Previous Next

Fire Mountains of the Islands

6. Calderas, Ignimbrites and the 1937 Eruption at Rabaul: 1914–1940

Great billowing clouds of smoke burst out on all sides: not only upwards, but in every direction, rolling in enormous bursts. The people rushed into the church to pray … Then came a rain of hot ash, so thick that we could not see a gleam of lamplight. I thought ‘This is the end’, felt my way through the darkness to the front, and asked the people, as loudly as I could, to beg God’s forgiveness.

Father Nollen (1937)

Garrison Life and Volcanoes

Australian military forces occupied Rabaul and ruled the Old Protectorate of German New Guinea for almost seven years following a short skirmish near Bitapaka in 1914. Rabaul became a battle-deprived backwater of the First World War when thousands were dying at Gallipoli and in the trenches on the Western Front. Garrison life for the Australians in New Guinea, wrote one war historian, ‘was a still lagoon, with the surf beating outside’.1 The Australians slipped readily into the ruling upper layer of the racially segregated society that had been created by the Germans in Rabaul. Immigrant Chinese, Malays, Ambonese, and also Japanese who had the same legal status as Europeans, represented the middle layer. Melanesians occupied the bottom layer. The Rabaul town plan was also racially segregated. This was ‘justified in the name of law, order, and disease control’ and supported by the military medical officer Ray Cilento who, two decades later as Sir Raphael Cilento, would investigate the medical implications of the 1937 eruption at Rabaul.2

German citizens in 1914 were allowed to remain in Rabaul if they signed statements of neutrality, but wartime tensions and suspicions remained between them and the Australians. German residents, and others, predicted that there would be an eruption at Rabaul in 1917.3 This was based on the belief that the 1878 eruption at Rabaul had taken place 40 years after a previous one in 1838 — presumably the poorly dated Sulphur Creek eruption reported by Father Boegershauser — and another eruption, therefore, was imminent. There was discussion about the matter in the garrison newspaper, the Rabaul Record. Potential evacuation routes were mentioned briefly in one article, which ended, however, on a rather fatalistic note that carried some militaristic historicity:

Do like the gallant Roman soldier who, at the destruction of Pompeii, died in his sentry box. Remain at your post. When the critical moment approaches you may — to cheer yourself up — sing ‘Let me like a soldier fall’.4


Figure 36. The large volcanoes of Kabiu (left) and Turagunan (right) together with Tavurvur, lower down on the left, form the background to this photograph of Australian troops and a six-inch gun mounted at Fort Raluana at the entrance to Blanche Bay in about 1918.

Source: Australian War Memorial (H01987).

Captain C.H. Massey, who was an officer in the Australian garrison and had had previous geological training, recorded the effects of an earthquake and tsunami in January 1916 which had caused damage to shipping and the wharfs at Rabaul. He noted that local earthquakes within the harbour seemed to originate from the area of Tavurvur volcano.5 Some local residents believed that earthquakes felt at Rabaul happened during heavy rain, but Massey did not agree that there was such a direct connection. He did suggest, however, that there may be a relationship to high ocean tides and therefore to the north-west monsoon or ‘wet’ season, and so, in turn, to ‘earth warping’ caused by the changing seasonal positions of the Moon and Sun relative to the Earth. He also stated his disbelief that volcanic eruptions could be predicted accurately using a precise 40-year periodicity.


Figure 37. Four Australian soldiers of the occupying force at Rabaul during the First World War, together with a local guide, are looking into the 1878 crater of Tavurvur volcano.

Source: Johnson & Threlfall (1985, figure on p. 12). Geoscience Australia (M2447-31A).

Massey concluded that the large flooded depression of Simpson Harbour was the original ‘crater’ — actually a caldera — of Rabaul volcano and that it was the source of the thick deposits of ash, pumice, and scoria making up the surrounding countryside to the west and south, notably at Bitapaka. These early geological observations by Massey relate to the origin of a series of calderas that form Blanche Bay, and to the especially large volcanic eruptions that accompanied caldera formation at Rabaul. They are matters that would occupy the attention of many subsequent volcanic geologists.

The Australian military authorities did not take any specific measures to monitor the volcanoes and earthquakes at Rabaul; although, in a letter written in September 1920, General G.R. Johnston, head of the Australian garrison, recommended the installation of a seismograph at Rabaul, evidently without a successful outcome.6 Rabaul and the former German colony came under an Australian civil rule in 1921, which lasted until 1942. The captured territory — following the Treaty of Versailles in 1919 — was a Class C Mandated Territory under the League of Nations, and this allowed the Australians to rule it almost as if it was part of Australia. There were subsequent high-level discussions about amalgamating the Mandated Territory with the Australian Territory of Papua, and about relationships with the British Solomon Islands Protectorate, but the three immediately adjacent territories of the British Empire remained administratively separate. Scientific interests of the new civil administration of the Mandated Territory were restricted largely to health, agriculture and anthropology and, as in previous German and Australian-military administrations, there was no interest in volcano monitoring — at least, until the disastrous 1937 eruption at Rabaul.

Australian Expedition along the Bismarck Volcanic Arc

Evan R. Stanley, the Government Geologist for the Territory of Papua, was a member of the Commonwealth Scientific Expedition to the new Mandated Territory in 1920–1921 — its aim, an assessment of the natural resources of the newly acquired territory. The expedition was led by Dr Campbell Brown, an authority in osmoridium, the natural metallic alloy used in the manufacture of pen nibs and surgical needles. Brown, who had connections with the Waterman Pen Company, had persuaded the Australian Prime Minister, W.M. ‘Billy’ Hughes, to have the Commonwealth Government fund the expedition. A 12-metre wooden ketch, the Wattle, was to be used as a mobile base.

