V. Nature of the Vents Formed in the Eruption.

The Great Fissure.—When the scene of the Tarawera eruption is examined it is found that a series of vents was formed extending along a line over nine miles in length, reaching from the north-east slope of Wahanga, through the heart of Ruawahia and Tarawera, down the side of the latter to Rotomahana and beyond, ending at a point about 600yds. to the north of Okaro Lake. This line has a mean direction nearly N.E. by E., or, more exactly, 58° E. of N., whilst that of the Taupo zone is 38° E. of N. It is not perfectly straight, but its deviations from a straight line are very slight.

The vents or craters along this line have been formed around portions of a great fissure which traversed the rocks of the earth's J crust. High-pressure steam escaping from below blew out the rocks forming the walls of the fissure, and thus produced crater-like hollows. Had the force of the explosions been uniform along the line, and the rocks of equal strength to resist the force of the escaping steam, the result would no doubt have been the formation of a deep narrow trench extending along the whole line. But the rocks differed in their power of resistance, and the supply of steam and its facility of escape varied at the different points. The eruption was further complicated by the rise of molten rock, saturated with steam, to the surface along the northern part of the fissure, and by the existence around Rotomahana of rocks which had long been subjected to the action of heated water rising from below.

As soon as the fissure was formed the high-pressure steam from below would tend to escape upwards. At certain points it would find a more ready exit, and, expanding explosively as it issued from the fissure, would blow out or tear away the surrounding rocks, thus forming a series of craters, separated by wider or narrower strips of ground which escaped the explosions. Such areas of ground are seen along the great Tarawera fissure, some of them extremely narrow, so that one cannot help wondering how they can have escaped the explosions, for they appear like mere walls dividing a crater into two; others are wider. The most considerable of the bridges is the one at the south-west end of Tarawera, where the former surface of the ground was left undisturbed for the distance of a quarter of a mile.

That the fissure actually passed through these bridges, which thus escaped simply by the more ready passage of the steam elsewhere, is page 41 shown by the narrow vertical cracks or fissures to be traced in most of the walls of rock which part successive craters. We may see here, moreover, that the line of fissure was not a line of fault, for the rocks on both sides of the fissure correspond exactly, there being no lowering of the rock on either side.

The materials ejected from the various craters naturally accumulated to the greatest thickness around their margins, and the size and depth of the craters were thus added to by the rock-débris piled up around them. The craters therefore are of double origin. They are, in the first place, hollowed out of the ground by explosions—i.e., they are primarily explosion-craters—and, secondly, the upper portions are built up around the vents. The relative shares of these two causes in the formation of the craters of the present eruption are by no means the same in all cases. In some, one-half of the depth of the crater is due to the building-up of debris around its edge; but in most cases a smaller proportion of the depth is due to this cause—sometimes much less, as in some of the craters towards the southwest of the line, where less than one-tenth of the total depth is so formed.

1. Craters on Wahanga, Ruawahia, and Tarawera.—These occupy a line rather more than four miles in length, reaching from a point on the north-east slope of Wahanga, 900ft. above Lake Tarawera, to the foot of Tarawera Mountain, 260ft. above the lake. The only considerable interruption of the line of craters is the bridge at the southwest margin of the flat top of Tarawera, having a breadth of a quarter of a mile (see map of the fissure). To the west of this the fissure down the side of Tarawera has received the name of "the Tarawera Chasm" whilst to the east it stretches across the mountain in the form of a deep trench with steep sides, divided by narrow partitions, here the rock of the mountain remains in position, into a series of oblong craters.

This section of the fissure is distinguished from the other by the circumstance that fresh augite-andesite scoriæ were ejected from its whole length, except for a short distance at each end. The lava which rose from below was so highly charged with superheated steam page 42 that it did not escape as a lava-stream, but was hurled out explosively, and is piled up in layers of black and red scoria; around the fissure.

That the lava welled up in the fissure in a molten state is abundantly shown not only by the vast quantities of scoriæ produced, but by various other facts. Bombs of the new lava are to he found abundantly in many places : these have assumed a spherical form whilst cooling from the molten condition in their passage through the air. At the bottom of one of the craters I found large masses of fresh black lava containing as much as 1,000 cubic feet. The rocky walls had been splashed with the molten lava, which had moulded itself in cindery, slaggy forms around the projecting rocks. Along the upper margin of the Tarawera Chasm the same moulding of the lava around projecting rocks was observed, showing that the ejected lava had fallen here whilst still in the state of fusion.

