The Mystery of the Tumuli

Over 400 mounds on New Caledonia in the southwest Pacific may mark a human presence predating the current earliest generally accepted dates by thousands of years.

Editor’s Note: The following paper, published exclusively here at Popular Archaeology, relates a detailed review of the investigation of enigmatic tumuli, or mounds, on the Isle of Pines, part of the island French territory of New Caledonia in the South Pacific. The origin and purpose of these tumuli have eluded investigators and scientists for decades, sparking various conjectural explanations for their existence. If the tumuli represent the work of humans, they would provide evidence for human occupation of the island thousands of years before the currently accepted dates for human habitation. The author suggests a new hypothesis for further study. 





The ancient migrations settling the South Pacific remain poorly understood. This is particularly so when the New Caledonia sequence is considered and where perhaps the greatest archaeological mystery of the region awaits resolution. Scattered across the Isle of Pines more than 400 tumuli, or mounds, shelter substantial cement blocks, many with a circular shaft running through their centre. The only tumulus fully excavated also revealed a large cone-shaped iron object buried directly beneath the shaft, function unknown. Described by two archaeologists as recently as 2015 as “a kind of archaeological nightmare” 1  the origin and purpose of the tumuli remains enigmatic, the mystery compounded by studies revealing dating thousands of years earlier than settlement models currently allow. The paper outlines various investigations over the last century before arguing that excavation data from two seminal studies (Chevalier 1963 and Lagarde 2017) demonstrate that non-anthropic explanations – notably the long-standing avian theory – fail to address the substantial data indicating a human presence on the island predating the Early Modern Era. A new hypothesis for the original purpose of the tumuli is presented, one based on the accumulated data.



Isle of Pines, Tumuli, New Caledonian pre-history, pre-Lapita culture


While the term “tumulus” – a mound – is accurate enough, the word can have an unfortunate association in English with burial practices. To remove the ambiguity provided by the fact that there are, of course, numerous other “mounds” of varying characteristics, both on the Isle of Pines and elsewhere, the term “tumulus” in this paper refers to what are more accurately described as a “cement-core tumulus.”

The cement interior of the tumuli is typically referred to as a “cylinder” in the various early reports. As this term conveys the implication of circularity, and these structures are actually cube-shaped, this paper will use the term “core” as more accurate terminology.



The Melanesian Kanak population is believed to have inhabited the islands of New Caledonia since the so-called Lapita period around 1350 BC. The first European arrivals on Kunie, the Isle of Pines, in the early nineteenth century noted the numerous tumuli, but the indigenous population ascribed no particular significance to them. This led to some early investigations and conjecture, but no sustained attention for many decades. Proper studies and initial excavations commenced only in 1959. Since then, scholarly opinion has advanced – and discarded – a range of ideas concerning the tumuli, but without a consensus to the present.


More than 400 large tumuli are dotted over much of the Isle of Pines with a concentration on the central so-called “Iron Plateau,” plus about 20 at particular locations in the south of the main island of New Caledonia. No systematic attempt has yet been made to locate tumuli elsewhere on the main island or, indeed, the other islands of the group. Early settlers assumed these were simply ancient burial mounds, but no graves, bones, pottery, charcoal, etc. have ever been found in them.

The locations of the tumuli generally appear to be quite random regardless of the terrain, although in a very few places a possible alignment or circle might be perceived in aerial images. More certain is the fact that most tumuli are rarely spaced less than 200 m apart, as geologist Jacques Avias noted in 1949.2 The tumuli average about 3 m in height with a circumference ranging between 15-25 m.3 While the majority of tumuli fall into this size range, a handful of oval shaped structures have also been described; one that was examined measured 3 m high and 16 m wide.4 

New investigations have revealed that as much lies beneath the tumuli as above; the mounds contain a large square or oblong core made of a high-grade concrete. The most detailed excavation to date of a tumulus was on the Isle of Pines in 1959, revealing a core almost 2.60 m tall and a substantial 1.8 m across. At several points along the height of the core it was encircled by rings of iron-oxide nodules.5 The oval-shaped tumulus noted earlier proved to have two concrete cores, each 2 m diameter, and placed 5 m apart.6 While the concrete cores so far examined appear to consist of homogenous material, the upper part of one of the Paita cores that was completely demolished contained irregular pockets of soil of varying sizes.7 With regard to the soil intrusions, excavation data remain too limited to draw any conclusions.

Most intriguingly, the Isle of Pines tumulus excavated by Chevalier in 1959 revealed that buried directly below the centre of the core was a symmetrical cone or top-shaped iron object, pointing down and surrounded by 3 rings of large iron nuggets. It was over 2 meters long.8 As will later be discussed, Chevalier’s excavations of two tumuli at Paita also revealed puzzling features below the concrete core. One of the tumuli completely demolished on Paita revealed a separate “vaguely circular” slab of cement beneath it, in the middle of which was a wide (90 cm) and deep (1.8 m) trench, oriented east-west, and two holes at different depths. The other Paita tumulus did not appear to have a comparable feature beneath the core.9

Another, most significant, feature was not noted until the excavations led by Jack Golson in 1959-1960: a circular, vertical, shaft running from the top of the core at its centre down to its base; this will be discussed further.


Figure 1: Known tumuli locations on the Isle of Pines (after D. Frimigacci, Notice explicative de la feuille Ile des Pins: Carte géologique à l’échelle, Paris: BRGM, 1986, 28). 


Figure 2A: One of hundreds of enigmatic “tumuli” on the Isle of Pines. This early photograph was published by R. H. Compton in 1917 and shows one of the rare double-cored mounds.


Figure 2B: A 1949 aerial view of the “Iron Plateau” of the Isle of Pines with scores of tumuli visible before reforestation of the area.


Figure 3: Today, hundreds of mounds lie in open land or in pine forests on the Isle of Pines. Photograph by the author, 2017. 



Some of the underlying issues for some aspects of New Caledonian and, more generally, for South Pacific pre-history remaining in somewhat of a stalemate were discussed in Roger Green & J. S. Mitchell’s 1983 paper, “New Caledonian Culture History: A Review of the Archaeological Sequence,”10 The authors provide the following reasons: firstly, over the years research has appeared in a great range of venues and is seldom integrated. Secondly, not all  findings have been published in readily available sources (and sometimes not published at all); thirdly, a perceived language barrier as the main sources are [obviously] in French and English; finally, one of the primary sources (Gifford & Shutler, 1956) has some deficiencies in its presentation of data – something Green and Mitchel discuss – and investigators since have not followed a consistent terminology or frame of reference.

A chronological examination of the main sources reveals the long-standing confusion over how to account for the tumuli within the accepted time-frame of human arrival in New Caledonia, as well as the factors outlined by Green and Mitchell just noted. To examine even a brief history of modern sciences’ attempt to understand these structures is to encounter a spectrum of human foibles, ranging from honest reporting with rational analysis to degrees of disingenuity and, above all, avoidance.