The expedition was a poor replication of those that had taken place in German times, and it turned out to be an embarrassment to the Australian Government and Hughes because of its cost and non-achievement of objectives.7 Stanley’s report to the Australian Government — one of the expedition’s few redeeming features — included information on the nature of the volcanic chain from Rabaul in the east, along the north coast of New Britain, to Manam and the Schouten islands, 1,000 kilometres to the west of Rabaul. His report also contained a recommendation that a ‘Geophysical Observatory be established in conjunction with a Volcano Observatory’ for public-safety purposes in the New Guinea region.8 Calls for volcano monitoring at Rabaul would be made later, too, by the media, but like Stanley’s recommendation, they elicited no support.9

Stanley visited the small caldera at Lolobau off the north coast of New Britain, but sailed past Galloseulo and Pago volcanoes further east without knowing, given his low vantage points at sea level, that these, too, occupied calderas. New Britain, in fact, has more young calderas than any other area of similar size in the whole south-east Asian and south-west Pacific region. Stanley noted that Pago, the volcano that had been in explosive activity in 1911–1918, was bare of vegetation, was emitting water vapour and had a ‘breached’ appearance. A series of lava extrusions had been underway at Pago since before the 1918 visit of a Roman Catholic missionary, Father Ischler, to the floor of the Witori caldera, and lava may have continued to flow until as late as 1923.10

Stanley was the first geologist to recognise the existence of the 13-kilometre-wide caldera of Dakataua at the northern end of Willaumez Peninsula, New Britain, where D’Entrecasteaux in 1793 had identified only Willaumez ‘island’. He also noted that Wangore, or Bulu, just south of Dakataua, was an ‘active’ volcano because its upper parts in 1921 were completely bare and there were emanations of water vapour from near the summit.


Figure 38. Pago is here seen in explosive eruption in 1918, together with the rubbly surface of a recently erupted lava flow in the foreground. This photograph appears to be the earliest taken of any volcanic eruption in the Near Oceania region.

Source: P. Ischler, Mission of the Sacred Heart. Originally published in an article by Stamm (1930, p. 89). See also Cooke (1981, Figure 8).

A large eruption that took place at Manam on 11 August 1919, and which was observed by local missionaries, was recorded by Stanley as a result of his visit to the Manam area in 1921. This eruption appears to have been much larger than those reported previously at Manam, but it may have been similar in size to the eruptions that took place at the volcano, for example, in 1957–1958. Stanley wrote:

At about 2 p.m. large volumes of steam and black vapour … soon obscured the sun, but the glow from the crater illuminated the lower clouds … lava commenced to flow down the Eastern side and in about 5 minutes reached the sea … causing the water to boil furiously [the ‘lava’ was probably a pyroclastic flow] … When the rumbling ceased a grey-brown halo of dust encircled the mountain, which gradually spread for miles over the mainland. This dust was deposited to a depth of several inches at Potsdamhafen [on the mainland opposite] …

The [Manam] villages of Zogari and Josa were covered with dust, scoria, and hot fragments of vesicular lava. Gardens and houses were destroyed, and the natives for two or three days after had to place banana and breadfruit leaves on the tracks to prevent their feet from being burnt. At night it was possible to read a newspaper 8 miles away by the reflection of the lava on the clouds … The natives of Manam consider these visitations are due to work of a bad ghost, and consequently they are continually making tambaran, i.e., a solemn dance accompanied by the beating of drums and the blowing of 6-ft bamboo pipes which they believe assists in appeasing the wrath of the ghost.11

Stanley, after such successful fieldwork, was beginning by 1923 to establish a reputation as a pioneer geologist in the New Guinea region, but on a visit to Adelaide in December 1924 accompanying his sick wife Helen, he contracted blood poisoning supervening on a facial pimple and died suddenly at the age of 39. The Territory of Papua Administration did not replace him.

The 13 years or so after Stanley’s death are notable in the Near Oceania region for a deficiency of reports on volcanic activity, although not simply because of a lack of scientists, such as Stanley, to provide them. Reported eruptions seem to have been significantly less numerous than those reported in, for example, the late nineteenth century during German times, and the deficiency may indeed reflect a true situation of fewer eruptions during this particular period. But the low number of reports matches, curiously, a global trend in decreased reporting of volcanic eruptions that has been linked to the Wall Street market crash of 1929, the subsequent Great Depression of the 1930s, and to a worldwide preoccupation with economic and political anxieties rather than popular interest in volcanic eruptions.12 Nevertheless, volcanology as a whole was advancing, and particularly so in relation to the origin of calderas, such as those found in New Britain, and to the discovery of rock type known as ignimbrite.


Figure 39. The caldera lake of Dakataua volcano is seen in this sketch made by Stanley in 1921 looking southwards from the north of the caldera. Makalia, or Benda, is a youthful volcano, left of centre within the caldera. The arrow on the right points to Wangore, Vangori, Bulu, or Bula volcano, which is well to the south of the caldera, and is shown emitting water vapour.

Source: Stanley (1923, Figure 8).

