104: Transactions of the Sontli African Pliilosopliical Society. 



are very common. Ordinarily these amygdules are romided, but in 

 the Drakensberg a pecuhar variety occurs in the shape of long, thin 

 pipes, about the thickness of a lead pencil. These often anastomose 

 and form branched bodies like some of the corals. The pipes stand 

 upright in the flow, and in typical cases there is at the bottom, at 

 the contact with the sandstone or underlying lava, a compact mass- 

 which has cooled quickly. Then, half an inch above the lower 

 surface, there is a zone of small vesicles ; then, again, a thin layer of 

 compact rock. Above this comes the regular layer of pipe amygdules- 

 standing thickly packed together, and averaging 4 or 5 inches in 

 length ; sometimes in the thicker flows there is another zone of 

 pipe-amygdules above this first one, and separated from it by a 

 layer of compact rock, but more frequently this second row consists- 

 of short pipes or vesicles elongated in a vertical direction ; five to- 

 eight separate vesicles have been counted arranged above one of the 

 pipes below. Such a prodigious development of large vesicles is 

 not known to occur elsewhere, and points to a great amount of 

 water-vapour held occluded in the molten magma than usually 

 happens ; it also points to the sudden eruption of the magma, which 

 did not allow time for the water-vapour to get expelled on the way 

 up from the regions of intense pressure. Agates in the form of pipe- 

 amygdules have been for a long time familiar to the diamond diggers 

 on the Vaal Eiver, where they occur with ordinary rounded agates ; 

 they have prol^ably been derived from the disintegration of the 

 Drakensberg and Maluti amygdaloids. Many of the lava-flows, 

 however, do not show this structure, but are filled with rounded 

 vesicles, and others again are compact crystalline rocks, from which 

 the excess of water-vapour was got rid of before extrusion. 



Some of the lava-flows immediately above the sandstones are full 

 of pieces of sedimentary rocks, which they have picked up in their 

 course, and held up in their substance when liquid. A good example 

 of this is to be found in Glen Alfred, where the base of the flow 

 is like ail agglomerate, so thickly studded are the bits of foreign 

 material. 



The two Deer Park volcanoes already mentioned are by no means 

 the only ones on this farm. The general trend of the whole group 

 in Matatiele is 60° E. of N., but the Deer Park volcanoes occupy a 

 line nearly at right angles to this direction, and on the south side of 

 the Drakensberg. I have been able to find altogether six distinct 

 vents in a distance of about three miles, the two referred to above 

 are the largest, the otliers are very much smaller, and the two ones 

 at the southern end of the line are only 6 and 4 yards across. 

 Possibly there are others, and I strongly suspect that the orchard by 