Early observers on the Isle of Pines realized that these scattered mounds represented a mystery and were quick to seek answers. As early as 1897, Frenchman Theophile-Auguste Mialaret reported “opening” a tumulus, but finding nothing of interest.11 In 1917, Englishman R. H. Compton reported in The Geographical Journal the “remarkable” dome-shaped mounds of earth: “…excavation has revealed no contents of interest…various suggestions have been put forward to explain them.” In both cases their efforts were watched, as Compton noted, “without fear by the indigenous people who, while claiming that they did not make the tumuli, could give no explanation for their origin.” He concluded: “I am quite unable to offer any suggestion as to their origin myself.”12

The 1949 paper by Avias included an aerial photo (reproduced here as Figure 2B) showing “nearly two hundred” tumuli on the central “Iron Plateau.” He proposed a series of ancient arrivals in the southern Pacific, predating the Melanesians, concluding “…at least the following hypothesis can be put forward: a civilization …preceded the present Kanak civilization…this civilization had a Neolithic industry more advanced than the indigenous people.”13

That the structures contained a massive concrete core remained unsuspected and unknown until proper excavations on the Isle of Pines began in 1959. Leading the post-war vanguard of serious investigators were two archaeologists, Jack Golson and Luc Chevalier.                

From 1959-1960 British-born Australian archaeologist Jack Golson led a team excavating four “tumuli” on the Isle of Pines, finding that there were at least 3 types of structures. Golson’s team went on to map more than 120 mounds and excavated three of them. They viewed the concrete cores as “anchors” and established, by probing their centers, that “many” other mounds also had a cement core. As briefly noted earlier, their excavations revealed that at least several of the cores had a vertical hole running through them, a feature they termed “postholes,” but could not determine what function the shaft might serve.

Golson published a brief report in 1961, “Report on New Zealand, Western Polynesia, New Caledonia, and Fiji” concluding: “The mystery of the tumuli is thus, despite this spate of activity, as great as ever. Who were these concrete makers of New Caledonia and what is the function of their constructions? Native tradition is silent and the archaeologist as yet ignorant.”14 In 1963 he published more data from his expedition in “Rapport sur les fouilles effectuees a l’ile des Pins (Nouvelle Caledonie) de Decembre 1959 li Fevrier 1960” and with Daniel Frimigacci in 1986, “Tout Ce Que Vous Pouvez Savoir Sur Les Tumulus, Meme Si Vous Osez Le Demander,’’15 providing a comprehensive summary of the topic.              

Luc Chevalier, the French archaeologist, was led by unexpected circumstances in September 1959 to shed further light on the tumuli. Based in Noumea as curator of the National Museum, Chevalier traveled to the Isle of Pines to investigate a report that local workers upgrading a road had encountered a solid white “rock” at the centre of a tumulus they were demolishing for road-building material. With a half-exposed mound as a starting point, Chevalier immediately began excavating until the southern side of the tumulus core was exposed and the unexpected object beneath it revealed. Chevalier describes it as an “iron oxide rock” about 2.40 m tall – almost as tall as the core itself – and a maximum diameter of 1.10 m, resembling a symmetrical upside-down “top” or cone, pointing down. At that point, the entire structure threatened to collapse; amid several small landslides excavation was halted. In 1963, Chevalier’s final report, “Le Probleme des Tumuli en Nouvelle-Caledonie” appeared in the Bulletin de la Societe des Etudes Melanesiennes;16  It deserves close reading.


Figure 4A, 4B (2 images): The exposed 2 meter high concrete core from the tumulus excavated by Chevalier in 1959, with a close-up view of large iron nodules embedded in it. Photographs by the author, 2018.


Figure 5: Chevalier’s 1963 excavation report (his Fig. 5) shows the interior of the mound after excavation. The rings of iron nodules and the top-shaped object buried beneath the central core are clearly shown.


Chevalier then examined other nearby tumuli on the Iron Plateau, determining that at least many of them also contained a central core made of “white mortar” and making particular note of the oval-shaped tumulus that contained two cores. Returning to Noumea, he then investigated several tumuli in the region of Paita just north of the capital. An identical situation to what had happened on the Isle of Pines involving road workers was replicated in Paita; this time uncovering two concrete-cored tumuli about 300 m apart. It was here that several intact Placostylus snail shells found embedded in the concrete matrix of the cores and were sent, with samples of the cement, to a laboratory in Paris for dating. His report shows images of the exposed concrete cores of 2 of the Paita tumuli being demolished.

Despite anticipating further excavations that would “verify the presence and shape of this [cone-shaped] stone under the other tumuli,”17 Chevalier’s next excavations at Paita would only raise more questions. Aside from the absence of any mention of Golson’s “post holes” in the concrete cores (about which Chevalier may not have been aware) in his reports, those he examined remained completely analogous to those in the Isle of Pines tumuli although, as noted earlier, the Paita cores had differing material beneath them and the cone-shaped object was not found in these part-excavations. Nor does Chevalier mention returning to the Isle of Pines, although he expressed optimism that a complete excavation campaign there would be possible the following year, ie. 1960.18

Among other significant aspects, Chevalier noted that on both the Isle of Pines and at Paita the excavations were culturally sterile; that is, no tools, pottery, charcoal etc. could be found. The closest thing to a cultural artifact was the apparent surplus of construction materials found near the two Paita tumuli excavated, a quantity of lime mortar mixed with silica near one and a 10 m wide elevation of silica gravel near the other. Both offered clear indications of human involvement likely, as Chevalier recorded, being the places where cement was prepared before being poured into the centre of the tumulus to form the core.19

He commented, “There is no doubt that the same people who made the tumuli on the Isle of Pines also made those in the region of Païta. The only difference between them is that the tumuli of Païta are made of a supply of silica present everywhere, while those on the Island of Pines used the “iron shot” which is found in abundance on the plateau….It is also necessary to determine the origin of most of the mortar: or is this the result of the calcination of coral?… it is curious to find no fragments of charcoal in the mortar. [It indicates] a space between the time of the construction of the tumuli and the pre-European Melanesian era.” The conclusion of his report, “Many Questions Without Answers,” summed up the facts:                                                                “The tumuli of New Caledonia are undoubtedly the result of an ancient human activity of which there is no trace in the local tradition.” Remember that the manufacture and use of lime was not known to the natives before the arrival of the missionaries in 1843. One is perplexed by the importance of the work done to erect these monuments. When one thinks that a tumulus represents on average, a volume of 200 m3 and that on the Isle of Pines alone there are more than 300 tumuli, requiring a very large workforce. However, this workforce during its work, had to leave some traces of its passage, its life, its activities …to date, no trace, no evidence, no vestige, either inside or outside the tumulus has been found.”

Chevalier closed with the following, very perceptive, observation:

“…a final question arises in our mind: the builders of these tumuli knew perfectly the principle of manufacturing and using the mortar for the erection of the cylinder. Why did they not use this knowledge for other works, other realizations, their dwellings, for example…? Would they have reserved this knowledge only for the tumuli, conferring to these monuments a particular character, even a sacred one?” 

Had more attention been given to these questions by those who came after Luc Chevalier, our understanding of these structures would have advanced more rapidly than it has and without the avoidance of facts and obfuscation in so much of the commentary since then.