Calderas and Ignimbrites

A pioneering study on the origin of calderas was published in Dutch in 1885 by R.D.M. Verbeek, a mining engineer of the Netherlands East Indies, following his investigation of the catastrophic eruption at Krakatau volcano in 1883.13 Verbeek proposed that a shallow and wide magma reservoir beneath some volcanoes is emptied so quickly during large explosive pumice eruptions that its roof collapses and disintegrates catastrophically. The formerly high ‘ancestral mountain’ of the volcano is engulfed and replaced by a low-rimmed but wide caldera at the surface. Other East Indies geologists and volcanologists who studied Krakatau subsequently — including the geodynamicist R.W. van Bemmelen and volcanologist Ch.E. Stehn who would later visit Rabaul — also supported this ‘emptying out and breaking down theory’.14 The theory eventually became more widely known to English-speaking geologists when an American volcanologist of Welsh extraction, Howel Williams, in 1941 published the benchmark paper ‘Calderas and their Origin’, which included extensive discussion of many calderas of general ‘Krakatau-type’.15


Figure 40. Magma in the shallow magma reservoir shown in the upper cross-section is erupted rapidly as a result of a powerful volcanic eruption, such as at Krakatau in 1883. This evisceration leaves the roof of the magma reservoir unsupported, and it collapses and disintegrates, as shown in the lower diagram. Subsequent geological studies of many eroded calderas have not revealed such strong disintegration of the roof rocks as shown in this cartoon, or in those shown by Howel Williams in his well-known studies of Krakatau-type calderas.

Source: Bemmelen (1929, detail from Figure 2).

A second important study on the origin of calderas was published in 1909 by three British geologists as the result of their detailed geological mapping of ancient volcanic rocks at Glen Coe in western Scotland. They proposed that a cylindrical mass of rock, several kilometres in diameter, had fallen vertically — piston-like — into an underlying reservoir of magma, which then ‘welled up around the subsiding mass, like liquor in a full bottle when the stopper settles home’.16 The upwelling magma, they suggested, produced eruptions from volcanoes arranged on the ring fault, and which filled the depression at the surface — the caldera — with volcanic materials. This process was called ‘cauldron subsidence’. Depressions that formed in this way later came to be called ‘Glen Coe-type’ calderas, and the published cartoon of the process became one of the most reproduced diagrams in the literature on volcanic geology, together with those used by Williams.

The two processes together — one based on the geology of the roots of an eroded volcano, the other on inferences from the surface on what had happened below — contain elements that are still used today in geological interpretations of the origin of calderas.


Figure 41. Volcanoes on a ring fault forming a Glen Coe-type of caldera are shown on the extreme left. Eruptions from a volcano in the centre of the caldera may later help fill it, as shown second from left. The roof above the rock cylinder on the extreme right is so strong that subterranean cauldron subsidence does not produce a caldera at the surface.

Source: Clough et al. (1909, Figure 14).

New Zealand was the site of another study that would have significant volcanological influence internationally. Pat Marshall mapped some of the geology of the Taupo Volcanic Zone in North Island and came across a distinctive type of volcanic rock that in 1935 he named ‘ignimbrite’.17 These rocks are characterised by distinctive, ubiquitous, disc-like blebs, some dark and glassy, which Marshall interpreted as collapsed pumice fragments. The fragments had been deposited from huge, pumice-bearing pyroclastic flows and subsequently squashed or flattened by the overlying weight of the deposit when the hot flow had come to rest. Ignimbrite or ‘fiery igneous cloud’ is similar to, and is as etymologically imprecise, as the French term nuée ardente or ‘glowing cloud’, but it has since been taken up with enthusiasm by volcanologists. ‘Ignimbrite’ is fundamentally a name for a pumice-bearing rock or deposit laid down by pyroclastic flows, but it is used also for the great volcanic-rock formations made up of the material, and even for the eruptive process that produced the deposit — ‘ignimbrite eruptions’.

Extensive research has been undertaken on ignimbrites and their origin, and there is now general agreement that they are produced mainly by collapse of a plinian eruption column. The high-rising eruption ‘stalk’ of the plinian column loses its upward momentum — either because the gas pressure at the volcano’s vent decreases or because the width of the vent increases by erosion caused by the inexorable expulsion of material — and the column collapses, crashing down onto the flanks of the volcano. Vast amounts of pumice, ash and gas then race outwards as gigantic and destructive pyroclastic flows, which may extend tens of kilometres from the volcanic vent. Not all of the flows are hot enough that when they come to rest they necessarily produce the flattened, disc-like, pumice fragments seen in the New Zealand examples. Ignimbrites are distinctive products of caldera-related eruptions, such as those that have taken place at Rabaul, Witori, Dakataua and Long Island.

Eruption at Rabaul in 1937

By May 1937 Rabaul had developed from its early German origins into a pleasant colonial town with shady streets, shops, sports facilities and commerce, which was dominated by the island trading companies of W.R. Carpenter and Burns Philp. It was, however, still strictly racially segregated and visibly European-controlled. The town and Territory were run by Australians for Australians, who were loyally British in their sentiments. Empire Day was celebrated on 24 May when Judge F. Beaumont ‘Monte’ Phillips reminded pupils at the Rabaul Public School of all the good things that the British Empire stood for. The worst of the Great Depression was over, but signs of war had again appeared following the rise of Nazism in Europe and the beginning of Japanese military expansion into east Asia.

