water-worn pebbles in our diluvium. On closer examination, however, it is found that this hypothesis cannot be the true one. In all the mountains which surround the plains in question, such as Longwood, the Takitimos, the Hokonuis, and the mountains to the east of Wyndham, quartz veins are of rare occurrence, and form a quite insignificant part of the whole. It is also seen that the detritus or gravel formed at the base of these ranges is of quite a different character to that under our notice. The gravel of the beds of the Oreti and Mataura formed from these mountains is, for the most part, blue in colour and composed of very hard sandstone or slate.

The upper plains of Southland, such as the Waimea, Mararoa, Otapiri, and Lora Plains, are composed of gravel of this kind, while the lower plains near the sea-level are composed of heavy beds of milk-white quartz of the kind we are speaking of; and it is further found by the bores that have been made, that these beds alternate with heavy beds of clay and seams of lignite to the depth of more than 200 feet. These beds of quartz could not, therefore, have been deposited from the mountains behind our plains on the landward side.

Similar difficulties stand in the way, if we suppose them to have been derived from the seaward side. The syenite of the Bluff hill contains plenty of quartz, but only as a component of a rock as hard as quartz itself; and which water wears into round balls remarkable for their great elasticity and hardness.

The Stewart Island granites could not account for the deposition of these beds anything more easily; nor the sandstones of Ruapuke, or the Greenhills.

The only remaining possible supposition is, that there were mountains of quartzin sitû, which were degraded on the spot, and left these beds to mark the place where they stood. This is so unlikely that it can hardly be entertained, as mountains of this kind occur nowhere else in the neighbourhood, and even if such had been the case here, the great hardness of such mountains must have resisted the denuding forces as much as the sandstones and the granites in their neighbourhood. No vestige or evidence of such mountains is anywhere seen, while the beds of clay and lignite would have to be otherwise accounted for.

However much geologists may object to it, the true theory of the formation of these pebbles seems to be, that they are silicified wood; and the more they are examined, the more convincing does the proof become that these beds represent, in one condition, the remains of ancient forests, just as the coal beds represent the same thing in another condition. On examining these pebbles closely, it is seen that, in almost every case, the appearance of wood structure can be detected. In some cases it is quite perfect; the

annual growth-rings, the medullary rays, and the vascular tissue being easily seen. Their crystallization is quite peculiar, differing entirely from reef-quartz in being vesicular, or something like what snow is to ice; and much softer than rock-quartz, so that in many cases they can be scratched with a knife. They are all flat-shaped, or knot-like; just as if they had been originally pieces of bark, or knees, or resinous knots, which had resisted the action of ordinary pntrefaction long enough to become completely silicified.

Specimens are found showing the different stages of the process, from lignite to perfect quartz. The small set accompanying this paper may be referred to and described. No. 1, from the Hokonuis, in conglomerate, is unmistakable wood, evidently a root of black pine (Podocarpus spicata) or kowhai (Sophora tetraptera). It is very hard, rings like clinkstone on being struck, and is dark blue in colour, evidently from the carbon not being quite oxidized out. In this respect it is exactly like a clay pipe when insufficiently burnt, part of the carbon of the nicotine remaining in the form of soot to stain the pipeclay blue. In every respect this specimen is perfect stone, giving sparks with steel, and with a specific gravity equal to quartz. Its perfect woody structure and charcoal colour alone betray its origin. In time the blue colour would no doubt have given place to white or grey, when the last vestige of its carbon had been oxidized to CO₂. No. 2 is from the top of the coal at the Nightcaps, and shows the wood first changed to lignite, on the under side, while the upper, or that exposed to the atmosphere, is becoming white, hard, and quartz-like, with a burnt appearance. No. 3 from the same locality shows this burnt appearance to such a degree that one would conclude on looking at it that it had been through the fire. Such, however, could not have been the case as it was detached from the solid seam by the writer.

These specimens show that carbon gets away from wood remains in all probability as CO₂ by slow combustion at ordinary temperature; and when silica is supplied in the same proportion by highly silicated water, the condition has in all probability been attained for the preservation of the structure, after every other trace of its original has disappeared.

Had the water absorbed by the decaying timber been unable to supply the silica in the proper proportion to replace the carbon as it oxidized, caverns in the quartz would probably have been formed, or a vesicular structure, if more nearly equal to the demand,—just what is often observed in these specimens. If the supply of silica was in exact proportion to the departing carbon, perfect opal would be the result; while if from increase of temperature from any cause, fermentation and putrefaction set in, the carbon would get away so rapidly that no silicification could take place,

and no remains whatever would be left to tell the story of the kings of the forest as we see them embalmed in these specimens in their mummy-cases of milk-white quartz.