The results of two chemical analyses of the components of the concrete (one from the Isle of Pines, one from a Paita tumulus) appeared in Chevalier’s 1963 report and are quite comparable; the primary disparity evident in the Fe and Si amounts reflecting the two very different environments of origin. The dating of the mortar and snail shells from a core excavated on Paita by Chevalier was published in a 1966 paper in Radiocarbon. This gave radiocarbon dates for surface and interior mortar (7070 ± 350 and 9600 ± 400 BP respectively), plus a date for Placostylus snail shells on the surface of a cylinder (12,900 ± 450 BP). Echoing Chevalier’s belief that the tumuli “testify to an important human activity completely extinct today…” the report’s authors noted that if the cores were really man-made “…they are by far the most ancient mortars known.”20 Radiocarbon dating by others since have served to confirm Chevalier’s very early dates. For example, a 1972 dating of tumulus cement yielded ages ranging from 4120 ± 90 to 7710 ± 70 BP.21

Space constraints prevent a full review here of those who have accepted this dating, but they include scholars such as Brookfield & Hart in 1971,22 Fr. Marie-Joseph Dubois (1976),23 Richard Shutler Jr. (1978),24 Peter Bellwood (1978),25 Roger Green & J. S. Mitchell (1983),26 Kerry Ross Howe (1984),27 John R. H. Gibbons & Fergus G. A. U. Clunie (1986),28 Christophe Sand (1999)29 and Patrick V. Kirch (2000 and 2017).30 

It is an understatement to say that the disconnect between the Placostylus and mortar dating  and the conventional understanding of human settlement in the Pacific is unbridgeable. To contemplate people arriving in New Caledonia as early as 12,000 years ago simply cannot be massaged into the accepted scenario. Consequently, many – perhaps most – of these researchers believed that the extraordinarily early snail shell dating had probably been of much earlier material, thus allowing for the actual tumulus construction to have taken place perhaps in the Lapita period, roughly 1500 to 500 BC. Other commentators simply ignored these data. No-one seems to have asked how multiple fragile snail shells survived intact for many thousands of years before finding their way onto the outside of a block of drying cement, for the shells were found on the surface of the cement, not in its interior. One shell can be clearly seen in Fig. 14 of Chevalier’s report.


Unsurprisingly, the inability of archaeology over many years to provide resolution as to the origins of the tumuli led to several proposals far from main-stream thinking. This began as early as 1949 with the “working hypothesis” of Jacques Avias already noted. After attempting in vain to locate patterns in the placement of the tumuli and any correspondences to the southern constellations Avias broadened his studies to all other traces from the past, such as petroglyphs and pottery styles across New Caledonia. His paper offered suggestions that a Neolithic culture predating the present Kanak civilization may have been influenced by, or even originated in, certain Eastern cultures, specifically the Chinese and Japanese. He developed an earlier thought expressed by P. Rivet in 1931 that from Japan the Ainu may have explored the Pacific islands.31 While not subscribing to such views themselves, Daniel Frimigaccci and Jack Golson also mentioned that the “most extreme hypotheses were advanced to explain their presence [the tumuli],” noting in particular “extraterrestrials or mysterious populations now extinct practicing stellar-solar rites.”32

One of the great ironies is that attempts to explain the tumuli by invoking incursions by distant explorers, inhabitants of the legendary continents of Mu, Atlantis or the Aroi Sun Kingdom,33 or extraterrestrials,34 now appear somewhat less fanciful than they once seemed, especially when more prosaic explanations are compared. This is no more so than the avian theory to be discussed next.


In late 1985, after bones of an extinct Megapode bird species were found in New Caledonia, Cecile Mourer-Chauvire & Francois Poplin published “Le Mystere de Tumulus de Nouvelle-Caledonie,”35 proposing that these birds may have formed each tumulus by scraping the ferritic soil into a giant mound in which the eggs could be deposited and incubated until hatched. Left unexplained was how this process somehow produced a massive, linear, block of cement in the interior, with a hole running through it and, in at least one case, a large cone-shaped iron structure buried directly below that. They concluded that “the giant bird hypothesis is just as reasonable as the theory that these mounds were built by ancient humans who knew how to make cement.” 


Figure 6: The conjectured formation of the tumuli by the large extinct bird Sylviornis neocaledoniae, seen here scraping soil into a mound for nesting.


From this quite cautious proposal, the idea caught hold. Notably, since his 1983 co-authored paper listed earlier, archaeologist Roger Green came to accept this naturalistic explanation for the mounds as outlined in his influential 1988 article, “Those mysterious mounds are for the birds.”36 Green’s article gives good background of how the Megapode theory was first pressed into service to explain the mounds. That redeeming feature aside, the reader needs to take some giant leaps of faith: the concrete cores are explained away as the wholly natural action of microorganisms, tiny globules of calcite in the soil somehow binding rocks and debris together. Later on, it was proposed that the nesting birds also introduced vegetation into the mounds which decayed, providing heat for the eggs. Subsequently, a refinement of the theory proposed that the Megapode tidily excreted into the hole in the top of the mound, the faeces becoming the heating agent for the eggs. Over time, the droppings are presumed to have fossilized and become the “cement” found today.

The Megapode theory failed to account for what qualified archaeologists actually found in the tumuli: the components of the concrete and the iron artifacts around and beneath the core. It typically did not concern its proponents that no trace of bird shell or other material that would normally indicate an avian presence had been found.

Of course, not all scholars accepted the avian theory. Some indicated their tacit acceptance of an anthropic origin for the tumuli, sometimes in regard to the political implications any notion of pre-Melanesian settlement presents. See, for example, John Connell, “New Caledonia or Kanaky? The political history of a French colony.37 Others remained open to a human origin for the tumuli without taking a firm position pro et con. A 1995 paper by Janelle Stevenson & John R. Dodson was open to earlier arrivals, noting the evidence for human occupation 28,000 years ago on Buka Island in the nearby Solomons. They noted the divided opinions on the mounds, stating that “no cultural material or bird remains” had been found, but avoided any direct mention of the cores.38 In 1999, Stevenson noted that human arrivals earlier than the Lapita culture (ca. 3200 years ago), would remain speculative “…as long as a human origin for the so-called tumuli of New Caledonia is entertained…” referencing a paper by Christophe Sand.39 

A 1997 review by Patrick V. Kirch of three 1995 publications by Christophe Sand (until 2019 Director of the Institute of Archaeology of New Caledonia and the Pacific in Noumea), critiqued the multiple earlier scholars whose analysis of data suggested the “specter” of an aceramic, cultural, origin. Sand’s publications were also criticized for not adequately discussing the Megapode theory.40  Kirch essentially maintained his pro-avian position throughout his 2000 book On the Road of the Winds – An Archaeological History of the Pacific Islands Before European Contact. It was republished in revised form in 2017, but while he accepted tumuli dating as early as “10,950 BC” it continued the inaccurate (and unreferenced) assertion that “egg shell” had been found in “the mounds.”41

The avian theory continued to be uncritically accepted when it reinforced more general positions. A representative statement from anthropologist Susan O’Connor argued in 2010 against the 1986 Gibbons & Clunie proposal for Pleistocene voyaging between the eastern edge of the Australian continent and New Caledonia. Their two primary arguments for the voyaging rested on the reduced distances needed for ocean crossings when sea levels were lower and the reported human derivation of a series of ‘tumuli’ or artificial mounds on the interior plateau of the Isle of Pines. O’Connor’s comment regarding the tumuli was that they dated to c. 12,000 BP, [and] have been shown as built by megapodes.”42

Beginning as a well-intentioned attempt at explanation, for many the “megapode” theory became a convenience to avoid dealing with the possibility of a pre-Lapita human presence. But it was a fiction all the same. After all, the actual components of the cement cores have been known since Golson and Chevalier’s work was published and, along with other items we would expect if birds were involved, neither Phosphorous nor any fragments of egg shells have been found in the tumuli.