Brigadier-General W. Ramsay McNicoll was Administrator of the Mandated Territory. He and his wife lived in Government House on Namanula Hill, where Albert Hahl had earlier enjoyed cool breezes, the westerly views of the town below and easterly ones across St Georges Channel. There was still no instrumental monitoring of the volcanoes at Rabaul, and other priorities had superceded any longer term concerns about volcanic hazards. This is perhaps not surprising for Australians, as active volcanoes were unknown in Australia itself, and there had been no further volcanic activity at Rabaul since 1878 — that is, well before the Australian takeover of the former German protectorate.

A sharp earthquake shook Rabaul town at about 1.20 pm on Friday 28 May 1937. Rabaul residents later recalled that this marked the beginning of the lead-up to the 1937 volcanic eruptions at Rabaul, although tremors were said to have been felt days previously near Vulcan Island, site of the 1878 eruption.18 Houses shook, trees swayed and crockery danced along the table on board the Montoro, which was discharging cargo at the Rabaul wharf. Damage and landslips were reported from out on the Kokopo Road at Karavia, near Vulcan, where the effects of the earthquake seemed greatest. Felt earthquakes became more numerous after five o’clock on the following morning, Saturday 29 May, and damage became more widespread. There were also reports of elevation and exposure of the sea floor at Vulcan Island, together with constant shaking, including at the slipway occupied by the steamer Durour on the nearby mainland.

A group of Europeans went out from Rabaul to investigate. Fish stranded on the exposed coral reef at Vulcan as a result of the uplift of the sea floor, were causing considerable interest, and were being collected by Tolai people from villages on the western shore of Blanche Bay, including many men who had by chance gathered at Tavana for an initiation ceremony of the tubuan male secret society.19 The European party took a boat out to see the rising sea floor but hurried back to shore just in time:

Right in our wake a blackish spiral of water was spouting 30 feet up … We had just passed over the spot … Then it started to crackle and explode and thunder, hurling up black stones and things, and was indescribable in its fury … [Minutes later the eruption was] in full force. Mountains of stones were being hurled thousands of feet high with deafening noise and sulphurous blinding smoke — black, black, black.20

The rapid rise of the Vulcan eruption column started at about 4.10 pm and provided little time for escape, especially from the villages close to the volcano. Pumice, gas and dust were being thrown upwards continuously from the new vent in typical plinian style. A surveyor later reported that the column reached a height of eight kilometres or more, but estimating such heights by eye from the ground nearby is notoriously difficult, and this minimum height could be an underestimate. It is, nevertheless, indicative that this plinian eruption was small compared to, say, those at Vesuvius in AD 79 or Krakatau in 1883.


Figure 42. The main road system and settlements of the Rabaul area shortly after the 1937 eruption are shown in this sketch map. Open red triangles are volcanic centres.

Source: Adapted from Johnson & Threlfall (1985, figure on p. 5, based on pre-war sketch map).

May is at the beginning of the south-east or ‘dry’ season, so the winds bent the eruption cloud towards the north-west, dumping the pumice and ash onto villages and people below. Small pyroclastic flows also emerged from the base of the eruption column and almost certainly engulfed people in their path, but these flows were not identified as such until much later from photographs and by geological study of the Vulcan deposits. The eruption cloud also contained water, at least some of it drawn up from the sea, and heavy rains fell to the ground in the north-west, causing streams to flood. Those who escaped from the Vulcan eruption later gave vivid descriptions of the eruption and its effects, telling stories of terror, anxiety, and loss.21 Father Nollen, for example, wrote of his experience of the Vulcan cloud at Malaguna on the north-west shore of Simpson Harbour. He thought about running away ‘But the black wall came on so fast that it was no use to think of flight’ and he feared for his life.22


Figure 43. The image depicts the near-vertical south-eastern side of the sub-Plinian eruption column from Vulcan. The photograph seems to have been taken late in the afternoon of 29 May 1937. Fronds of coconut trees, collapsed by ash loading, and floating pumice on the harbour waters, are shown in the two smaller photographs.

Source: Daily Telegraph, Tuesday 8 June 1937. Australian Consolidated Press Limited.

Evacuations began when people throughout the Rabaul area realised the seriousness of their situation. Many escapees in the west moved towards shelter at settlements to the south and particularly to the Vunapope Mission at Kokopo. McNicoll was visiting the Morobe Goldfields on the mainland, Chief Justice D.S. Wanliss had been acting in charge at Rabaul, but was cut off from Rabaul by road, and so Judge Phillips took over as the officer in charge of the Administration. Phillips was ‘a short sturdy man of great energy … he was in his element in a job such as this’, wrote Patrol Officer J. K. McCarthy, who himself experienced the drama of the unfolding disaster.23 People were moving spontaneously on foot and in vehicles out of the caldera using the obvious escape routes — up and over Tunnel Hill Road in the north-west, and eastwards onto Namanula Hill, site of the European hospital. Tunnel Hill evacuees who turned westwards along the north-coast road faired poorly as they had to endure ash fallout and mud rains from the Vulcan cloud.


Figure 44. Forked lighting illuminates the Vulcan eruption cloud in this photograph taken from the eastern side of Rabaul Harbour on the night of Saturday 29 May 1937.

Source: R. Davies. Johnson & Threlfall (1985, figure on p. 39). Geoscience Australia (G2008).

The town itself did not at first receive any ash falls, as the Vulcan cloud was being blown north-westwards, but the cloud expanded as the eruption developed and local winds dumped a few centimetres of ash on Rabaul. Night fell and the many distressed evacuees in relative safety could see the ongoing eruption lit up by powerful and near-continuous lightning flashes within the cloud. Phillips was able to have messages transmitted that evening from the radio on board a copra ship, the Golden Bear, which was still anchored at a wharf in the harbour. He first sent news of the eruption to the Prime Minister’s Department in Canberra, and later to the captain of the Montoro, which had left the harbour before the eruption, requesting that he return to Rabaul and standby.24 A ship-borne evacuation was being organised.

