The Cape Peninsula stretches from the surrounds of Cape Town southwards towards its termination in the Cape of Good Hope and Cape Point. The terrain is the classic Fimbost, a rugged country of exposed rock and loose sand covered by native Protea plants and newer immigrants such as Wattle. The geology is mostly sandstone, though there are a few exposures of limestone. This area of South Africa is not renowned overseas for its caves, the region further inland with its Kango caves is much better known. However, the Cape Peninsula is very interesting because its caves are in sandstone. The morphology of the cave passage is very similar to limestone caves and there is a considerable horizontal development (a maximum reported of 400m+), though nothing significant vertically.

Two distinct types of cave are evident, the classic rock-shelter and the more complex cave analogous to limestone development. Two rock shelters are worthy of mention; Elephant's Eye Cave in Constantiaberg which is a large entrance visible for many kilometers, and Muizenberg cave which has a smaller rear entrance passing through the mountain and has evidence of Stone Age habitation. These rock shelters appear to be classic erosion features in soft rock, the rear entrance in Muizenberg cave is related to a fracture in the rock. These caves which are little more than surface features, are widely distributed in the region. The more interesting and more complex caves are confined to one small part of the hills of the peninsula.

Behind the coastal towns of Muizenberg and Kalk Bay is a small range of rocky hills. Here, three hills, Kalk Bay Mountain, Ridge Peak and the aptly named Cave Peak contain over eighty caves. These vary in complexity from Avernus, a small chamber with two 10m entrance passages to Ronan's Well with over 400m of tight strenuous passage. The surface morphology of the hills are strangely karst-like, though sink features are absent. I did make the elementary acid test on the rock and it didn't seem calcareous. To reach the very top of Cave Peak required a scramble and easy climb. The cave entrances were at the foot of a summit block where a steep slope intersected a 10-20m crag. The hilltops are of exposed rock, highly eroded and faulted. The cave development is controlled by these faults, chiefly vertical faulting though some features show a definite influence of the horizontal.

A typical and very popular cave is Boomslang Cave. This cuts the ridge, offering an easy through trip for many a local hill walker. The "front entrance" of the cave is typical of the caves I saw. A vertical fault intersecting a rock face had been eroded deeply into the face, the entrance was at the back of a small ravine above a pile of boulders. The passage continued along the line of the fault, usually it was about 2m wide and up to about 15m in height. When we were maybe 40m or so inside the cave a hole was visible in the roof; a search on the surface failed to find this pitch entrance. The passage closed in and a stoop round led to a little chamber with about 10cm of water on the floor. There were three or four ways out of this room, one got low and wet very quickly, another led into a wide passage of stooping height and then into a largish circular chamber of about 10m diameter. From this chamber there were several small passages, one leading back into the chamber and two leading to the outside world on the other side of the ridge. We took one of these and it too was associated with a visible vertical fault. However about 5m from the new entrance the passage entered a little widening and the exit was controlled by a horizontal fault which created a small flat-out crawl. Following this, we exited having tunnelled completely through the hill.

The cave passages between the crawl and the main chamber were dry and their floors were covered in a coarse grey sand. The passages associated with the vertical entrance fault had a thinner sand flooring and also had extensive damp or wet lengths. The presence of water suggests that these are erosive rather than tectonic features, although development is clearly controlled by the extensive faulting. Sandstone caves are not unique to this region of the world, French expeditions to Venezuela found huge sheer sided pits (two hundred metres deep!) in the sandstone Roraima (Lost World) region. These pits had cave passage at their bottoms. Although there are no similar caves in Britain the sandstone in the Orkneys has huge sea eroded gorges many times longer than they are wide in its cliffs, these "geos" are similar fault controlled structures.

Brian Derby