Bay, or Robinson's Bay, should show a double slope in cross-section, an upper one consequent on the cycle which produced the terrace, and another consequent on the lowering of the floor of the harbour, which has left the shelf now high above base-level. Finally, the last suggestion would be reasonable if one could definitely demonstrate that the beds which were planed to form the shelf all belonged to an earlier episode. Although it is probable that this is the case, it cannot be taken for granted in view of the general absence of outcrops, and the uncertain relationship of the beds that are visible. If this could be proved I should be inclined to favour the last hypothesis. Thus it is that where the break in the internal profile of the caldera is most definite, it cannot be proved to be associated in its entirety with an earlier volcanic series. Summary. The foregoing gives an account of the field and other relations of the beds which may be concerned with this problem, and the following is a summary of the evidence in favour of there being an igneous substratum exposed to view on which the cone of Akaroa has been constructed. 1. There is first of all the presence of plutonic rocks on Onawe Peninsula. These are not definite proof of the existence of an older igneous substratum, since it is possible that the syenite may be connected in some way with the magma which produced the trachyte of the dykes, especially as there is some chemical affinity between the two. All the same, trachyte dykes do penetrate the syenite and in individual cases must therefore be of later date. The syenite may, however, be more closely related to the earlier trachyte flows and tuffs, and it would then indicate the presence of an igneous foundation to the cone. Also, there is little doubt that the gabbro has metamorphosed some of the basaltic volcanics and, if so, it proves that pre-gabbro volcanics were in position before its intrusion, and even if the gabbro be regarded as the deep-seated equivalent of some of the basic rocks of the cone, it is almost certain that the metamorphosed rocks belong to an earlier epoch. 2. The next circumstance of importance is the occurrence of trachytoid rocks over a wide area in the middle of the caldera (see Plate 8). These rocks, whether they occur as flows, or as tuffs, or as the manifestations between the Carenage and Tikao, are entirely different from the flows of the cone, which are basic in composition throughout, and they certainly suggest an earlier volcanic episode when the rocks were different. It must be noted, however, that rocks of varied composition have been known to be extruded from the same volcanic focus at approximately the same time. The trachyte tuff is well-bedded and has an inclination which changes rapidly in direction and amount, the angles ranging up to 20°. This may be taken as evidence of folding, especially as the overlying basic rocks show a sympathetic distortion. So the waviness in angle is not due to the deposition of material from the air or a pre-existing irregular surface. There does not appear