In the part of the fissure extending from its beginning on the slope of Wahanga as far as the Tarawera Chasm we may count twelve craters, separated by partitions of varying height. The first two craters are comparatively small and of no great depth; the others are long and deep, varying in depth from 300ft. to 800ft., and in width from 150yds. to 300yds., the deepest crater being that crossing the gap between Wahanga and Ruawahia. The sides of the craters are almost without exception so steep, in places being quite vertical, that it is not possible to descend into them. The only craters which I was able to get down into were the two nearest the south end of Tarawera. An aneroid reading at the bottom of the second of these showed it to be 450ft. below the highest point of its margin to the north-east. At this time steam was issuing from numerous points in the fissured rocks at the bottom.

The first crater begins low down on the slope of Wahanga. It is a long and narrow fissure, not more than 80ft. deep. The most interesting feature it presents is that it cuts across a small hill, or, rather, a spur of Wahanga, which had been divided by a gully worn by water descending from the mountain. The fissure shows a total disregard of the surface-features of the ground, cuts into the highest part of the spur, and crosses the gully at right angles. The spur consists of a lava-stream, covered above with a rhyolitic breccia. The end of the crater away from the mountain shows around its margin principally fragments of the old light-coloured rhyolites, hut nearer the mountain these become more largely mixed with black scoriæ, and then with red. Apparently the fresh lava was only thrown out from the part of this crater nearest the mountain.

The next crater is also of relatively small size, but shows far more scoriæ around its edges. The third and fourth craters are much larger and deeper, and are comparatively narrow. They lie immediately under the south-east wall-like face of Wahanga, part of which has been blown away. The margin of the craters to the north-west is formed by the flat top of Wahanga itself (3,500ft. in height), and from it there is a sheer precipitous descent of over 1,500ft. to the bottom of the craters. The opposite margin is necessarily much lower. The former wall is composed of the massive rhyolite lava of the mountain; in the opposite wall rhyolite breccias are to be seen.

The line of craters crosses the gap between Wahanga and Ruawahia. and, striking the north-east angle of the latter, cuts across the centre of Ruawahia and Tarawera. Three craters may be counted occupying the gap between the summits, or they may be regarded as the three divisions of one crater, for the partitions are very narrow, a mere knife-edge of the original surface being left, too narrow to support any thickness of débris, whereas the other margins of the craters are raised by a hundred feet or so of scoria; and other fragments of ejected rock. The craters are of great depth; the one nearest Ruawahia stretches down 800ft. below the gap. The southern part of this crater is excavated out of the side of Ruawahia, which just above reaches its greatest elevation of 3,770ft. From this point down to the bottom of the crater there is therefore a precipitous fall of 1,500ft.

The walls of the fissure here show very interesting sections, quite different in character from those seen elsewhere. Looking at the side of Ruawahia exposed at the end of a crater, we sec, underlying the ash and scoriæ recently deposited, first a flat cap of massive rhyolite lava, and below that again inclined beds of white rhyolitic ash or pumice, which form a striking contrast with the dark-red or black scoriæ. The pumice-beds show a very complicated arrangement, showing that the section has been taken across the rim of a cone and crater composed of pumice and other light-coloured rhyolites. Just under Ruawahia the strata in the centre dipped inwards, evidently towards the hollow of a former crater. At the sides the strata showed an inward dip towards the centre of the section, and then an outward dip away from the mountain. In the sides of the fissure stretching towards Wahanga a dip away from Ruawahia could be readily observed.

The line of craters is continued through the middle of the plateau-top of the mountain as far as the bridge above the "Chasm;" the depth these craters is greatest (600ft.) towards the north end of Rua- page 44 wahia, and diminishes gradually towards the opposite end. The craters are hollowed out of the rhyolite lava forming the top of the mountain: the sides are very steep here, at times quite vertical. On top of the old lava a small amount of rhyolitic breccias is observed filling up the hollows between the peaks which marked the top of the mountain. Above these is deposited a great thickness of scoriae from the recent eruption, varying from 80ft. to 200ft., the greatest thickness being found near the north-east part of Ruawahia. The upper parts of the crater-walls, where they are formed by the newly deposited and incoherent materials, are naturally less steep than the lower parts where the fissure passes through the massive rhyolites.

