64 Transactions of the South African Philosophical Society. 



characteristic "peg- structure," but this feature is not so well 

 developed as in the melilite of Spiegel Eiver and Matjes Fontein ; 

 irregular patches of a colourless, almost isotropic, mineral are basal 

 sections of the melilite. There is usually one cleavage crack in the 

 lath-shaped sections, and the edges of the lath are irregular owing 

 to the' surrounding grains, particularly of magnetite and perofskite, 

 projecting slightly into the melilite. The melilite gives the usual 

 silica jelly on treatment with weak acid. The large olivine crystals 

 are usually corroded but otherwise unaltered; they reach half an 

 inch in length. The ground mass and smaller mineral constituents 

 of the rock are enclosed within the corrosion cavities. No glassy 

 matter appears to be left in this rock, but the fibrous ground may 

 have been derived from glass. 



This rock can be called a melilite-basalt, and its points of resem- 

 blance to others of a similar nature will be referred to later. 



The contact with the surrounding sedimentary rocks can only be 

 seen on the south-western edge of the pipe, but the thin sandstones 

 and shales exposed there are distinctly hardened within a few feet 

 of the pipe. A columnar structure, with the columns usually at 

 right angles to the pipe wall, is well developed at places ; sometimes 

 the columns are arranged radially in bunches. The strata dip away 

 from the neck at various angles up to 20°. 



The pipe No. Ill is an irregularly shaped area of melilite-basalt 

 with more melilite than the rock from No. I., but no pale-brown 

 mica ; olivine, perofskite, and magnetite are abundant. It contains 

 inclusions of an earlier consolidated melilite-basalt with much pale 

 biotite. The melilite is arranged in a stream-like fashion round the 

 large olivine crystals. No tuff or breccia has been found in this 

 vent, but its junction with the surrounding rocks is not exposed, and 

 as there are no artificial excavations in it, the absence of tuff may be 

 more apparent than real. 



No. IV. is an elliptical area of melilite-basalt about 300 feet long 

 and 100 feet wide. Its junction with the sedimentary rocks is 

 hidden. 



The remaining necks — Nos. V.-VIII. — are filled with rocks of quite 

 different nature from those already described. They consist of 

 altered vesicular lavas and rocks containing so much debris derived 

 from sandstones and shale that they must be looked upon as 

 agglomerates and not as igneous rocks charged with foreign matter 

 No melilite-basalt has been noticed in these pipes. The boundaries of 

 these vents as marked on the plan (Fig. 1) are not accurate ; there 

 is so much sand and gravel washed down from the hills to the west 

 that they are for the most part hidden. There is no doubt as to the 



