F. P. Mennell — 8oulh African Petrographij. 359 



rounded grains of olivine, this being the only case in which I am 

 able to record olivine from the rocks of the Zambesi Valley. The 

 groundmass is largely of the usual type, a granular aggregate of 

 lath-shaped felspar, augite, and magnetite, but there is also a con- 

 siderable amount of glassy material, which is either brown in colour 

 and isotropic, or bright orange, with an imperfect spherulitic structure 

 between crossed nicols. 



Though by no means the most abundant type, South Africa 

 affords some good examples of olivine basalt. A beautiful one ^ 

 occurs amongst the rocks which surround the volcanic pipes of 

 the Kimberley diamond-mines, from which they were no doubt 

 originally extended. It is holocrystalline, with a well-marked 

 ophitic structure, and I know of no rock in which the distinction 

 between olivine and the pyroxenes is so well indicated. Both augite 

 and enstatite are present, and their cleavages are well developed, 

 pinacoidal ones being shown by both rhombic and monoclinic variety 

 in addition to that parallel to the prism. Twinning is very common, 

 and the felspars penetrate the crystals in all directions. The olivine, 

 on the other hand, shows distinctly higher refraction and stronger 

 double refraction than the pyroxenes, and the absence of cleavage is in 

 striking contrast to its conspicuous development in their case. It occurs 

 in rounded grains and granular aggregates, which are not as a rule 

 penetrated by the felspars, and are traversed by irregular cracks, 

 sometimes rendered more apparent by incipient serpen tinization, 

 though in no case has this proceeded far. The felspars give lath- 

 shaped sections showing repeated twinning, and evidently belong 

 to a variety near anorthite, the maximum extinction angle being 

 about 40°. There are also larger allotriomorphic crystals, which are 

 untwinned but are strongly zoned in layers of different composition. 

 Rather large irregular granules of ilmenite are somewhat sparingly 

 distributed, and perofskite also appears to be present as an occasional 

 accessory. 



We now come to the consideration of the dyke rocks, which are 

 more evenly distributed among the various types. I have adopted 

 the term granophyre for the acid division, as it is etymologically 

 a highly appropriate term, but not without some misgivings owing 

 to its having unfortunately been restricted by many authors to those 

 rocks in which the groundmass assumes a micrographic structure. 

 Quartz felsite would perhaps be less open to objection in some 

 respects, but any such double-barrelled expression is very cumbrous 

 when a prefix becomes necessary. With this explanation I do not 

 think that my usage of the term need give rise to any misconception. 



Some beautiful granophyres, in the ordinary acceptation of the 

 word, occur near Fort Gwelo in Matabeleland. Large phenocrysts 

 of felspars (chiefly plagioclase), and more rarely of corroded quartz, 

 are set in a groundmass which is partly spherulitic and partly 

 ' microgranitic' The former type largely predominates, and the 

 spherulites are frequently seen to be composed of radially disposed 



1 This is uo doubt the rock referred to by Professor Bonuey, Geol. Mag., 1897, 

 p. 449. 