The flat top of the mountain was, as already described, marked by smaller peaks or hillocks, and was therefore far from being even The newly-deposited matter, also, has not fallen regularly, but has formed along certain parts of the margin of the fissure deposits at least twice as thick as those along neighbouring parts. The margin of the fissure is therefore marked by undulating beds of scoriæ, which rise into low conical hills in which the layers slope outwards. The outline of the mountain, as viewed from a distance, has also undergone a change; it shows more conspicuous peaks than before, and is somewhat higher. The highest point of the mountain is at the north-east end of Ruawahia, and just on the west side of the fissure: this point, according to Mr. Percy Smith's observations, is 3,770ft. above the sea. The mountain was formerly 3,606ft. high, thus showing a gain of 164ft. The gain is naturally greatest near the fissure; farther away the deposit of ashes is less thick, and the height of the mountain less affected, so that the surface shows a slope away from the fissure on each side.

The freshly-deposited material on the mountain consists most largely of scoria; of augite-andesite, which give to the new strata, as seen in section along the margin of the fissure, a prevailing colour of dark-red, so that in the sides of the fissure the line between the old and new rock is very obvious. (See Pl. p. 44.) Amongst these darker scoria; are fragments of the older light-coloured rock, and towards the surface they are often very abundant. The surface of the mountain along the bridge above the Tarawera Chasm is remarkable for the profusion of angular blocks of rhyolite with which the ground is strewn. Most of these are about a foot in diameter, but: many of them are very much larger.

Near the edge of the fissure the augite-andesite scoria; are red in colour, whilst farther away they are black. This change of colour from black to red is due to the action of various volcanic gases and page break

Tarawera Fissure.

Shewing old surface of Mountain.

Looking South

page break page 45 air on the scoriæ whilst still hot, and may be imitated in the laboratory. If the black scoriæ be heated to a temperature below red-heat in a tube through which a current of air and steam is passed, the scoriæ slowly turn red, owing to the oxidation of the iron present in them, and its conversion into ferric oxide. If hydrochloric-acid gas be added to the air and steam passing over the scoriæ, the change of colour takes place much more rapidly, whilst parts of the scoriæ are attacked still more strongly, and become a bright yellow or orange-red colour, from a stain of perchloride of iron formed by the action of tie hydrochloric acid on the oxides of iron.

A precisely similar action on a larger scale took place amongst the lot scoriæ which accumulated on the margins of the craters. The steam, which escaped abundantly from the fissured rocks on the mountain after the eruption, was mixed with hydrochloric-acid gas, so that walking near the fissure was often rendered difficult owing to the choking character of the fumes rising from the ground. During the month of July the ground along the sides of the Tarawera Chasm appeared yellowish-green, as viewed from a distance, owing to the presence of large quantities of perchloride of iron. This yellowish tinge was popularly attributed to sulphur, as is often the case. After heavy rain the colour disappeared, the perchloride of iron being removed in solution by the rain-water. The beds of scoriæ along the margin of the Tarawera Chasm presented a most brilliant appearance, as their prevailing tint of dark-red or chocolate was marked by bands of bright orange-red or yellow where the perchloride of iron had gathered.

The ash near the margin of the fissure on the mountain seems to have undergone a very large amount of chemical change. The ground has in many places been traversed by hundreds of cracks, usually parallel to the margins of the craters, though sometimes at right angles to them. Along these cracks the acid vapours have escaped, and have acted upon the rocks. The perchloride of iron and other soluble products formed have been removed by the rain, and an earthy residue of bleached and decomposed rock left behind. The fissures, which began as narrow cracks, have increased in width and depth until some of them are hundreds of yards in length and 10ft. or more in breadth, with a depth of 5ft., 10ft., or more.

It should further be mentioned that the loosely-compacted materials which accumulated around the craters are gradually slipping down into them, so that already long talus-slopes of the fallen débris have collected below. By this constant fall of material into the craters they are naturally becoming filled up, and the depth greatly reduced. page 46 One of the difficulties in the way of exploring the bottom of the fissure on the mountain (so far as it is accessible) consists in the possibility of the sudden fall of an avalanche of loose scoriæ from above.