The avian theory was dealt its final fatal blow on 30 March 2016 with publication of a paper by Trevor H. Worthy et al. titled “Osteology Supports a Stem-Galliform Affinity for the Giant Extinct Flightless Bird Sylviornis neocaledoniae (Sylviornithidae, Galloanseres),” in PLOS ONE.43 Significantly, Christophe Sand was a co-author. Page 53 of the paper discusses the avian explanation for the tumuli (all emphasis added):

Sylviornis neocaledoniae has been identified as a possible maker of large mounds or tumulion La Grande Terre and Ile des Pins. However, there is no evidence of association of this species with these mounds and eggshell has not been found in them, despite the calcareous nature of at least those on Ile des Pins. 

Given that our analyses show that S. neocaledoniae has no enhanced morphological adaptations for digging, such as those displayed by certain megapodes, it is most unlikely that it is responsible for these tumuli.

Instead these tumuli are likely to be formed by an interplay of vegetation and erosion as suggested for similar mounds elsewhere. Even if mounds were associated with eggshell, then the largest of all Megapodius species, the New Caledonian endemic and extinct M. molistructor Balouet and Olson, 1989 would be a more likely contender for their builder.

The paper concluded as follows (incidentally validating the comments made by Golson44 in 1963 about the extinct bird under discussion):

Our finding that it is a stem galliform (and thus not a megapode) makes it most unlikely to have been a mound-builder, as this scenario would require both mound-building and ectothermic incubation to have evolved twice. Furthermore, Sylviornis neocaledoniae shows no specific adaptation for digging to facilitate mound-building, unlike all extant megapodiids, making it even more unlikely that it exhibited ectothermic incubation.


Buried within the text reproduced above lay the author’s replacement theory to explain the tumuli:                          

Instead these tumuli are likely to be formed by an interplay of vegetation and erosion as suggested for similar mounds elsewhere. (emphasis added).


Figure 7A, 7B (2 images): A view showing so-called “mima-like mounds” and a graphic illustrating two hypothesized erosion scenarios resulting in the formation of mima-like mounds. Graphic kindly provided by Nathalie Diaz, University of Lausanne.  



Thus, what could be termed the “Environmental theory” was introduced. In essence, it proposes that “calcite concretion” may have developed around the root complex of a long rotted tree, leading to “differential erosion” in the landscape. Such structures are variously known by names such as “mima-like mounds” or “heuweltjie earth mounds” and are found in various environments around the world. However, an examination of this phenomenon (including sources cited in notes 83-85) reveals stark differences between such formations and the tumuli. They are utterly unlike the tumuli under discussion in this paper. Not only do the former have smaller dimensions, but these formations lack the tumuli’s central concrete pillar, not to mention its central shaft and iron appendages. The suggestion is perhaps even less plausible than the avian theory it replaces.


In 2008, an important development began with publication by Frederique Valentin & Christophe Sand of a paper, “Prehistoric Burials from New Caledonia (Southern Melanesia): A Review.”45 The paper included a detailed description of the Tüü mound on the Isle of Pines excavated by Daniel Frimigacci, noting that the Tüü mound likely had original dimensions approximating the concrete-cored tumuli, but its interior was taken up with 3 levels of human burials; a base level of loose bones, with 2 higher levels each containing a full human skeleton. Dating of the bones yielded a range between 1930 and 1845 BP and appeared to be primary, original, interments. The salient points of the text follow (all emphasis added):

The study revealed two stratigraphic levels containing interments (Frimigacci, 1979, pers. comm. 2000). The lower level, above a basal level dated to 1930+/–70 BP (UW 655, calibrated A.D. 50 (80) 245), contained at least one complete adult skeleton, dated to 1440 +/– 35 BP (UW 766, calibrated A.D. 560 (635) 660). The body was on its right side, apparently in a hyperflexed position, knees against the chest (Frimigacci, 1979:22, figure 15).

The second level contained a complete skeleton of a male about 40 years old (Charpin, 1983). It was on its back in extension.

The surface level included many disarticulated and scattered human bones that Frimigacci (1979) regards as representing incomplete skeletons. A bone sample returned the early date of 1845 +/– 65 BP (UW 765, calibrated A.D. 30 (160) 370).

According to Frimigacci (1979: 21), “ce tertre funéraire est une sépulture collective, il a été édifié au fur et à mesure des inhumations” [the funeral mound is a collective grave which was built up by successive inhumations]…..The inhumations were apparently primary and definitive, displaying variations in positional modes and an expansion of the tumulus by addition of sedimentary levels.

By including this description and dating for a prehistoric burial on the Isle of Pines, the straightforward report effectively again differentiated burial mounds from the concrete-cored tumuli (which were left unmentioned throughout the paper). Surprisingly, this very description became the linchpin of another theory in 2017 when Louis Lagarde published his paper, ““Were those mysterious mounds really for the birds?46 It is probably the most important study of the tumuli published since Chevaliers 1963 report.

Positioning itself, in part at least, as a response to Roger Green’s influential argument in support of the avian theory, the paper reviews the history of scientific enquiry into the tumuli, giving particular credit to the pioneering work of Daniel Frimigacci in mapping tumuli in the 1970’s and also exploring the coastal burial mounds such as the one at Tüü. It then maps the avian hypothesis in some detail, as well as various earlier attempts to explain the tumuli geologically. Results of the 2006-2010 survey led by Lagarde, some never previously published, then serve to demonstrate why the tumuli must have an anthropic origin.

For all this material we can be grateful, but despite it the report works up to a most unexpected detour, for notwithstanding the long-published excavation reports and the author’s own wide-ranging experiences, the report then proposes that the burial mounds are essentially the same as the tumuli, only the construction material varying according to the location. The detour is first introduced by mentioning an oral tradition claimed by a local informant mentioning that “posts were planted on the top of earth mounds and scenes of “torture.” This claim had been reported in 1986 by Frimigacci and Golson,47 and while Lagarde rightly warns about placing too much weight upon it, he nevertheless uses it to introduce his primary conclusion in the paragraph that immediately follows.

In this next discussion, Evidence of funerary practices, the dating of a burial mound in Gadji  to ca. 2309-2003 and 2208 BP is described as “similar” to the Tüü first century AD dating; on that basis the plateau tumuli are then declared to have the same ritual, funerary, function as the earth mound burials. This stunning claim completely ignores the various carbon-dating results and, as we shall see, the uncontested facts established in tumuli excavations. Lagarde is careful to first note a fundamental difference between the burial mounds and the tumuli (one has burials, the other does not), but then introduces the perplexing claim that the tumuli are void of any “archaeological material”:

Given this [dating] data, it seems legitimate to suggest a new interpretation as funerary structures for the earth mounds. One major difference separates the plateau structures from the coral plain ones: in the plateau tumuli, no evidence of archaeological material is to be found, whereas every open tumulus of the limestone environment (through Frimigacci’s or our more recent fieldwork) seems to contain human remains (Frimigacci 1986: 29).

