railway station also showed that the silt was stratified, but this section is now obliterated by weathering, and I have thought it important to call the attention of geologists to the section behind the dock while it is still fresh; for no doubt it will soon become obliterated like the one behind the railway station.

The origin of this silt deposit is of considerable interest, as it is important evidence in discussing the question of the latest oscillations of level in New Zealand. Dr. von Haast, in his Report on the Geology of Canterbury and Westland (1879), p. 367, calls it “The Loëss Formation,” compares it with the loëss deposits of China described by Baron von Richthofen, and says that “the general character and position of the principal loëss (or loam) beds in this province prove clearly that they have been formed by the modus operendi pointed out by Von Richthofen.” There is, however, one difference which he mentions, “and that is the absence in the Canterbury beds of the peculiar small marly nodules so common on the Rhine, the Danube, and in China.”

The modus operendi in question is the following: The fine particles of earth carried down the slopes by the rain are partly retained by the grass growing on the slopes, and the dust blown across the land by the wind is also retained by the grass, the roots of which also decay and assist in raising the ground;—so that the formation is a mass of grass covered with fine earth and sand brought by the wind and the rain, and has, of course, an entirely subaërial origin. It is characterized by being unstratified, and by having a “peculiar vertical capillary texture,” caused by the decay of the roots of the grass.

I am afraid that the well-marked stratification of the base of the Lyttelton silt deposit can hardly be reconciled with this method of formation, and there are several other difficulties which cannot, I think, be explained on the theory of a sub-aërial origin.

In the first place the deposit is widely distributed and rests upon beds of very different mineral composition. According to Dr. von Haast it is found at the foot of Mount Grey and on the Moeraki Downs, where it lies upon tertiary argillaceous and calcareous sandstones; at the Malvern Hills, where it rests upon secondary sandstones and slates. At the southern end of the Canterbury Plains it occurs from the Orari to Timaru, where it is found on tertiary sedimentary and volcanic rocks; and in some places it lies on the shingle of the Canterbury Plains. South of Timaru it can be traced beyond the Waihao, and in the valley of the Waitaki south of Elephant Hill. From my own observations I know that it is largely developed at Oamaru, where the base is also stratified, and that it extends as far south as Moeraki Peninsula. It is difficult to understand how so

widely extended a deposit, resting on such different rocks, could be formed in the way suggested by Dr. von Haast; and it is also difficult to understand why, on this hypothesis, the deposit should be found only on the coast near the mouths of the great rivers from the Waimakariri to the Waitaki, and not elsewhere.

In the second place the limit in height of the deposit must be noticed. In the “Transactions of the New Zealand Institute,” vol. vi., 1873, p. 423, Dr. v. Haast says that on Banks Peninsula it extends to a height of 800 feet above the sea; and in his Report on the Geology of Canterbury he says that at Timaru it reaches to near the summit of Mt. Horrible, which is 1,272 feet high. No other data can be obtained at present, and the difficulty here is to explain why, on the subaërial hypothesis, its height on Banks Peninsula should be limited to 800 feet, notwithstanding that grass grows, rain falls, and the wind blows at much greater altitudes.

In the third place the silt deposit is not confined to the slopes. At Oamaru it covers the very highest points of Oamaru Cape, as may be seen in fig. 7, page 55, of my report already quoted; and in many parts of Banks Peninsula it is better developed on the ridges than in the valleys, as the accompanying sketch of the cliffs between Little Akaloa and Mackintosh Bay will show (fig. 2); so that the action of rain in its formation must be eliminated.

Fig. 2, a, silt deposit; b, volcanic rocks.

The last point to be noticed is the fossils contained in the deposit. Usually it is quite unfossiliferous, but Dr. von Haast states that moa bones and land shells have been found in it somewhere on Banks Peninsula, but no definite locality is named. The occurrence of land shells is remarkable considering how rare they are now in New Zealand, and that they are almost entirely confined to the bush. I have never heard of any land shells having been found on grass land except in crevices of limestone, or under blocks of the same rock; and there is no limestone on Banks Peninsula, while, according to Dr. von Haast's hypothesis, the formation accumulated on open grass land.

At Oamaru marine shells of still living species are found in the lower part of the deposit up to a height of 60 feet above the sea. A list of the species will be found at page 70 of my “Report on the Geology of Otago,” 1875.