After what has been said concerning the fissure on top of the mountain, it will not be necessary to devote much time to a description of the Tarawera Chasm. The Chasm consists of a portion of the fissure about a mile and a quarter in length, stretching from the edge of the plateau down the side of the mountain to its foot. (See Pl. p. 46.) The fissure is continuous the whole way down, but varies in width and depth: the bottom is marked by various crater-like hollows, which are doubtless the points whence the discharge of steam or lava was most energetic. The uppermost part of the fissure is the most considerable, a wide and deep crater having been formed here. At the back of the crater its wall is formed by the vertical rocks, rising up more than 1,000ft. to the top of Tarawera; at the sides the height of the crater walls diminishes. The amount of scoria; deposited around the upper part of the Chasm shows that it must have been very active, and it was doubtless the more ready escape of the volcanic forces along the side of the mountain which allowed the bridge of ground above to remain undisturbed.

Further down the Chasm becomes narrower and less deep, and ends at the foot of the mountain in a small crater, which when I first saw it, in July, 1886, contained a small lake. Later on this disappeared, and examination of the bottom shows that in the massive rhyolite forming the wall at the back a fissure exists which penetrates the hill, and is probably a portion of the main fissure. The rocks are in contact above, but diverge below, so that it is possible to advance a considerable distance into the mountain along the fissure.

Part of the site of the Tarawera Chasm was occupied by large trees, and during the eruption these, like the rocks, were blown up. Large stumps of trees found near Rotomahana are supposed to have come from this place, as no trees of so large a size grew on the other parts of the disturbed ground.

2. Craters from the Foot of Tarawera to Rotomahana Crater.— The character of the eruption along this and the remaining section of the fissure was different from that on the mountain. Here, on what we may call the lower country, no molten lava rose to the surface, and there was therefore no ejection of fresh scoriæ or bombs. The matter ejected consisted of the old rocks of the locality fractured and reduced to ash by the violence of the steam which escaped in such abundance. The eruption on the lower ground may therefore be page break

The Chasm

South-West end of Tarawera.

page 47 spoken of as a hydrothermal eruption, whereas that from the mountain was a true volcanic eruption. The terms "hydrothermal" and "volcanic" are used here in a very narrow and restricted sense, for most volcanic phenomena are also strictly hydrothermal, the immediate cause of volcanic eruptions being generally admitted to be the presence of large quantities of water and gases in masses of heated rocks. Whether the heat be sufficient to melt the rock or not is relatively a matter of minor importance. It is, however, both usual and convenient to distinguish between these two phases of volcanic activity.

The question which next suggests itself is, What was the source of the high-pressure steam, the active agent in the ejection of material from the remainder of the fissure? Were the rocks along the fissure at no very great depth from the surface charged with water and at a high temperature, and so ready to explode as soon as the opening of the fissure occurred? Or was the steam derived from a greater depth Rom the surface? Is it possible that the deeper part of the fissure across the lower ground was injected with molten rock just as the fissure along the mountain was, although the lava did not appear at the surface?

It would be interesting if we could know why the lava rose in the fissure along the mountain, where there was the additional resistance of half a mile of rock to overcome, rather than along the plain. It is scarcely a sufficient answer to say that the source whence this lava was derived lay at such a depth below the mountain that the additional thickness of half a mile of rock made no appreciable difference, for the fissure extended through the plain as well as through the mountain. A more satisfactory explanation may perhaps be found the supposition that the source of the lava lay immediately under the mountain, and that the old lavas of the mountain, notwithstanding their apparently firm and massive character, really offered, by reason of their abundant joints, comparatively little resistance.

The fissure extended to the distance of five miles beyond the mountain, and there is evidence that the whole of this fissure was of Beep-seated origin. It therefore seems somewhat remarkable that no lava appeared in the fissure along the low ground, and we are led to inquire whether there is anything in the character of the country here which may have prevented the rise of lava up the fissure, supposing it to have been injected into the fissure at a greater depth. The only agent which can be supposed to have obstructed the rise of molten rock is the large supply of water available on the plain. Not only was there Rotomahana, but also Rotomakariri, and other lakelets, as well as a considerable supply of underground water in the porous page 48 strata. What the behaviour of molten rock in a half-opened fissure would be, when the surface waters were let down upon it, it is impossible to say with any certainty, but the result would doubtless be the explosive conversion of the water into steam, and the consequent cooling of the lava. Whether any fragments of the fractured lava would be thrown up to and reach the surface would depend upon the depth in the fissure at which the explosions occurred. The occurrence of the fragments of the fresh lava amongst the ejecta from the fissure would then furnish some evidence of the nature of the action which had taken place.

