. In some sections, however, a pleochroism from orange to a 

 pale-green was observed. The mineral is sensibly uniaxial. These characters agree 

 on the whole with those of the penninite variety of chlorite. In the text-books, how- 

 ever, the double refraction of penninite is stated to be extremely low, whereas the 

 mineral in question has a double refraction not much lower than that of biotite itself. 



The pyroxene occurs in small euhedral crystals, or as granular aggregates em- 

 bedded in the plates of calcite and biotite. It is grass-green in colour, with only a 

 slight pleochroism from green to a yellowish-green. The cross-sections are frequently 

 rectangular, with well-marked cleavages parallel to the rectangular sides. There is 

 occasionally a slight truncation of the corners giving rise to octagonal sections. The 

 extinction diagonally bisects the cleavages and the rectangular sections. The crystals 

 are therefore prismatic with only a feeble development of the pinacoid faces. There 

 is great dispersion of the bisectrices, and some of the sections fail to give perfect 

 extinction in white light. Many of the pyroxene crystals are mere shells, the interiors 

 having been entirely replaced by calcite. 



The biotite and augite are embedded in a groundmass of large allotriomorphic 

 plates of calcite. Within the latter, or enveloping biotite or augite, are found areas 

 of a colourless mineral with very low double refraction and a dark-blue polarisation 

 tint. There is occasionally some approach to a prismatic form, and on the thin edge 

 of sections a cleavage may be observed parallel to the direction of elongation, which 

 is also the negative or "fast-ray" direction. The extinction is straight in regard to 

 the cleavage, the mineral is optically uniaxial, and has a refractive index considerably 

 higher than that of Canada balsam. These characters agree with those of melilite, 

 but the characteristic peg structure of this mineral has not been observed. 



There are numerous irregular patches of a black mineral, which in very thin 

 section becomes translucent and of a beautiful blue colour. The identification of 

 this mineral as anatase was put beyond doubt by finding it growing, in small 

 rectangular tablets, out of decomposing ilmenite. Apatite is also very abundant in 

 small euhedral crystals. In one of the slides (209d) one of the bands or schlieren 

 consists of biotite and granular pyroxene in a groundmass consisting entirely of 

 grains of apatite. In another section a considerable amount of soda-orthoclase 

 appears in large anhedral plates, with which the biotite forms peculiar poikilitic inter- 

 growths. Notwithstanding the alteration of the rock, and the rather uncertain 

 identification of melilite, the general character of this rock is similar to that of the 

 alnoite group, as defined by Rosenbusch,* and it is here provisionally classed with 

 that group. The type alnoite of Alno, Sweden, differs from the rock here described 

 only in containing serpentinised olivine, garnet, and perofskite as the titaniferous 



* Rosenbusch, Gesteinlehre, 1910, p. 304. 



