10 
Transactions of the Boijal Society of South Africa. 
ing the squares enclosed by the outlines and afterwards cutting out and 
weighing. The felspar was estimated by difference]. Both rhombic 
and monoclinic pyroxenes are present. The relative proportions of these 
vary ; in many rocks the monoclinic pyroxene is more abundant than the 
rhombic species, and in places excludes the latter ; moreover there are two 
monoclinic species. They are all very pale coloured to almost colourless 
in thin sections, the diopside in some rocks is diallagic. 
The specific gravity of five representative specimens of this type gave 
the values: 2-86, 2-87, 2-88, 2-89, 2-91. These exceed the figures for 
anorthosite (2-70) {ibid., Adams, 131 J), but on the other hand are low 
for a typical norite (2-97). 
In the text-books the definition of norite as regards the relative 
amounts of pyroxenic constituents to the felspar and to one another is 
somewhat conflicting. It is variously stated that a norite contains 
plagioclase and pyroxene in about equal amounts and rhombic pyroxene 
is the principal pyroxenic constituent, or that an appreciable amount, 
or an equal amount of rhombic pyroxene makes the rock a norite rather 
than a gabbro, whilst the anorthosites are very low in pyroxenes, frequently 
containing not more than 5 per cent. 
The rock type under consideration, therefore, whilst approaching the 
anorthosites is still nearer to the norites (or to the gabbros). And 
interpreting the definition of norite broadly, together with a consideration 
of the other parts of the intrusion in which rhombic pyroxene is so 
strongly represented, it is perhaps preferable to regard the type as a 
whole a felspar-rich norite. Individual specimens nevertheless, according 
to their mineralogical composition, may be designated felspar-rich gabbro. 
Locally the amount of pyroxene appears to increase considerably until 
the rock may be a typical norite, when the pyroxenes are idiomorphic. 
Enstatitite (see Plate III, fig. 4). — This remarkably persistent rock 
occupies the low-lying broad basic margins of the Great Dyke continuously 
throughout the area mapped. It is probably the predominant type. A 
precisely similar rock is described by F. H. Hatch (" The Geology of the 
Marico District, Transvaal," Trans. Geol. Soc. S. Af., vol. vii., pp. 4-5, 
1904). 
In the Selukwe area it forms low ground, which, were it not for the 
stream sections, affords but few exposures. Occasionally it forms the 
flanks of the Selundi range extending well up the slopes. The two rivers, 
U mtebekwe and Little Tebekwe, flow along the pyroxenite margins, which 
have an average width of about three-quarters of a mile, but may be 
considerably more or less ; nowhere does the rock form the whole width of 
the intrusion as it does in the Belingwe portion {loc. cit., Wagner, map) 
and north of the Selukwe district. 
The rock not uncommonly forms small kopjes 20-50 feet high, 