Fragments of the augite-andesite lava occur everywhere amongst the ejecta from the portion of the fissure along the lower country; but it is certain that a large proportion of these, and perhaps all, came from the mountain, and so the evidence is rendered valueless, It may be mentioned here as bearing upon this point that near some of the southern craters, more especially the Black and Echo Craters, large quantities of augite-andesite scoriæ occur; and it appeared to me at first probable that they had been ejected from this part. If it could be shown that the molten lava rose to the surface not only along the fissure in the mountain, but also at certain points near the south end of the line, the probability of its injection into the fissure along the whole extent would be greatly strengthened. Further close examination of the neighbourhood of these craters does not strengthen the impression, but seems rather to show that the fresh scoriæ found near them came from the mountain. Although such an injection of molten lava into the fissure south of the mountain cannot be proved, it is by no means improbable, and would explain some of the characters of the fissure. We must, however, look to another circumstance to explain the great size of the Rotomahana Crater. Around Rotomahana the ground everywhere had for centuries shown signs of vigorous hydrothermal activity. There can be little doubt that at moderate depths from the surface the rocks were saturated with water at a temperature far above its ordinary boiling-point, and that this water was simply kept from flashing into steam by the pressure of the overlying rock. If that pressure could have been relieved by the removal of the surface-layers of the ground the superheated water would have been explosively converted into steam. There were present, therefore, around Rotomahana all the conditions requisite for a hydrothermal explosion except the relief of the pressure due to the overlying rocks. The formation of the fissure during the eruption supplied the last necessary condition, and the result was therefore the stupendous hydrothermal explosion of page 49 June, 1886. The large area of ground around Rotomahana which had long been subject to hydrothermal action explains satisfactorily why the width of the crater is so much greater there.

But between Rotomahana and the mountain on the one side and to the south-west of Rotomahana on the other there are large sections of the fissure which passed through ground not known to have shown any signs of hydrothermal activity. Even in the sections afforded in the sides of the craters any alteration of the strata by hydrothermal action is generally wanting. Here therefore it must have been entirely the force of the steam rising from a deeper part of the fissure which caused the ejection of the surface-rocks. Along the central part of the Rotomahana Crater, where the fissure crosses it, the explosive force of steam rising in the fissure would be added to that already present in the surface-rocks, and hence it was that the crater was excavated to the greatest depth along the line of fissure, whilst away on each side the depth is so much less.

Resuming the description of the Great Fissure, we find below the Tarawera Chasm a bridge of undisturbed rocks about 100yds. in width. Then just under the slope of Tarawera is a circular crater about 250ft. deep, the bottom of which is occupied by a green lake. Close to the south-east side of this is another distinct but smaller crater. This occupies part of the site of a small isolated hill of rhyolitic lava which stood up from the plain at a little distance away from the slope of Tarawera. From the resemblance which the shape of this hill bore to the mountain above, it had been popularly called "the Baby Volcano," and, if we may judge from a section of what remains of it as exhibited in the wall of one of the craters, it had a mode of formation analogous to that of Tarawera.

A partition of no great thickness of rock divides the Green-Lake Crater from what is practically one continuous crater stretching from this point to the foot of Te Hape-o-Toroa, and including the site of the former Lake Rotomahana. This has a varying width and a length of a little over three miles, whilst its area is equal to three square miles. From its southern margin a hill of hard rhyolitic rocks known as "the Banded Hill" projects into the area, and partly divides it into two areas the one to the east may be termed the Roto-makariri Crater, whilst that to the west naturally receives the name of Rotomahana Crater. (See Map II.)

The walls of the Green-Lake Crater are chiefly formed of hard rhyolite lavas, but at the beginning of the Rotomakariri Crater there is a change to softer rocks—the horizontal beds of pumice and rhyolitic sands and breccias, partly lake-beds, described in a former part of page 50 this report. Owing to the soft and incoherent nature of these strata the volcanic forces have been able to blow out a greater quantity of material, and the fissure at once assumes a greater width. For the distance of a mile the fissure has a mean width of a third of a mile, and at its bottom contains a lake which has been called Hotomakariri, though not occupying the site of the old lake of that name. (See Pl. p. 46.) The sides of this part of the crater are 250ft. to 300ft. high. To the east there is a cliff with nearly vertical face, whilst on the opposite side the slope is more gradual. For a few days after the eruption large clouds of steam were seen to rise from the position of this lake; but by the middle of July the activity had almost disappeared, and only a few curls of steam were then rising from its margin. On the eastern side the upper 40ft. to 60ft. of the cliff are formed of grey volcanic ash, containing, however, a considerable quantity of black scoriæ from the mountain. Away from the margin of the fissure the thickness of the ash decreases rather rapidly, so that at the distance of 500yds. the tops of the tea-tree stems may be seen projecting from the ash. On the western side of the fissure the ash reaches a much greater thickness. The level of the water in the new Hotomakariri two month after the eruption was 983ft. above the sea, or 57ft. below the level of Lake Tarawera.

