VI. Nature and Composition of Ejected Matter.

It has been necessary, in speaking of the various craters formed during the eruption, to make frequent reference to the characters of the ejecta from them. It will therefore only be necessary to give here a brief summary of such facts as have already been referred to.

Immense quantities of volcanic ash, bombs, scoriæ, and fragments page 56 of all sizes up to blocks 10ft. in diameter were ejected during the eruption. As regards composition, these belong to two distinct types of volcanic rocks—viz., (1) rhyolites; (2) augite-andesite. All the rocks which occurred in the locality previous to the present eruption were various forms of rhyolitic lavas or tuffs : the augite-andesite which was ejected from the Tarawera Mountain is a rock new to the locality. The rhyolites are rocks of a distinctly acid type, i.e., they are rich in silica, whereas the newer rock is basic.

It appears to be a well-established generalisation that at volcanic vents at which eruptive action is continued through long periods of time the lavas of different chemical composition appear in a definite order. Von Richthofen and other geologists have, indeed, maintained that in all volcanic districts an unvarying order of sequence of the different lavas can be traced. This order is as follows: (1) andesites (and propylites), (2) trachytes, (3) rhyolites, (4) basalts, Other geologists maintain that so many exceptions to the law occur when stringently laid down as to invalidate it. "When, however, the law is stated in the wider form that the first lavas to appear in a volcanic district are those of intermediate composition, and that these are followed by the acid and then the basic lavas, it is found to hold good in a greater number of eases.

The appearance of augite-andesite lava for the first time in the Tarawera district marks an important point in its history. The augite-andesites are of somewhat basic composition, though they are not so highly basic as the basalts. If the new lava from Tarawera be correctly identified as an augite-andesite, we have here augite-andesite succeeding rhyolite, a sequence which docs not follow Von Richthofen's law, but conforms to a certain extent to the wider generalisation : that is to say, we have a basic rock following an acid one.

In any case the appearance of the new kind of lava may be taken to indicate the establishment of a new stage in the volcanic evolution of the district; and if, now that the ancient volcano has roused itself to new life, other eruptions take place from the mountain, we may expect the products to be basic lavas, and not acid rhyolites as in a former period.

1. Rhyolites.—The only rocks occurring on or near the surface in the neighbourhood previous to the eruption were rhyolites. Some of the varieties of rhyolitic lavas occurring here have been mentioned in the account of the geology of Tarawera. Among the volcanic blocks recently ejected many other varieties occur, as, for instance, dark-green and grey perlites of typical structure, and felsitic and porphyritic forms of the most varied hues. Near the Tarawera Chasm huge page 57 masses of a grey perfectly laminated rhyolite lie amongst the débris. This variety is very similar to the lithoidite found in such abundance in the vast pumice-deposits around Lake Taupo; it consists of alternate laminæ of a light and dark colour, these often testifying by their curved and contorted forms to the unequal flow in the lava as it

It has been commonly stated that fresh obsidian bombs and pumice were ejected from Tarawera during the eruption^* As it would be somewhat remarkable to find two distinct forms of fresh lava ejected simultaneously from the same volcano, it may be well to give here the evidence bearing on the point. Previous to the eruption obsidian bombs, often of very perfect form, were known to exist on Haroharo, a mountain to the south of Rotoiti. The writer, when visiting Rotoiti a month after the eruption, was shown one of these by Mr. Fraser, who had obtained it long before the eruption. Next day a visit to the further end of the lake was made by boat, but no bombs of the kind were noticed. A few days later heavy rain fell, and owing to the nature of the volcanic ash the water ran off rapidly, scouring out the watercourses, and in many places cutting deeply into the surface-rocks. After these floods obsidian bombs were found floating at the farther end of Rotoiti, where the watercourses come down from Haroharo. It is clear, therefore, that the bombs must have been derived from the latter mountain, and not from the ejecta of the Tarawera eruption. This is confirmed by the absence of such bombs from the slopes of Tarawera and other parts of the ash-fields. Fresh pumice is also said to have been found. Pumice-fragments, in very small quantity, are to be found in various parts of the area covered with ash, but even when of somewhat fresh appearance there is no reason to believe that they are of recent formation. Occasionally they are coated with patches of black lava; these fragments are probably derived from the pumice-strata exposed in the gap between Wahanga and Ruawahia. It will be seen, therefore, that the statement that two kinds of fresh lava were ejected is without foundation.

2. Augite-andesite.—The molten rock which rose to the surface the fissure on the mountain was so highly charged with steam that, instead of overflowing the fissure and giving rise to a lava-stream, it was ejected explosively from the crater, chiefly in the form of porous scoria; or fine sand and dust. From the Tarawera Chasm, however, numerous bombs were ejected : these are much less porous than the scoriæ, and form round black masses lin. to 8in. in diameter,

^* See also appendix to "Report of Tarawera Eruption," by Mr. S. Percy Smith.

page 58 with cracked surfaces. They never show the elongated or spindle-shaped forms with twisted ends which are so common amongst the bombs of the Auckland basaltic cones. The scoriæ and dust are widely distributed over the whole country affected by the eruption, whilst near the fissure larger and more solid blocks of the lava are found.

The rock has been identified as an augite-andesite; but in some respects it approximates to the basalts. The solid portions have a black and slightly resinous appearance, and almost homogeneous texture. The specific gravity is 2.93. Under the microscope thin slices of the rock show rather numerous small crystals of felspar and yellowish augite. The felspars are mostly small and in ledge-sham sections, but a few rather larger ones are present; they are nearly all plagioclases, but there are a few distinct sanidines. The augite is in small irregular crystals and granules. The ground-mass of the rock is a grey glass with abundant crystallites (globulites, longulites, and microliths) of translucent substance, and grains of magnetite. Much of the augite should perhaps be considered as belonging to the ground-mass.

