that I have examined are: — the raised beach just mentioned, which is about half a mile from the river's mouth, and 40 or 50 feet above the present sea-level ; on the south bank of the Zwartkops, near Rawson's Bridge ; and along the banks of the creek and Ferreira's River, towards Cradocktown. No trace of these shell-beds is found on the surrounding highlands.
This formation is also found protruding through the red clay that forms a portion of the flat, between the outlier marked No. 1 in the section (fig. 4) near the Salt Vlei and the sandstone on the coast near Port Elizabeth. See/and g g in that section (p. 508).
In each locality mentioned this deposit extends a considerable distance. Thus it can be traced from a spot where it is exposed by a cutting through the north bank of the creek, on the Grahamstown Road, across the raised flat over which the road passes towards the Zwartkops, until you reach the wide alluvial flats that extend from the south side of the river. Again it stretches from the road mentioned to the sand-hills along the coast. This portion, however, is covered with soil, thickly interspersed with shingle, mainly composed of pebbles of quartzite, from the distant hills. Wherever any wild animal has burrowed through the soil and shingle, or where any excavation has been made, the shells of this deposit are almost invariably exposed to view. In like manner it evidently extends towards Port Elizabeth — as it there again makes its appearance, as indicated in the section alluded to (fig. 4). We must note that we have here the shingly beds with the red clay superposed ; I shall have again to refer to the last-named deposit.
I have found the shells from these ancient shell-banks on the coast near the mouth of the creek, evidently washed out and spread over the surface.
Returning to our first point of departure, we can trace the shell- bed, exposed at intervals, along the banks of the creek and Ferreira's River, towards Cradocktown, as before mentioned ; and, after passing the mud-flats on the south side of the Zwartkops River, we find it again, extending along the banks of the same for more than half a mile. In this locality it was evidently at one time of much greater extent than at present, and must have occupied the very course of the present river ; but it is being rapidly worn away by the action of the stream.
In some parts (at the bridge) this deposit is nearly 30 feet thick. It is composed principally of sand, the upper portion being interlaminated with bands of limestone. This is especially the case at the western (or what was its inland and most sheltered) extremity. In these bands are imbedded large quantities of shells. At this spot a large and characteristic Panopoea is found. It is generally buried in the lower part of the stratum, and rather frequently with both valves entire. This is the only bank in which I have found this shell, with the exception of a solitary specimen in the raised beach before mentioned. At the creek and the banks of Ferreira's River, I have not found any indications of it. At this place (Rawson's Bridge) a large Solen is also very plentiful, as well as a Mactra and