At the cud of the new lake the fissure suddenly widens, and its bottom forms an uneven plain, bounded to the south-east by nearly vertical cliffs 250ft. in height, and to the south-west separated from Rotomahana Crater by the Banded Hill. To the north-west its boundary is formed by the Star Hill, the slopes of which are excavated by craters now extinct, but which poured forth volumes of steam for many weeks after the eruption. The flat is traversed by numerous broad watercourses, which have spread large quantities of ash over it. Small streams of water flow from the foot of the cliffs, The water of two of these is cold and pure; but that of a third which issues from the cliff nearer the mountain was warm, and strongly charged with salts of iron, oxide of iron being deposited copiously along its channel. It probably has its source in the rain-water which drains from the mountain, and the iron will be derived from the perchlorides, &c., dissolved out of the scoriæ by the rain.

Towards the centre of the flat is a crater-hollow, containing a lake. During the month of July steam was rising from the margin of this crater; but otherwise there was no activity left on the flat until a point nearer the Star Hill was reached. Between the latter hill and the Banded Hill is a narrower portion of the fissure leading to Rotomahana Crater. Here two crater-lakes, with steep sides, page 51 mark the direct line of fissure. They occupy the site of the old Rotomakariri (indicated in Map II. by dotted lines).

Near the top of the Star Hill, and close to the margin of the fissure, is a detached crater some 200ft. deep. This emitted much steam for many weeks after the eruption, and even during the following summer the water at the bottom was boiling at one side. About half-way down the steep sides of the crater the old land-surface could be distinguished, and close to this was a patch of green ferns—an unexpected and remarkable sight in a crater which had been so active, especially as no other trace of vegetable life was to be seen for miles around in the waste of steaming water and volcanic ash.

The Rotomahana Crater has a length of one mile and a half in the direction of the Great Fissure, and a breadth which somewhat exceeds this amount; whilst its total area, as estimated by Mr. Percy Smith, is slightly over two square miles. Its sides are steep, and some 200ft. to 800ft. high. In certain parts, as, for instance, the southern bay, there is a vertical cliff; but elsewhere the ash around the crater has slipped down, and lies at an angle of about 30°. Below this steeper part the bottom of the crater gradually slopes down to the central hollow. Shortly after the eruption the greater part of the floor of the crater was covered with mud-volcanoes, "mud-geysers," and fumaroles, which gave off dense clouds of steam, rising high into the air. At the beginning of July I noticed one particular line of ten or twelve mud-geysers near the deepest part of the crater which were throwing up fountains of liquid mud. The position of the line was approximately along the middle line of the Great Fissure, Towards the end of the month a lake of hot muddy water had collected in the deepest part of the crater, just along the line of fissure, and crossing the former site of Rotomahana Lake, as will be seen in Map II., where the former position of the lake and terraces is shown by broken lines. The level of the new lake was 565ft. above the sea, or about 500ft. below the old level of Rotomahana.

The crater receives the drainage from a considerable area of the surrounding country, and this, rushing down the watercourses, carries with it large quantities of mud and stones. The mud and stones at the edges of the crater have slipped down, and so the level of the crater-floor has been raised, and the size of the lake in the centre increased. By the following summer the lake had grown to a size at least equal to that of the original Rotomahana. As time goes on, the lake, having no outlet, will doubtless increase still further in size, and may even extend its area so as to form one continuous lake as far as the end of Rotomakariri. If, however, the level ever rises page 52 high enough to overflow the barrier of a hundred feet or so of mud and ashes with which the Kaiwaka Valley—the former outlet of Rotomahana—is blocked up, the great fall down to Tarawera Lake will give the water such excavating power that a few hours will probably see a deep channel hollowed out of the soft ash, and the relative levels of the two lakes again established.