From this point of view our plant and forest remains are disposed of in nature in three different ways, viz.:—

1. They rot and mix with the soil, where the carbon slowly oxidizes in the earth. This is proved by experiment. The air of the soil is found to contain far more carbon dioxide than the atmosphere, and thus the CO₂ of the soil is far greatest during the summer months, when the temperature is high. Pettenkoffer (Watt's Chem. Dic. 3, Sup., p. 133) found that the quantity of CO₂ in the air of the soil increases very gradually from the greatest depth examined by him—about fourteen feet—upwards to the surface, and that during August and September, at Munich, it was five times greater than it was in January. This can only be from the gradual oxidizing of the woody matter of the soil—at least the presumption is very strong that it is so, although some are of opinion that it may be obtained from some of the lowest forms of animal life.

2. The remains of plants and trees may oxidize so gradually that, in a silicious soil where they absorb silicious water, they may be silicified, and may thus form vast gravel beds of quartz, or of nodules of sandstone composed of quartz, lime, magnesia, potash, etc., in combination, according as the trees or the vegetation were rich in these. In this way our lignite beds may pass by oxidation into sandstones or slate or marl, according as the original vegetation was rich in silica, alumina, or lime, and according as the water absorbed by it was rich in these elements.

3. Or these remains may—by being excluded from the atmosphere by accident, or where deposited in great thickness—form beds and seams of coal which may resist for a long time the oxidizing influence of the air. Coal seams are almost always found to have been protected from the air and from silicious water by dense beds of fireclay above and below, impervious to water and air and other elements inducing change. These deposits depending only on rare and accidental conditions will, therefore, be the exception, and will be the least common way in which the carbon of these remains is disposed of.

These considerations lead us to suspect that vegetation may have had more to do with the formation of many of onr sandstone rocks than is generally supposed. Many things strengthen such a supposition, such as the ash and plant beds so frequently met with, the eminently concretionary character of many, almost all of them, and the strange absence of fossil remains from many of our sandstones. The red and blue slates of our

Maitai formation, for instance, are not generally so metamorphosed as to have destroyed the fossil remains which were almost sure to have occurred had these beds been laid down by the agency of water.

The remaining specimens numbered 4, 8, 9, 10 still the structure of wood, but are completely converted into quartz, with a specific gravity of 2.6 to 2.8, and having grains of magnetic black sand, or thin laminæ of mica here and there between the growth-rings of the original wood, and in the caverns of the structure. These specimens may be picked up in thousands in our streets and in our gravel pits and cuttings, indeed scarcely a piece of quartz can be picked up which does not show woody structure. No. 5 is silicified wood resembling chert where the woody fibre is quite distinctly seen. No. 13 has woody fibre very fine and dense, but with true veins of crystalline quartz transverse to the fibre, just as if the wood in the lignite stage had, in shrinking, cracked and admitted the silicious water to deposit, amidst the chemical changes going on, true crystalline quartz.

The water-worn condition of these pebbles must have resulted from a submergence, probably very slight, of the plains for some time. Indeed the lignite beds alternating with beds of clay and quartz gravel prove conclusively that this was the case, and that such alternations of level must have taken place, a great many times, during probably long periods, since we meet with thin seams of lignite, alternating with clay and gravel in the most natural way, for more than 200 feet in the bores that have been put down in the neighbourhood of Invercargill. These plains, then, on which we live and move to-day, but slightly elevated above the tide, have been so (only sometimes just as much below tide-mark as they are now above) for long periods, during which immense forests grew, decayed, and became quartz gravel, while for correspondingly long periods the tide washed over them, covering up with clay the deposits of timber to make lignite of them, and polishing the pebbles which had passed into a more advanced stage of change through the oxidizing of the carbon of these vegetable remains.

This natural oxidizing of the carbon of the vegetable world at ordinary temperatures, or at temperatures considerably elevated under the surface, is probably a process which has not been comprehended in all its magnitude and importance. The small amount of carbon dioxide in the atmosphere (only about 3 in 10,000) has probably a misleading effect, leading us to conclude that the process must be very insignificant when the product is so small.

When we consider, however, that all the growing forests of the world, nay, the entire vegetable kingdom, derives its carbon principally from the carbon dioxide of the atmosphere, it will be comprehended what an enormous supply will be wanted. It will want little less than the oxidizing of