Those who spent the night on Namanula Hill saw the extent of the ash fallout and damage to Rabaul town at dawn the next morning, Sunday 30 May. People streamed down the road from Namanula during the morning to assemble in their thousands at Nodup on St Georges Channel, where a flotilla of ships and boats was assembling to take off the evacuees. The Golden Bear had escaped safely from the harbour in morning daylight, dashing past Vulcan, which was still in eruption and building up a new volcanic cone. The ship reached Nodup where other vessels were gathering. Montoro arrived there in early afternoon, and thousands of people were ferried from the shore in small boats out to the larger, waiting vessels. The transfer was reasonably orderly although slow, but anxiety increased when the evacuees became aware of another dark eruption cloud towering up behind Kabiu peak immediately above them. An eruption had started at Tavurvur volcano at about 1.00 pm, in near-simultaneous activity with that at Vulcan — as in 1878. It was another ‘double eruption’. The vessels made their way southwards towards Kokopo and the mission at Vunapope, which was becoming the main evacuation centre for the whole Rabaul area. The Montoro, carrying perhaps 5,000–6,000 people, stopped at night at the entrance to Rabaul Harbour, and Brett Hilder, an officer on board, recorded:

Most of us stayed up to watch the satanic celebrations; the two volcanoes, on each side of the entrance, were throwing up a solid jet of red-hot dust and stones to a great height and the two columns appeared to meet somewhere over the town of Rabaul. The lightning was fantastic … . The noise of eruption and thunder, and the thump of falling rocks was continuous, while our mass of human cargo had plenty to look at while they stood packed on all the decks.25

Heroic mission staff at Vunapope organised the care and shelter for the thousands of refugees now concentrated in the relative safety of Kokopo, site of the former German capital, and villages unaffected by the eruptions generously took in and fed those less fortunate. McNicoll returned from the mainland on Monday 31 May, flying in over the pumice-covered waters of Simpson Harbour to the new airstrip at Vunakanau. Australian newspapers published the Rabaul story that week as front-page news, and the editor of the Rabaul Times printed a special issue of his own newspaper at the Vunapope mission, detailing for residents the dramatic events of the past few days.26

The Administration began noticing a marked decline in activity at the two volcanoes and, within days, made a firm decision to return to Rabaul immediately, to begin cleaning up, and to encourage people to return there without delay. This prompted criticism from some of the European community, and certainly the bold decision — which turned out to be correct in the short term when eruptions ceased altogether at both volcanoes — was based neither on any scientific assessment of the potential of the volcanoes for further activity, nor on any thorough discussion on the suitability of the Rabaul site as a safe town or capital. The laborious clean-up began. Ash was removed from roofs and drains, water tanks and gutters were cleared, and health precautions were implemented. Damage to Rabaul buildings was relatively light, as the combined thickness of the pale pumiceous ash from Vulcan and the overlying dark, muddy, black ash from Tavurvur was only a few centimetres. But the ash had stripped protective vegetation from the slopes of the volcanoes and walls of the caldera, and heavy rains caused flooding in the town and the formation of deep gullies, adding to the problems of regenerating Rabaul.

The Administration’s initial focus was directed mostly towards Rabaul’s restoration and, evidently, much less towards determining the immediate fate of the villages close to Vulcan. Sir George Pearce, the Australian Minister for External Affairs, reported to the Australian Parliament as late as 18 June that ‘The deaths of natives definitely known at present number less than twenty’,27 and the Administrator in Rabaul would not be able to inform him formally until mid-July that, in fact, hundreds had perished. The final statistic, published in 1939, was 505 villagers ‘killed by being buried, crushed, asphyxiated or drowned’, plus two Europeans.28 McNicoll, in late June 1937, nominated ten Europeans for honours and decorations for their relief work during the disaster. Phillips was top of the list, and he eventually received a CBE, Commander of the Order of the British Empire. He would later become a key decision-maker in the days preceding the disastrous eruption at Lamington volcano in 1951.


Figure 45. Both Vulcan on the left and Tavurvur on the right are shown in reduced activity in this view from Taliligap looking northwards towards Rabaul, and photographed perhaps around 31 May 1937. Pumice covers the waters of the harbour. Vulcan has almost reached its full height of about 225 metres above sea level.

Source: Johnson & Threlfall (1985, figure on p. 85). Geoscience Australia (M2447-33A).

There were no awards for New Guineans. On the contrary, the Administration claimed that the new land that had formed at Vulcan was government-owned, even though survivors from nearby villages such as Valaur and Tavana had taken it over for traditional gardening soon after the eruption had ceased. Strong Tolai protests continued until after the Second World War in a dispute that reflected fundamental differences between European and traditional approaches to land-ownership laws. The Administration at one stage even issued a licence for a race course to be built on the new land at Vulcan. The dispute was not resolved until it was dealt with by the Supreme Court in Australia and settled in favour of the Tolai villagers.