The sites of the Pink and White Terraces, marvels of natural beauty, which attracted visitors from all parts of the world, have been identified with sufficient closeness by the surveys of Mr. Percy Smith, and it is shown that the ground around them and under them has been blown out for several hundred feet. Fragments of the sinter of which they were composed have been found amongst the débris around. The White Terrace was situated a little to the north of some bold, jagged rocks of massive rhyolite which stand up inside the edge of the Rotomahana Crater, and have received the name of Pinnacle Rocks from their peculiar form. These rocks were greatly fissured, and even in the summer after the eruption much steam issued from the cracks. The two plates (p. 52) show Rotomahana before and after the eruption; they are from photographs taken from the same point of view.

On the opposite side of the fissure the Banded Hill is also composed of massive rhyolite. It is the remains of the hills which lav to the east of Rotomahana Lake, and the whole of the western face has been blown off. Six weeks after the eruption, when the steam arising from Rotomahana had abated sufficiently to allow one to sec into the crater, the side of this hill presented a remarkable and very beautiful sight, especially if viewed from the point projecting into Rotomahana on the south-west. Its face was greatly fissured, and from the cracks there issued thousands of jets of high-pressure steam, like the steam escaping from the safety-valve of a steam-engine, whilst the roar of the out-rushing steam reminded one of the sound of a busy railway-station multiplied a thousandfold.

The southern portion of Rotomahana crater differs a little from the other parts. Here the crater is bounded by high cliffs of horizontal strata, composed of volcanic sandstone and breccias, whilst, the floor of the crater shows an extraordinary confusion of huge blocks of brown sandstone, tumbled about in the wildest disorder: between them is the common deposit of grey ashes. When the writer passed across this place six weeks after the eruption, a river of boiling water was found rushing out from beneath a heap of such blocks of sand stone. Here and there the blocks are piled up around funnel-shaped hollows, which mark the spots where the steam that shattered the rocks escaped.

White Terrace and Lake Rotomahana

Rotomahana

Where the White Terrace stood.

The Fissure from Top of Black Crater.

Looking-Across Rotomahana to Tarawera in a North-East Direction.

Crater North of Black Crater

shewing how fissure has cut through the spurs of Hape-o-Toroa Hill.

During the summer after the eruption the activity in the Rotomahana Crater had greatly subsided. In many places the rise of steam was choked by the accumulation of mud brought down from the sides of the crater. The parts which remained most active were near the Pinnacle Rocks, and between these and the Hot Lake, and some very active fumaroles between the former site of the Pink Terrace and Te Hape-o-Toroa.

The deposit around Hotomahana consists chiefly of the surface-rocks reduced to the state of dust and sand. Larger fragments and blocks of all sizes occur, mixed with the finer matter, as well as scoriæ from the mountain. The deposit when moistened by the rain formed a mud, which made walking very difficult; but, on drying, it set to form a mudstone or kind of soft concrete.

3. Smaller Craters to the South-west.—Under Te Hape-o-Toroa the Rotomahana Crater narrows, and soon ends abruptly. The fissure is continued as a scries of distinct and separate craters, all of which are of relatively small size. In the middle of July, when first examined them, they were ten in number, and occupied a length of more than one mile and three-quarters. Shortly afterwards, however, the three small ones nearest to Rotomahana Crater were destroyed by the water draining in that direction. This part of the fissure corresponds to the lower part of a valley along which the Haumi Stream formerly flowed on its way to Rotomahana, and is now occupied by a watercourse with vertical sides which reach the right of 200ft.

One of the most interesting points in this section of the fissure is its independence of the surface-features of the country. The seventh crater of the series, counting from the south (see Map II.), lies at the bottom of the Haumi Valley, and it might have seemed most natural for the fissure to have continued its course up the valley, which is nearly straight. Instead of doing so, however, it (in the ft sixth crater) strikes up the side of a spur from Te Hape-o-Toroa (see Pl. 9), p. 53), and continues along the high ground a course parallel with the valley close below. The next two craters (the Black and No. 4 Craters) are situated on the top of the spur with the ground sloping down on either side; whilst the third crater is blown out of the side of the spur towards its end. Further on the fissure still shows a disregard of surface-features, for, whilst the Echo Crater Begins at one end in a valley, its further end is on top of a hill. The Southern Crater, the last of the chain and the termination of the fissure so far as can be seen, is likewise on top of a hill.