More such claims follow:

…after sixty years of archaeological activity on Isle of Pines, the hypothesis of an ancient human settlement that would predate Lapita by thousands of years should be abandoned. Every dig and survey campaign has only brought back evidence of material culture in connection with already known assemblages from mainland New Caledonia and its global, well-known, diverse cultural traditions of the last three millennia.

it has been impossible to discover any kind of artifact that would indicate a pre-Lapita culture, on Grande Terre or the Loyalty islands. The same goes for Isle of Pines, where no evidence of any culture that would pre-date Lapita has ever been found, after our four-year research campaign.

This hypothesis forces to unite both types of tumuli in the same function, despite the lack of archaeological evidence coming from the plateau ones. This is however very much in phase with the results obtained by Frimigacci. The plateau earth mounds and those of the calcareous seashore plain, are structurally and morphologically identical, their differences coming only from the environment in which they were erected.

Typologically, they are identical and linguistically, the people of Isle of Pines call both kinds purè: furthermore they stand in similar proportions in a population/space ratio, as hundreds of them are to be found in both chrome and coral environments. If one of the types was much rarer, if language could differentiate the structures, if the inner structure of the earth mounds was indeed typologically different, we would have separated the tumuli into two categories. On the contrary, we stand confident with the hypothesis of a unique anthropogenic tradition in which minor differences are to be explained by the surrounding environment.

This is not our idea, as Frimigacci was the first to point out the similarities existing between the two populations of tumuli, and to try to explain the use of the plateau ones by the contents of the seashore ones (Frimigacci 1986:32). His archaeological excavation of the human remains in the Tüu tumulus is for us the key finding of the past sixty years.

Arguing at some length that human bones can completely vanish under conditions of soil acidity and exposure to the elements via the “tiny limestone blocks” covering the tumuli, Lagarde concludes:

  …we suggest the possibility that the tumuli are of anthropic origin, bear a ritual/funerary function and that at least some of them were used as funerary mounds for primary burials.

We hope to have brought forward that the tumuli of the Isle of Pines are more archaeologically complex than straightforward, natural, bird-made structures. Although being the last pieces of evidence that could have supported the theory of a extremely ancient human settlement of the New Caledonian archipelago, they actually belong to the recent three millennia of Austronesian presence. In our opinion, they illustrate, at least partly, the funerary practices of a time span of five to six centuries, between the end of the first millennium BC and the first centuries AD.

How can we reconcile the fact that the interiors of the burial mounds and the tumuli are, in truth, actually radically different? What can account for such a disconnect between the uncontested archaeological evidence and such explanations as this report offers? Are we seeing the rise of a new default position to replace the discredited avian theory? If so, then as with all the other attempts at explanation the tumuli it comes at the expense of the archaeological evidence. As at the time of writing such questions remain unresolved.48      


The author’s explorations on the Isle of Pines commenced in 2017 and first undertook a detailed examination of the original tumulus excavated by Chevalier. This was to reveal that this concrete core had a circular vertical shaft throughout its length. As Chevalier’s excavation report did not mention such a significant feature this unanticipated find was initially perplexing, but it was concluded that his excavation had not cleared the entire summit of the tumulus before work was abruptly halted. Thus, the opening at the very centre of the top of the core—1 foot/30 cm in diameter—was missed by him. After Chevalier returned to the main island, local workers continued removing the mound’s iron gravel as a source of road-making material until only the concrete core remained but the shaft would have been of little interest even if noticed. In fact, as already noted, Jack Golson’s 1959-1960 survey reports had mentioned this feature in many of the tumuli he and his team had examined. In his notes, Golson designated the concrete cores as “anchors” and the shafts as “postholes,” but could not suggest what purpose the latter served.49 


Figure 8: Chevalier’s 1963 excavation report (see Fig. 5) redrawn with the circular shaft in the concrete core added.


Figure 9A, 9B (2 images): Views looking down and up showing the 30 cm wide circular shaft running vertically through the middle of the concrete core of the tumulus excavated by Chevalier in 1959. Photographs by the author 2018, 2020.



From abundant large excavation debris still surrounding the tumulus, the author obtained samples of the interior concrete and submitted them for elemental composition analysis in an Australian laboratory.50  Following is a comparison of the results from Chevalier’s analysis of the first tumulus excavated on the Isle of Pines with the author’s 2018 analysis of the same structure, of the interior mortar in both cases:

    Chevalier Analysis/Author Analysis

Loss on ignition % 33.20 41.8                         

Fe2O3 26.57 6.6

Fe     18.60       nil recorded

Al2O3     4.00 0.4

CaO 34.25   48.2

MgO traces 0.7

SiO2   1.80       < 0.05

SO3 not recorded 0.1

MgO not recorded 0.7

K2O not recorded 0.03

Na2O not recorded       < 0.05

P2O5 not recorded       < 0.05

TiO2 not recorded         < 0.05

Mn2O3 not recorded 0.08

SrO not recorded 0.1

Cl not recorded   0.011


A construction sequence can now be proposed that incorporates all the features of the tumuli now known. It is based upon Luc Chevalier’s 1963 proposal for how the cement core was built on the Isle of Pines, but substantially expands and updates it.


Figure 10: Construction sequence of the cement core on the Isle of Pines as proposed by Chevalier, 1963 [his note 6].


Figure 11: Author’s proposed construction sequence in cross-section, incorporating all known features.


Stage 1

The necessary first step required excavation of a hole through the topsoil and more than 2 meters deep into the ferruginous clay layer beneath.

Stage 2

This was followed by placement in the hole of the almost 8 foot/2.4 meters long Fe cone-shaped object, its point downwards and with at least 3 rings of Fe nuggets placed around it, perhaps to stabilize it in a vertical position. Their presence, incidentally, makes it clear that the cone-shaped object is not an artifact from iron being leached from the surface and somehow precipitating into a geometrical shape below the tumulus.

Stage 3A

Vertical placement of a tubular pillar or “pylon” (a smooth tree trunk?) of unknown length centered above the cone. Current excavation data lacks the resolution to determine which of these two possibilities is correct: (A) the pylon may have been placed in direct contact with the rounded top of the cone, or:

Stage 3B

  1. the pylon may have been separated from the top of the cone by a shallow bed of Fe nodules and soil, the scenario utilized in this model.

Stage 4

Over a bed of medium-sized Fe nodules, the first cement layer was poured into the hole, thus stabilizing the pylon in position. At ground level the base of the cement block was surrounded (delineated?) by large Fe nodules, while a circular base layer of Fe gravel was added around the cement core to the same height.

Local materials, consisting primarily of iron gravel but inevitably also some soil, were scraped up in a circular fashion around the structure, thus differentiating the scraped area from the normal surroundings, creating the “aureole” effect still visible in aerial images, as Chevalier noted. Throughout its erection, the outside surface of the cement core was reinforced with various sized Fe nodules, in contrast to the softer cement interior.

Stage 5

The second layer of cement was added, maintaining the same dimensions as the first layer. Likewise, the Fe gravel was built up around the core to the same level.

Stage 6

The third and final layer of cement was added, maintaining the same dimensions and the Fe gravel built up around the core to the same level.

Stage 7

A final, substantial, layer of Fe gravel was added to entirely cover the cement core, completing the rounded shape of the tumulus. The small size and particularly the weight of the gravel is such that it would have rapidly filled the hole in the cement core, assuring us that the pylon, or at least a “cap” of some type, was in place when the gravel covering was added to the tumulus. The simplest explanation is that the pylon remained in place throughout.