Subsequent Investigations at Rabaul

The question of Rabaul’s suitability as a capital town did not disappear and there began a series of investigations into its future. The first of these started on 10 July 1937 when Sir Raphael Cilento, tropical-medical expert and director of health and medical services for Queensland, visited Rabaul to deal with the immediate health concerns in the reoccupied town, as well as with medical aspects of any future developments involving the capital. Cilento urged a meticulous and intensified reintroduction of routine medical services, and commented on the psychological benefits to the community that had resulted from the Administration immediately reoccupying the town. He also addressed future medical needs and concluded his report with the principle that residential areas for white people should be set apart from the rest of the population, not only for health reasons but also for the physical safety of white women.29 His proposal to settle Europeans on Namanula Hill was, however, impractical owing to the small amount of land available there.

Volcanologist Dr Charles E. Stehn, Director of the Netherlands Vulcanological Survey, and Dr W.G. Woolnough, Geological Adviser to the Commonwealth, arrived to advise the Australian Government on the suitability of Rabaul as the main administrative centre for the Mandated Territory. Woolnough was an Australian geologist, whereas the German-born Stehn had volcanological experience derived from studies undertaken, for example, at Krakatau, which could be applied to assessing the situation at Rabaul volcano.30 These two senior men were joined by Norman H. ‘Norm’ Fisher, a young Australian who, since 1934, had been the Government Geologist of the Mandated Territory of New Guinea, based on the mainland at Wau on the Morobe Goldfields.31 The three scientists undertook fieldwork together, Stehn and Fisher particularly building up an understanding of the past history of the Rabaul volcanoes. Fisher would play a major part in the aftermath of the 1937 Rabaul eruption and, indeed, would pioneer establishment of instrumental monitoring of the Rabaul volcanoes.

Stehn and Woolnough reached different conclusions as a result of their investigations, an unfortunate circumstance that did not help decision-makers in the Australian Government. Stehn, based on his experience with active volcanoes and instrumental monitoring in the Netherlands Indies, believed that a well-equipped volcanological observatory should be capable of providing warnings of impending eruptions sufficiently far in advance to enable the Rabaul population to be removed to places of safety. Woolnough on the other hand, stated bluntly that ‘… the advantages of retention of the capital in its present site and the provision of elaborate warning systems should not be entertained’.32 He concluded that the required scale of such an observatory, the uncertainties of eruption prediction, the generally small industrial and commercial development in the town, and the fact that the harbour might become blocked by future eruptions anyway, were all reasons why an observatory should not be constructed at Rabaul and why the town should be abandoned as the capital. Woolnough’s opinion, however, was ignored.

Instrumented observatories were, by 1937, a well-established feature of some of the world’s best known volcanoes, the idea deriving from the much earlier concept of astronomical observatories where instruments were directed at the stars, rather than at the internal and surface behaviour of a volcanic Earth. The first volcanological observatories were built in Italy, at Vesuvius in 1841–1845, and at Etna in 1878–1881. The Americans had established an observatory at Kilauea volcano, Hawaii, in 1912, and Japanese observatories had been built on Aso by 1928 and at Asama by 1933. Furthermore, E.R. Stanley had drawn the attention of the Australian Government to the need for volcanological observatories in the Territory of New Guinea after his attendance at the Pan-Pacific Scientific Conference in Honolulu in 1920.33 Stehn and Fisher were, therefore, building on something of a volcanological tradition.


Figure 46. Volcanologist C.E. Stehn is shown here in the foreground returning from the pumice-covered Durour, a vessel left stranded on a slipway south-west of the former Vulcan Island.

Source: N.H. Fisher. Johnson & Threlfall (1985, figure on p. 124). Geoscience Australia (no registered number).

Other inquiries into the suitability of Rabaul as a capital followed the investigation by Stehn and Woolnough, but no definite decisions were made for four years. Furthermore, the idea of monitoring volcanic activity instrumentally at Rabaul for early warning purposes was being kept alive. Fisher began to spend more time at Rabaul after completion of the investigation and he visited Stehn in the Netherlands Indies in 1939 for volcanological training. Fisher, strongly influenced by Stehn’s views, in 1939 published his conclusions on the geological history of the Rabaul area, promoting the view that a huge volcano — perhaps more than 2,700 metres high — or a group of somewhat lower cones, once covered the area now occupied by the Blanche Bay caldera. The smaller volcanoes of Kabiu, Tovanumbatir and Turagunan were simply ‘parasitic cones’ on the flanks of this great ancestral mountain.

Fisher wrote that ‘A tremendous outburst or series of outbursts … blew most of the mountain to fragments’, scattering pieces of the central volcano around the countryside. This, he went on, was followed by subsidences around the periphery of the resulting explosion crater, forming the Blanche Bay caldera, and then a large eruption finally produced huge thicknesses of ‘pumice ash’.34 This interpretation by Fisher is not the mechanism of formation of Krakatau-type calderas, as promoted later by Howell Williams, but rather reflected the view, which was held by many geologists at that time, that calderas were the result of volcanoes that ‘blew up’, or exploded outwards, rather than collapsing wholesale into underlying magma reservoirs. Furthermore, Fisher and Stehn did not identify ignimbrites at Rabaul. Much later work would, however, reveal them and, indeed, a much more complex geological origin for Blanche Bay. Fisher also compiled at this time a volcano inventory for the Mandated Territory, in which the full number of young calderas in New Britain was recognised for the first time, including those at Pago and at Galloseulo where a ‘crater lake’ — the crater was in fact a caldera —had been recently reported by the pilot of an aircraft delivering mail.35

A routine of visual observations of the volcanoes and measurements of temperatures at hot springs and fumaroles around the harbour was established by Fisher at Rabaul, and funds then became available for construction of a small observatory and cellar for installation of equipment. These were built in 1940, high on the northern rim of the caldera — now called Observatory Ridge — below the south-western flank of Tovanumbatir. The observatory had spectacularly uninterrupted views overlooking the town towards Tavurvur, Vulcan and the other young volcanoes of Blanche Bay. Fisher had well-defined plans for the observatory to contain a German-made, two-component Wiechert seismograph, two tiltmeters of Italian manufacture for measuring changes in ground slope, an earthquake ‘annunciator’ that triggered an alarm when larger earthquakes took place, together with other ancillary equipment.