These craters have been drilled by the force of the explosions out page 54 of the surface-rocks. They have very steep and sometimes vertical sides; their depth is in parts over 250ft. The rocks which form the walls of these craters from No. 3 to No. 6 are chiefly rhyolite lavas but in part they are coarse volcanic breccias. In the Echo and Southern Craters, however, the rock is of a very distinct character: it is a compact rhyolitic tuff, usually white or buff in colour.

Around the last two craters large areas of the ground are covered with very large angular blocks of this tuff. These are particularly numerous in certain places, as if the fire of blocks had been directed obliquely towards them. As they lie quite on the surface they must have been ejected after the finer ash had ceased to fall. The hills in this neighbourhood were covered with volcanic ash as dry and fine as flour, of a greyish-cream colour. The surface of the ash was covered with hundreds of shallow conical pits, ranging from 1ft. to 30ft. in diameter. These marked the spots where blocks hurled out of the southern craters had fallen back to the ground. Often the blocks had fallen with such force that they were buried in the ash; in other places the blocks could be seen projecting from the centre of the hollow. One of the largest of these, 500yds. away from Echo Crater lay in a hollow 27ft. in diameter; it was 5ft. in length, and must have weighed over seven tons. The distance which so heavy a block had been thrown sufficiently testifies to the violence of the escaping steam,

Several of these craters to the south-west were very active for many days after the eruption. The Black Crater may be specially mentioned amongst these as ejecting large quantities of mud and stones. The ash and stones around its margin, however, only reach a thickness of from 10ft. to 20ft. Five weeks later its activity was only shown by a number of boiling mud-holes, and during the following summer these were reduced to a single boiling pool, throwing up fountains of mud to the height of 10ft.

Echo Crater (sec Pl. p. 54) is the largest of this series. It is about a quarter of a mile in length, and its steep sides reach a considerable height at its south-west end; but towards the opposite end they fall, as the crater here is blown out of the valley near the Inferno Crater, from which it is only separated by the space of 50yds. When I visited it six weeks after the eruption the bottom was covered with water, which in many places was boiling tumultuously, giving rise to large clouds of steam. Towards the valley mentioned above, where the walls of the Echo Crater are lowest, there was a copious overflow of boiling water, which formed a river of considerable size. This must have been over 60ft. in width at its beginning, whilst farther down the Haumi Valley it became very much wider and shallower.

The Fissure

Looking Across Echo Lake and Inferno Craters in a North-East Direction.

The Southern Crater.

Looking North-East along [unclear: Fis]

A river of steaming and nearly boiling water of such size is a very remarkable phenomenon, and, like the river in the south bay of Rotomahana Crater, must point to a very abundant supply of underground water in the neighbourhood of the fissure. Farther down the valley the river ended in a wide pool opposite the end of No. 6 Crater. The eddies in the middle of this pool showed that the water was pouring down into some underground channel, along which it doubtless found its way into the Rotomahana Crater. It is to the action of this stream in the first place that we must attribute the formation of the deep watercourse which cut through the small craters near Rotomahana. A week or two later the channel was dry owing to the failure of the water rising in the Echo Crater. During the following summer the lake had shrunk so as to occupy only the southern part of the crater, whilst all hydrothermal activity was confined to the escape of a little steam close to the walls.

The rocks thrown out of all these craters to the south-west consist of the rhyolitic lavas and tuffs of which the country was composed and there was no ejection here of the black augite-andesite scoriæ as on the mountain. Large quantities of fresh scoriæ are, however, found near the Echo Crater, as well as bombs and small angular blocks of the same rock; and I was inclined at first to think that they came from this crater, and, perhaps, also from the Black Crater. More careful examination, however, during the summer, did not tend to confirm this opinion; the scoriæ and bombs are not specially numerous near the margins of those craters, but are rather more abundant at a distance, and there is nothing to remind one of the abundant accumulation of fresh scoriæ found at the edges of the fissure on Tarawera, where it is quite clear that they were ejected. The apparent abundance of the black scoriæ here is partly due to the smaller admixture of the light-coloured ash, and partly to a local thickening due to these parts being in the line of fire of fragments thrown out obliquely from the chasm in the side of Tarawera. Such an oblique discharge of fragments may appear rather startling at first, but it certainly took place on a small scale from some of the smaller craters, and was indeed witnessed by some observers.