Other varieties of the rock differ chiefly in containing a larger or smaller proportion of glass, and in the size and abundance of the augite-crystals. Some specimens have a large proportion of brownish glass. Olivine was found in a few small crystals in some specimens, but as a rule it is quite absent. Chemical tests also fail to show its presence. It cannot be regarded as an essential constituent.

On account of the absence of olivine as an essential constituent, the presence of abundant glass with microliths, and of a little sanidine, the rock has been identified as an augite-andesite. Still, it must be admitted that the rock approximates to the basalts, as is shown further by its chemical analysis. For this, as well as for analyses of the ash, I am indebted to Mr. J. A. Pond, Colonial Analyst. In an Appendix will be found the analyses of lapilli of this rock from various parts of the ash deposit. The percentage of silica varies from 50.9 to 52.5, an amount which is lower than that usually found in augite-andesites, whilst it is not higher than that found in some basalts.^* But whether we call the rock an augite-andesite or a basalt without olivine^† docs not affect the results stated above, that

^* The silica in the basalts is said to vary from 45 to 55 per cent. (Geikie), or from 36.68 to 53.76 per cent. (Green : "Physical Geology"). The average amount of silica in augite-andesite is 57.15 per cent.

^† Messrs. Hague and Iddings ("American Journal of Science," Vol. xxvii., p. 456) state that in the Western States various transitional forms from basalt into augite-andesite are to be found.

page 59 the lava of the late eruption is basic, is new to the locality, and follows the acid rhyolites.

It should be mentioned that the black augite-andesite is frequently filled with light-coloured fragments of the rhyolites. This is not surprising when we consider that the molten lava must have forced its way between rhyolitic rocks. Some of these included fragments are large pieces of unweathered lava, as large as a man's head, and of the same variety as that which forms the top of Tarawera Mountain; other fragments are of all sizes down to small splinters and grains no larger than a pin's head. Sometimes these fragments are fused, and more or less completely blended with the rock that contains them. To this origin we must attribute some crystals of quartz and weathered sanidine which were detected in a specimen of the black lava.

Volcanic Ash.—After what has been already said on this subject, it is only necessary to speak of the general characters of the ash, more especially as found at a distance from the fissure. The ash consists everywhere of a mixture of rhyolite and augite-andesite, reduced to a more or less finely divided state. In tracing the ash to its origin from various parts of the fissure, we must remember that from the part across the mountain the ejecta consisted mainly of augite-andesite, but also partly of rhyolite, whilst from the rest of the fissure along the lower ground the ejecta were all rhyolitic. The latter included, however, not only fragments of rhyolitic lava, but also all Kinds of rhyolitic débris, older ash, sands, partly waterworn fragments from lake-beds, compact tuff of still older date, fumarole-clays formed by the energetic hydrothermal action which had been going on so long around Rotomahana, mud from the bottom of Rotomahana Lake, sinter, and the soil and decomposed surface-rock (an appreciable item over so wide an area of disturbed ground).

Though the ash is everywhere a mixture of andesite and rhyolite, yet in certain parts either the one or the other largely predominates, and the colour accordingly varies between light-grey or nearly white and a dark-grey or black. The ash shows a good deal of variation in the size of its component particles, even at points situated at equal distances from the fissure. It almost always contains a quantity of fine impalpable dust, which envelops the coarser particles, and may so completely hide their character as to give a misleading impression of the composition of the ash.

It was therefore necessary to employ some method of mechanical separation of the particles of different degrees of fineness, so as to permit a comparison between the ashes from different places. The finest dust was separated by washing with water, and then the coarser page 60 particles separated by sieves. The portion which passed through a sieve of thirty meshes to the inch may be termed sand; the coarser part remaining behind may be termed grit, or, when coarser than peas it should be called lapilli.

The coarser particles can be identified by the naked eye; the finer ones require the aid of the microscope. In this way we find that the ash is composed of portions of augite-andesite lava of all sizes, from an inch in diameter down to the finest flour-like powder, and similar particles of rhyolites, including a small proportion of pumice. The ash contains the characteristic minerals of the rocks, separated from one another. The chief minerals recognisable are quartz and felspar (both sanidine and plagioclase); and less abundant are black and yellow mica, magnetite, and hornblende. Augite, iron-pyrites, and hypersthene were also identified, but only in occasional fragments. Volcanic glass, chiefly derived from the glassy rhyolites, formed a considerable proportion of the colourless fragments.

Two shells of a water-snail were found, one in a sample of ash from Tauranga, the other in the ash from Cape Runaway. These were probably blown out of Rotomahana.

The violence of the explosion was greatest on the mountain the column of steam laden with ash must have risen highest there We find therefore that at a distance from Tarawera the coarsest particles of the ash are fine scoriæ or lapilli of the augite-andesite. Sometimes by far the greatest part of the ash is composed of andesite. The fine dust which forms part of the ash has a varying composition: At the Wairoa it forms over 40 per cent. of the ash; it is white and earthy, but is not plastic, and can scarcely be called clay; it may contain a small proportion of fumarole-clays, but consists more largely of very finely divided rhyolite and andesite (chiefly the former). At Cape Runaway it had a rather dark-grey colour, and was chiefly composed of andesite.

One remarkable fact shown by the eruption is that, apart from the explosion of lava, the explosive action of steam rising along the fissure had the power of reducing the surface-rocks of moderate hardness to a very fine state of division. The fine flour-like dust which lies on the surface around the southern craters consists largely rhyolitic tuff reduced to dust, and must have been produced in this way by what may be termed hydrothermal explosion only.