Figure 47. N.H. Fisher is seen here taking temperatures in an embayment on the northern side of Vulcan cone in August 1937. Temperature monitoring can, in some circumstances, be an effective way of determining whether a volcano is becoming restless or not.

Source: Johnson & Threlfall (1985, figure on p. 124). Geoscience Australia (GB-2552, detail from original photograph).

There were plans to establish observation posts in tunnels in the caldera walls near Tavurvur and Vulcan, and to equip these with tiltmeters and small pendulum devices — tromometers — to measure long-period ground motions, together with telephone lines to the central observatory.36 Purchase of these instruments and funding to develop extra observation posts were severely curtailed by outbreak of the Second World War in Europe. Fisher finally had to design his own seismographs and annunciator for the Observatory Ridge cellar and have them built in Rabaul by the Public Works Department.37 Nevertheless, the Rabaul Volcanological Observatory (RVO) had been established, and Fisher became the first volcanologist to be based permanently in the Near Oceania region, assisted at times during 1937–1942 by surveyor L.E. ‘Les’ Clout and by geologists C. L. ‘Clem’ Knight and L.C. ‘Lyn’ Noakes.


Figure 48. The stark, utilitarian appearance of the original volcanological observatory at Rabaul, seen here sometime in 1940–1941, as well as a primitive set of monitoring instruments housed in the building’s cellar, belies the historical importance of the observatory’s construction and of its expectation of early warnings of volcanic eruptions at Rabaul.

Source: N.H. Fisher (likely photographer). Johnson & Threlfall (1985, figure on p. 130). Geoscience Australia (GB-1352).

Construction of the observatory at Rabaul in 1940 also marked, historically, the start of a new scientific era in using instrumental data to help understand the geophysical nature of the volcanoes in the region and to establish how eruptions might be forecast for the benefit of at-risk communities. It was, significantly and for the first time, an acknowledgement from government authorities that instrumental monitoring and early warnings of eruptions might make a difference in mitigating the effects of volcanic disasters in the Territory of New Guinea. The decades up to 1994 would be marked by a history of development of the work of RVO scientists and technicians, and of international volcanology in general. RVO would face several challenges, but particularly in 1994 when there would be a repeat of the near-simultaneous eruptions that had taken place in 1937 at Tavurvur and Vulcan.


Anonymous, 1917a. ‘Earth Tremor’, Rabaul Record, 2, no. 2, pp. 3–4.

Anonymous, 1917b. ‘More Earth Tremors’, Rabaul Record, 2, no. 3, p. 3.

Anonymous, 1932. ‘No Scientific Records Kept of Rabaul’s Quakes’, Pacific Islands Monthly, 3, no. 5, p. 26.

Arculus, A. & R.W. Johnson, 1981. 1937 Rabaul Eruptions, Papua New Guinea: Translations of Contemporary Accounts by German Missionaries. Bureau of Mineral Resources, Canberra, Report 229.

Bemmelen, R.W. van, 1929. ‘Het caldera probleem’, De Mijningenieur, Vereeniging van Ingenieurs en Geologen, Bandoeng, 10, no. 5, pp. 101–12.

——, 1949. ‘Charles Edgar Stehn (1884–1945), Nécrologie’, Bulletin Volcanologique, 8, pp. 133–37.

Chinnery, S.J., 1998. Malaguna Road: The Papua and New Guinea Diaries of Sarah Chinnery. National Library of Australia, Canberra.

Cilento, R., 1937. ‘Report on the Medical Significance of the Recent Eruption in Blanche Bay, Territory of New Guinea’. National Australian Archives CRS A518, Item X836/4.

Clough, C.T., H.B. Maufe & E.B. Bailey, 1909. ‘The Cauldron-subsidence of Glen Coe, and the Associated Igneous Phenomena’, Quarterly Journal of the Geological Society, 65, pp. 611–78.

Cooke, R.J.S., 1981. ‘Eruptions at Pago Volcano, 1911–1933’, in R.W. Johnson (ed.), Cooke-Ravian Volume of Volcanological Papers. Geological Survey of Papua New Guinea Memoir, 10, pp. 135–46.

Cummins, J.J., 1917. ‘Early Experiences in New Britain’, Rabaul Record, 2, no. 7, pp. 9–10.

Davies, H.L., 1987. ‘Evan Richard Stanley, 1885–1924: Pioneer Geologist in Papua New Guinea’, BMR Journal of Australian Geology and Geophysics, 10, pp. 153–77.

Fisher, N.H., 1939a. ‘Geology and Vulcanology of Blanche Bay, and the Surrounding Area, New Britain’, Territory of New Guinea Geological Bulletin, no. 1.

——, 1939b. ‘Report on the Volcanoes of the Territory of New Guinea’, Territory of New Guinea Geological Bulletin, no. 2.

——, 1940. ‘Note on the Vulcanological Observatory at Rabaul’, Bulletin Volcanologique, 6, pp. 185–87.

Gould Fisher, F., 1994. Raphael Cilento: A Biography. University of Queensland Press, St Lucia.

