The Parnell grit, like the Cheltenham breccia, presents a somewhat bedded appearance, due to the linear arrangement of fragments of similar size. In this case, however, the bands are much more regular, and the coarsest fragments—the size, perhaps, of marbles—are all at the bottom, from which point the bed grows gradually finer till it shades off into a sandstone. The outcrops of the Parnell grit everywhere present the most beautiful examples of the shading-off of one rock into another. The red rounded lapilli of the lowest layers are gradually replaced by smaller ones, and these by yet smaller, till at length they can no longer be observed with the naked eye, and on a fresh fracture the rock has all the appearance of a blue sandstone; nor is it possible to say where the sandstone begins and the breccia ends. The lapilli are very uniform in size in each layer, which seems to indicate a distant origin. The average thickness of the bed is about 18 ft. Fossils do not seem to be as abundant as in the older beds.
The weathering is extremely characteristic and useful. In the coarser parts it is similar to that of the Cheltenham breccia, but in the upper parts concretionary or spheroidal structure is usually developed. Round the shells of brown iron-oxide the dark colour which the bed usually presents is absent, so that the concentric layers are plainly visible. This type of weathering seems to be almost confined to this bed, none of the Waitemata sandstones exhibiting it; and it is invaluable as a means of detecting the presence of the bed in inland outcrops, where the colour has generally all been leached out and the bed is left an (apparently) white crumbly sandstone. Under sea-water neither this bed nor the Cheltenham breccia weathers more than a few inches, and the feldspars remain comparatively fresh.
As at Cheltenham, a very common feature of the bed is the occurrence of zeolite “dykes,” dividing the surface of the grit into irregular polygonal plates. These dykes, when split, frequently show beautiful dendritic manganese markings. The markings closely resemble those formed by moist emery-
powder when, after grinding a section, one slowly withdraws the slide at right angles to the iron plate, the markings occurring both on the plate and on the slide. Possibly the dendritic markings on the zeolites were formed in an analogous manner.
The lava fragments are so small and so oxidized that I was not successful in making any microscopic sections. My sections of the grit were scarcely more successful, the rock crumbling away before it was thin enough to be of much service. In one or two instances fragments of lava could be recognised in the sections, containing feldspar and six-sided augite phenocrysts and probably an augite-andesite.
Among the crystals in the tuff corroded quartz grains occur, which may be due to the mingling of sandy sediment, or, since Sir James Hector says that the grit contains fragments of trachytes, may be derived from them or from more acid rocks. By far the most numerous crystals are small oblong feldspars sometimes altered and milky, at other times fresh and bright, with good cleavage. These are very plentiful throughout the bed. A few broken fragments of augite crystals are also present, but the large perfect augites described as occurring in the Cheltenham breccia are conspicuous by their absence. The grit is largely composed of fragments of greenish sandstones and slates which may be Palæozoic rocks. Large blocks of sandstone are absent, and current bedding is unusual. Iron-pyrites is plentifully disseminated in bright-yellow flecks. The black matrix is often studded with scarlet scoria and white feldspars, and forms a handsome rock. From the plentiful occurrence of feldspar and augite the grit may be best described as an augite-andesite tuff.
I do not consider the source of the Parnell grit by any means as certain as that of the Cheltenham breccia, but, on the whole, the evidence is perhaps in favour of a source near Cape Colville. Before giving what evidence there is in support of this I must freely admit that the grit may have come from the Waitakerei vents, like the other volcanic beds of the series. But in the first place it is a good deal younger than the Cheltenham bed, as is shown by the fact that it crosses the Maitai ridge, which apparently was above water when the former bed was deposited; and so it is quite possible that the Waitakerei vents may have become quiescent. In the second place, it is a bed with a very wide distribution, extending from Ponsonby to Turanga Creek, from Manukau to St. Helier's. Yet it is not a coarse breccia. Such a widespread bed must have been, one would think, the result of very violent eruptions, and if the eruptions at Waitakerei were very violent the bed at the Manukau ought to contain some
coarse lumps. But, if it came from Coromandel, a great outburst might result in just such a bed at so considerable a distance. In the third place, it grows coarser, on the whole, in an easterly direction. At the Manukau the bed at the White Bluff, though evidently the same, is yet rather finer than at St. John's, to the east. I cannot say that I have observed any increase of coarseness between Ponsonby and Parnell, but between Parnell and St. John's College there is a distinct increase. At Tamaki Head the bed is coarser than at St. John's, while the outcrop in the reef is, on the whole, intermediate between the two. It is very noteworthy that Mr. Park was so struck by the increase of coarseness at Howick that he described the bed there as “much coarser” than at Parnell. Had he seen the outcrops at St. John's, St. Helier's, and Tamaki Head he would hardly have noticed the increase, so gradual is it; but it becomes noticeable when we compare places far apart. Ponsonby is eight or nine miles from Howick. At a spot between Little Muddy Creek and Avondale I found what appeared to me an outcrop of the Parnell grit. The bed weathered in concretions, and was, as usual near vegetation, leached of its colour, and crumbly. It was much finer than near Auckland, and only 10 ft. thick, and the sandstones with which it was interbedded seemed to be lying on Waitakerei lavas, but this I could not decide. It is so like the usual outcrops of the Parnell grit that I see no strong reason for doubting its identity. If this be accepted, the Parnell grit evidently did not come from the Waitakerei vents; and I think the evidence from coarseness is entirely in favour of its having come from the east, not the west. Lastly, we have the opinion of so excellent a geologist as Sir James Hector, already quoted: “The Parnell grit, as far as I have seen, contains no fragments of the volcanic rocks of the district, but is greensand, with well-rolled pebbles of cherty slate, quartzite, and other Palæozoic rocks, and occasional fragments of old trachyte and basic rocks of Cape Colville.” I might add that the volcanic inclusions are invariably in the form of lapilli, full of steam-vesicles, such as one would expect to have come from a considerable distance.