Stage 8

It is unclear whether a final, thin, layer of soil was added to cover the entire tumulus or whether the soil already mixed in with the gravel sufficed. In either event, the tumulus was now complete with only the pylon protruding from the hole in the cement block.

Stage 9

At some point the pylon was removed (or if organic, such as a tree trunk, perhaps rotted away) and the tumulus assumed its innocuous modern appearance, giving no hint of the substantial cement and iron structures within and below it.

At most tumuli in the present, the vertical hole in the cement core is covered over by the vegetation growing in the tumulus covering; for that reason, the hole was not readily visible to early investigators. Data are currently insufficient to ascertain whether the shafts became clogged with debris after removal of the pylon. While apparently intended to provide additional stability and strength, the very substantial covering of iron gravel and soil also effectively served to protect the tumuli from weathering and disguised their nature until very recent times.51 


Of the numerous reasons for building any structure elevated above its surrounds (to provide a higher vantage point, to elevate something for increased visibility, to use gravity to move something (eg. water) downwards, to funnel something (eg. smoke) upwards and so on), only a small number are potentially applicable to the tumuli. These are as follows:

1. A burial site. As demonstrated earlier, whatever else they may be, the structures are manifestly not burial mounds. No skeletal material, or the cultural artifacts associated with burials, have yet been observed in the concrete-cored tumuli. Human burials, of course, are found in a variety of mounds or sub-surface sites that do not resemble in any way the structures under discussion here. The human burial mounds are smaller in height and extent, have no cement core at their centre and are generally found in situations and locales more commonly associated with funeral remains.

Purely as a note of interest, in 2019 the author and others examined a large concrete-cored tumulus on private land at Gadji which had been disturbed by road construction; a small number of loose human bones were located just beneath the surface (a depth of ca. 15 cm), an act of later, accretional, burial using the surface of the mound as a convenience. Here, in a classic situation of an apparent exception proving the rule, the obviously recent very shallow burial helps emphasize that these structures were not originally intended to inter human remains, a distinction readily apparent to most observers over the years.52


Figure 12A, 12B (2 images): The relatively recent burial just below the surface of a large tumulus in Gadji. The damaged concrete core in the centre of the tumulus and the burial location on the right side are both visible in image 12A. Images by author, 2018.


2. As an elevated vantage point or watchtower. This proposal lacks any merit as abundant opportunities for higher views are afforded by adjacent land features and trees. And, in any case, it does not explain the cement core construction and the sheer number of tumuli built.

3. As a ceremonial or ritual centre. This concept finds some support in a handful of locations where rock-lined paths and other cultural indicators have been found. Most recently, the 2006-2010 survey on the Isle of Pines identified pathways, depressions and large ferralitic boulders atop 6 tumuli. While ceramic sherds remained “extremely rare” on the Iron Plateau, indications of eaten seashells from the coast or “fossil uplifted limestone blocks from the forest” were noted at 8 tumuli.53 Investigation is still required to determine if these works are contemporaneous with the tumulus, or a later expansion by others. Logically, the likelihood that the stones date to a much later period seems to be strengthened by their rarity. 

The oral tradition noted earlier that mentioned that “posts were planted on the top of earth mounds and scenes of “torture”” does not offer support for ritual use any more than it does for a funerary function; a post or tree trunk does not require a large concrete base to ensure it stands upright. If these traditions have any substance they must refer to a usage postdating the tumuli construction.

4. As an artistic expression. The most obvious objection to the structures being an art form that developed uniquely in the area, is that their most impressive features – the interior cement block surrounded by iron nodules – was completely covered over with earth and clearly not intended to be visible. Nor would a circular earthen mound require such an interior, one that remained unseen. We would expect that an object of art would be adorned with other features, either individualistic or symbols common to the community. Instead, with the exception of the few locations noted in #3, we find only conformity, an austere plainness that suggests a strictly utilitarian, functional, purpose, not an artistic one.


By showing what the tumuli are not and cannot be, the purpose of these structures can now be proposed with some certainty. From an engineering viewpoint the large cement blocks within the tumuli have the primary purpose of stabilizing a circular pillar or pylon, standing in the vertical shaft of the core but later removed or decomposed, for no trace of them now remains.Nor is there any hint of what the pillar/pylon may have supported and why it  required such a degree of stabilization, as evidenced by the massive cement base. Determining this is now the challenge going forward. Ascertaining the function of the cone-shaped metal object beneath the shaft may offer the most promising line of future investigation.

Nothing to date suggests that the tumuli construction took place over an extended period where we would expect to see improvisations or the development of different styles. Everything suggests a one-time concentrated effort.

Only two of the Paita tumuli preserve any trace – surplus raw material – of the construction process and, just possibly, of a stage of early experimentation with local materials. If so, we can speculate that the silica-based process may have been abandoned in favor of the iron resources readily available on the Isle of Pines.


More answers will reveal themselves when modern archaeologists engage more fully with the structures than they have heretofore. It would require little effort to completely reveal the interiors of those structures that are already partly and even fully exposed; in particular, the large cone-shaped iron structure which is the most likely feature to fully reveal the final answers to the mystery. Was this a type of anchor, such as our modern anchor spears, for the pylon or pillar sitting directly above? While this artifact has only been reported from one site, it beggars belief to think that Chevalier chanced upon the one tumulus that had a symmetrical cone-shaped metal object positioned exactly below the shaft in the central core; it must be an integral part of other tumuli. For that matter, we cannot be sure that nothing else lies beneath the cone itself – no-one has looked.

The incongruity of the scientific establishment, particularly archaeologists and historians, in having Chevalier’s excavation reports since 1963 while still proposing solutions that do not, in any way, account for what lies within these structures is staggering. The tumuli, after all, remain readily available for examination. This inaction must represent, in large part at least, the instinctive avoidance of something that challenges present paradigms about human settlement in the southern Pacific.

Surely the time is long past when data seemingly at odds with our current thinking are ignored. Perhaps in them lies a key to a broader understanding of what influences and events have shaped the Pacific’s distant past even if, as the data in his paper suggests, it requires a reconsideration of what we can possibly term the “Neolithic” explanation. It is hoped that this paper will encourage archaeologists and historians of this region to re-engage with facts and move beyond demonstrably inadequate theories in solving the puzzle these structures still represent.

In the 21st century, the enigma remains. And if Chevalier’s questions were perceptive in 1963, decades later they can be re-stated and expanded:

The identity of the tumuli builders remains, of course, the crux of the mystery, but they certainly predated the arrival of the Melanesians who became the generally regarded indigenous culture and had no ability to manufacture concrete. Who were they? Where did their ability to make high-quality concrete come from at a time when this technology was unknown, certainly in the Pacific at least?

What possible motivation or purpose could cause a people to invest so much effort to erect more than 400 of these structures, each averaging a volume of 500 cubic meters?54   

Why did such a massive effort leave no other traces of its life and activities—no tools, bones, charcoal, pottery or other cultural artifacts, either within the tumulus or around it?

Why did they not use this ability to construct other works, their dwellings, their burials and their sacred places before, during, or after the tumuli construction?

Over time, why was the ability to manufacture such a useful material as concrete not exported to other islands? Why was this capacity not passed down to the Melanesians who became the generally regarded indigenous culture? Indeed, why did this capacity not arise elsewhere in the region?