Hilder, B., 1961. Navigator in the South Seas. Percival Marshall & Co., London.

Johnson, R.W. & N.A. Threlfall, 1985. Volcano Town: The 1937–43 Rabaul Eruptions. Robert Brown and Associates, Bathurst.

Mackenzie, S.S., 1987 (1927). The Australians at Rabaul: The Capture and Administration of the German Possessions in the Southern Pacific. University of Queensland Press, St Lucia, in association with the Australian War Memorial, Canberra.

Marshall, P., 1935. ‘Acid Rocks of the Taupo-Rotorua Volcanic District’, Transactions of the Royal Society of New Zealand, 64, pp. 1–44.

Massey, Captain C.H., 1918. ‘Vulcanicity of Rabaul District’, Rabaul Record, 3, no. 1, pp. 8–10.

——, 1923. ‘Notes on the Physiography of Eastern New Guinea and Surrounding Island Groups: With Special Reference to the Volcanic Features of the Rabaul District of New Britain’, Proceedings of the Royal Society of Queensland, 35, pp. 85–108.

McCarthy, J.K., 1971. ‘When Matupit Blew, It Was Time to Go’, Post Courier (Port Moresby), 20 July, p. 5.

Neumann, K., 1996. Rabaul Yu Swit Moa Yet: Surviving the 1994 Volcanic Eruption. Oxford University Press.

Pearce, G., 1937. ‘Volcanic Disturbances at Rabaul’, Commonwealth of Australia, Parliamentary Debates, Session 1937 (1 GEO.VI), 153, pp. 43–44.

Simkin, T. & R.S. Fiske, 1983. Krakatau 1883: The Volcanic Eruption and its Effects. Smithsonian Institution Press, Washington D.C.

Simkin, T., & L. Siebert, 1994. Volcanoes of the World: A Regional Directory, Gazetteer, and Chronology of Volcanism during the Last 10,000 Years. Geoscience Press, Tucson.

Stamm, J., 1930. ‘Wie die Fischreusen mich beinahe um meinen ersten Tauftag in Nakanai brachten’, Hiltruper Monatshefte, 47, pp. 85–91.

Stanley, E.R., 1923. ‘Report on the Salient Geological Features and Natural Resources of the New Guinea Territory including Notes of Dialectics and Ethnology’. Report on the Territory of New Guinea, Commonwealth of Australia Parliamentary Paper 18 of 1923, Appendix B.

Stehn, Ch. E., 1929. ‘The Geology and Volcanism of the Krakatau Group’, Proceedings of the Fourth Pacific Science Congress, Batavia, pp. 1–55.

Stehn, Ch.E. & W.G. Woolnough, 1937. ‘Report on Vulcanological and Seismological Investigations at Rabaul’, Commonwealth of Australia Parliamentary Paper 84 of 1937, pp. 149–58.

Thomas, G., 1937. ‘Our Volcanic Issue’, Rabaul Times, no. 633, Friday 4 June.

Verbeek, R.D.M., 1885. Krakatau. Batavia.

Wilkinson, R., 1996. Rocks to Riches: The Story of Australia’s National Geological Survey. Allen & Unwin, St Leonards.

Williams, H., 1941. ‘Calderas and their Origin’, Bulletin of the Department of Geological Sciences, 25, pp. 239–46.

1 Mackenzie (1987), p. 316.

2 Gould Fisher (1994), p. 17.

3 Anonymous (1917a) and Cummins (1917).

4 Anonymous (1917b), p. 3.

5 Massey (1918, 1923).

6 Stanley (1923), p. 44.

7 Davies (1987).

8 Stanley (1923), p. 92.

9 Anonymous (1932).

10 Cooke (1981).

11 Stanley (1923), pp. 52–53.

12 Simkin & Siebert (1994).

13 Verbeek (1885). Simkin & Fiske (1983) provided an English translation of Verbeek’s classic memoir on the occasion of the centenary of the 1883 eruption at Krakatau.

14 Bemmelen (1929; the quotation is from the English summary on p. 111) and Stehn (1929). Simkin & Fiske (1983) republished the Stehn paper in their book on Krakatau.

15 Williams (1941).

16 Clough et al. (1909), p. 664. Glen Coe perhaps is better known for the genocide in 1692 of members of the Maclain branch of the MacDonald clan by a Campbell regiment of government troops.

17 Marshall (1935).

18 Fisher (1939a) and Johnson & Threlfall (1985).

19 See, for example, Neumann (1996).

20 Chinnery (1998), pp. 210–11.

21 Collections of stories were given by Arculus & Johnson (1981), Johnson & Threlfall (1985) and Neumann (1996).

22 Arculus & Johnson (1981), p. 29.

23 McCarthy (1971), p. 5.

24 Official government correspondence and other documents concerning the 1937 eruption are housed in files of the Territories Branch at the National Archives of Australia, Canberra.

25 Hilder (1961), pp. 54–55.

26 Thomas (1937).

27 Pearce (1937), p. 44.

28 Fisher (1939a), p. 24.

29 Cilento (1937).

30 Bemmelen (1949).

31 Wilkinson (1996).

32 Stehn & Woolnough (1937), p. 157.

33 Stanley (1923).

34 Fisher (1939a), p. 13.

35 Fisher (1939b). The crater lake is Lake Hargy.

36 Fisher (1940).

37 N.H. Fisher (personal communications, 1982, 2006) and Johnson & Threlfall (1985).

Previous Next