The prescient questions of Chevalier, pondering the mystery of a distant, unknown, people who knew how to make durable concrete on a large scale, still await resolution.



Conflicts of Interest



About the Author

Warren Aston is an independent researcher based in Brisbane, Australia. He studied archaeology at the University of Queensland, and has been involved in archaeological projects in Mexico and Oman over several decades. He can be reached at: 






1.“The [Isle of Pines] is considered to be a kind of “archaeological nightmare” due to the presence of no less than 200 earth mounds…” Louis Lagarde & Andre-John Ouetcho, “Rocks, pottery and bird bones: new evidence on the material culture of Isle of Pines (New Caledonia) during its 3000-year-long chronology,” in Christophe Sand et al. eds. The Lapita Cultural Complex in time and space: expansion routes, chronologies and typologies (Noumea: IANCP, 2015), Archeologia Pasifika 4 series.

2. Jacques Avias, “Contribution a la Prehistoire de L’Oceanie: les Tumuli des Plateaux de Fer en Nouvelle-Caledonie” in Journal de la Societe des Oceanistes (1949), 15-50 (author transl.) Emphasis added.

3. Louis Lagarde, “Geological anomalies?” in ““Were those mysterious mounds really for the birds?” Reappraising the Isle of Pines’ puzzling tumuli (New Caledonia)” in For a history of Oceanian prehistory: historiographical approaches to French-speaking archeology in the Pacific (Noumea: Pacific CREDO publications, 2017). Emphasis added.

4. Luc Chevalier, “Le problème des tumuli en Nouvelle-Calédonie,”Bulletin périodique de la Société des Études mélanésiennes 14-17 (Noumea: Société des Études mélanésiennes, 1963), 24-42. Emphasis added. The section titled “Tumulus A Cylindre Double” reports the examination of an oval double-cored tumulus and notes two others visible in aerial images.                                                                          Three related reports by Chevalier are also on file at the New Caledonia Museum in Noumea: “Observations faites sur un tumulus ouvert a LIIe des Pins” (3 pages) dated October 6, 1959; “Note sur les tumuli de Nouvelle-Caledonie” (3 pages dated October 21, 1960 regarding the mortar analysis) and a 5 page handwritten draft dated January 15, 1961 of the article that would appear in 1963.

5. Chevalier (1963), Figs. 4, 5. 

6. Chevalier (1963). See Fig. 7 sketch in “Tumulus A Cylindre Double.”

7. Chevalier (1963), Fig.10.

8. Chevalier (1963), Fig. 5.

9. See Chevalier (1963) Figs.10 and 11 with notes.

10. Roger Curtis Green & J. S. Mitchell, “New Caledonian Culture History: A Review of the Archaeological Sequence” in New Zealand Journal of Archaeology vol. 5 (1983), 19-68. Emphasis added.

11. Theophile Auguste Mialaret, L’île des Pins, son passé, son présent, son avenir: colonisation & ressources agricoles (Paris, J. André, 1897). Reprinted in 2012 (Paris: Hachette Livre BNF). Emphasis added.

12. R. H. Compton, “New Caledonia and the Isle of Pines,” The Geographical Journal vol. 49, no. 2 (London: Feb. 1917), 81-103, esp. 102-103. Emphasis added. 

13. Jacques Avias (1949), 15-50. Emphasis added.

14. Jack Golson, “The Tumuli of New Caledonia” in “Report on New Zealand, Western Polynesia, New Caledonia, and Fiji,” Asian Perspectives vol. 5, no. 2 (Winter, 1961), 170-172.

15. Jack Golson, “Rapport sur les fouilles effectuees a l’ile des Pins (Nouvelle Caledonie) de Decembre 1959 li Fevrier 1960,” in Bulletin périodique de la Société des Études mélanésiennes N.S. 14-17 (1963), 11-24; Daniel Frimigacci & Jack Golson, “Tout Ce Que Vous Pouvez Savoir Sur Les Tumulus, Meme Si Vous Osez Le Demander” (1986), copy in author’s possession. Emphasis added.

16. Chevalier (1963).

17. Chevalier (1963), Fig. 5 in “Fouilles Sous Le Cylindre.”

18. Chevalier (1963), see “Tumulus A Cylindre Double.”

19. Chevalier (1963), see “Elevations Secondaires” and “Fouilles Au Tumulus no 2.”

20. G. Delabrias et al. “GIF Natural Radiocarbon Measurements 11,” Radiocarbon, vol. 8 (Yale: The American Journal of Science,1966), 88-89 (Gif-298, 299 and 300).

21. T. A. Rafter et al. “New Zealand radiocarbon reference standards” Proceedings of the 8th International Conference on Radiocarbon Dating. Vol. 2, (1972), 625-675, as referenced in Lagarde “Were those mysterious mounds really for the birds?” (2017). The same paragraph cites Dubois (1976) for the 12,900 ybp Placostylus dating although the dating was actually from Chevalier, as reported in G. Delabrias et al. (1966) – see note 20.

22. H. C. Brookfield & Doreen Hart, Melanesia: A geographical interpretation of an island world (London: Methuen, 1971), 78 references both Chevalier and Golson’s work and summarized: “…there seem to have been earlier – perhaps much earlier – inhabitants of New Caledonia, who built tumuli of a concrete made of coral lime…” Emphasis added.

23. Fr. Marie-Joseph Dubois, “Trouvailles à l’île des Pins, Nouvelle-Calédonie,” Journal de la Société des Océanistes 51-52 (1976), 233-239. See especially the section Purè. Emphasis added.

24. Richard Shutler Jr. “Radiocarbon dating and Oceanic prehistory” in Archaeology and Physical Anthropology in Oceania 13, 2&3 (1978), 222 accepted that “…by 10,000 years ago, a non-Austronesian, aceramic, pre-Neolithic, tumuli-building people were in Island Melanesia, on New Caledonia and Ile de Pines…” Emphasis added.

25. Peter Bellwood’s opus Man‘s Conquest of the Pacific (Auckland: Collins, 1978), 250 concluded: “The Isle of Pines also provides one of the more unusual archaeological mysteries of the Pacific….[dating gives] a very odd range of dates between 1000 and 6000 BC. which may be of questionable value.” However he notes: “Furthermore, it is not at all certain that the mounds are man-made, but if they are, then perhaps a Lapita connection may be found, or they may even have been built by some unknown pre-Lapita inhabitants of New Caledonia.” Emphasis added.

26. Roger Green & J. S. Mitchell (1983) sensibly accept that the concrete cores are an important indicator of a “cultural” origin, dismissing attempts to label them as “natural” by highlighting Chevalier’s analysis of the concrete. They noted that some snail shells attached to the cores and in the cement matrix provided the oldest dating and believed some dating inaccuracies might exist, but, despite preferring a later dating, still accepted that the tumuli are manmade and likely ancient: “…the enigmatic tumuli, or concrete-cored mounds, provide a number of issues, the answers to which remain uncertain. However, there are several lines of evidence to suggest that these structures are man-made…at present these tumuli present a problem which only further archaeology and more data will resolve…it seems necessary to continue attributing at least some of these tumuli with their cylinders and other features to human activities, as they provide a range of evidence not easily explained by any set of natural events…It is only if some much earlier age estimates for tumuli are adopted that they might be taken to imply initial occupation by a group of non-Austronesian speakers.” All emphasis added.                                                                                                                                        

27. Kerry Ross Howe in Where the Waves Fall (Honolulu: University of Hawaii Press, 1984), 31 states: “New Caledonia poses one of the more interesting problems in Oceanic archaeology…the conical tumuli with an inner cylinder composed of hard lime mortar. They are clearly of human makingThree radiocarbon dates put their age at about 7000, 9500, and 13,000 years…These amazing dates may simply result from the testing of material that was long dead before being used to build the tumuli. On the other hand it is possible that the dates are correct.” Emphasis added.

28. John R. H. Gibbons & Fergus G. A. U. Clunie, “Sea Level Changes and Pacific Prehistory: New Insight into Early Human Settlement of Oceania” in The Journal of Pacific History vol. 21, no. 2 (April, 1986), 58-82 accepted that at least some of the 400 tumuli were “apparently man-made.” Emphasis added.

29. Christophe Sand, “Lapita and non-Lapita ware during New Caledonia’s first millennium of Austronesian settlement” in The Pacific from 5000 to 2000 BP (Paris: Institut de recherche pour le developpement, 1999), 139-159. Sand treated the dating as established: “The discovery of large tumuli, dated between 13,000 and 4000 BP, was at the same time seen as indication of an old settlement of southern Melanesia by modern humans…,” later noting “The significance of this dating remains unclear…” Emphasis added.

30. Patrick Vinton Kirch, On the Road of the Winds – An Archaeological History of the Pacific Islands Before European Contact (Berkeley: UC Press, 2000 and 2017), 135 accepted tumuli dating as early as “10,950 BC” while still maintaining support for the avian explanation. Emphasis added.

31. Avias (1949), 48.   

32. Frimigacci & Golson (1986), 1. 

33. For example, see Lagarde, “Isle of Pines and the tumuli research history” in “Were those mysterious mounds really for the birds?” (2017), noting particularly publications of engineer Bernard Brou, whose 1977 book, Préhistoire et société traditionnelle de la Nouvelle-Calédonie (Noumea: Publications de la Société d’Études Historiques de la Nouvelle-Calédonie) served to popularize the migration theories of Avias, although Brou himself had favoured a geological explanation. 

34. In 2018, the author published a response to the 2017 release of alleged US intelligence documents that included a claim that traces remained on the Isle of Pines of an extraterrestrial visit ca. 11,500 BP. Such an explanation will seem fantastic in direct proportion to how current one’s knowledge is of astronomical advances; the author retains an open mind.

35. Francois Poplin & Cecile Mourer-Chauvire, “Le Mystere de Tumulus de Nouvelle-Caledonie,” in La Recherche 16 (Paris, September, 1985), 1094. The same authors published “Sylviornis neocaledoniae (Aves, Galliformes, Megapodiidae), oiseau Géant éteint de l’ile des Pins (Nouvelle- Calédonie),” Geobios vol. 18, no. 1 (December 1985), 73-105. science/article/pii/S0016699585801820 All emphasis added.

36. Roger Green, “Those mysterious mounds are for the birds,” Archaeology in New Zealand vol. 31, no. 3 (1989), 153-158. Emphasis added.

37. John Connell, New Caledonia or Kanaky? The political history of a French colony. Pacific Research Monograph 16 (Canberra: Research School of Pacific Studies, ANU: 1987), 1-2. Emphasis added.

38. Janelle Stevenson & John R. Dodson, “Palaeo-environmental Evidence for Human Settlement of New Caledonia” in Archaeology in Oceania vol. 30, no. 1 (1995), 36-41. Emphasis added.

39. Janelle Stevenson, “Human Impact from the Paleo-environmental Record on New Caledonia” in Jean-Christophe Galipaud & Ian Lilley, eds. Le Pacifique de 5000 h 2000 avant le present / The Pacific from 5000 to 2000 BP (Paris: Institut de Recherche pour le Developpement, 1999), 251-258. Emphasis added.

40. Patrick V. Kirch, “Resolving the Enigmas of New Caledonian and Melanesian Prehistory: A Review,” Asian Perspectives vol. 36, no. 2 (Fall 1997), 232-244]. An email exchange between the author and Kirch in 2017 served to emphasize his support for the avian hypothesis. Emphasis added.      

41. Patrick Vinton Kirch (2000 and 2017). The egg shell claim appears on p.135 in the 2007 edition.

42. Sue O’Connor, “Pleistocene Migration and Colonization in the Indo-Pacific Region” in Anderson et al. eds. The Global Origins and Development of Seafaring (Cambridge: McDonald Institute for Archaeological Research, 2010), 46. 

43. Trevor H. Worthy et al. “Osteology Supports a Stem-Galliform Affinity for the Giant Extinct Flightless Bird Sylviornis neocaledoniae (Sylviornithidae, Galloanseres),” PLOS ONE 11(3): e0150871 (2016). All emphasis added.

44. Golson (1963).

45. Frederique Valentin & Christophe Sand, “THE TÜÜ BURIAL MOUND, ISLE OF PINES” in “Prehistoric Burials from New Caledonia (Southern Melanesia): A Review,” Journal of Austronesian Studies 2 (1) (Taiwan: National Museum of Prehistory, June 2008), 5. All emphasis added.

46. Louis Lagarde, ““Were those mysterious mounds really for the birds?” (2017). All emphasis added.

47. Frimigacci & Golson (1986), 31.

48. Louis Lagarde did not respond to multiple queries in early 2020 seeking clarification of his claim that the plateau tumuli contain “no evidence of archaeological material.”

49. Discussed in Golson (1961), 171. Additional instances where these features were found appear in unpublished excavation notes of Jack Golson in possession of the author.

50. Cement Australia, Darra Laboratory, Queensland, NATA No. 188, dated July 19, 2018.

51. A basic comparison of equal quantities by the author yielded these results: 125 g of iron gravel was heavier than dry soil at 91 g, and even soil fully saturated with water at 113 g.

52. A review of earlier literature discussing the tumuli will reveal how often the clear distinction between burial mounds and the tumuli was apparent to skilled observers. One of the earliest such accounts, for example, is Alphonse Riesenfeld’s wide-ranging Megalithic Culture of Melanesia (Leiden: Brill, 1950), 539 – 572, an interesting commentary on the times and while dated, still of value.

53. See Lagarde (2017) “Anthropic aspects of the plateau tumuli” in “Were those mysterious mounds really for the birds?”in Lagarde (Doctoral Thesis, 2012), Peuplement, dynamiques internes et relations externes dans un ensemble géographique cohérent de Mélanésie insulaire: L’exemple de l’Ile des Pins en Nouvelle-Calédonie (Paris: Université Paris I – Panthéon-Sorbonne), Tome 1, espec. 244-245). Emphasis added.                                                                       

54. Lagarde provides an estimate of tumuli volumes as averaging about 500 cubic meters; see Lagarde, “Geological anomalies?” in “Were those mysterious mounds really for the birds?” (2017).


Warren Aston is an independent researcher based in Brisbane, Australia. He studied archaeology at the University of Queensland, and has been involved in archaeological projects in Mexico and Oman over several decades. He can be reached at:

Subscribe to Popular Archaeology Premium. Still the industry's best value at only $9.00 annually.