WWW Resources on the Marshall Islands

MARSHALL ISLANDS - Coral Gravel Spreads


Anthropology - Marshall Islands -Settlements

Archaeology - Marshall Islands -Settlements

Architecture - Marshall Islands -Settlements

Geography - Marshall Islands -Settlements

Marshall Islands - Anthropology - Settlements

Marshall Islands - Archaeology - Houses

Marshall Islands - Archaeology - Settlements

Marshall Islands - Architecture - Traditional Houses

Marshall Islands - Geography - Settlements

Marshall Islands - Mile Atoll - Archaeology

Marshall Islands - Mile Atoll - History


The report reviews the present knowledge on coral gravel spreads or scatters, the most common archaeological site type in the Marshall Islands. Hithertoprehistoric and historic gravel spreads could not be distinguished. A thesis to distinguish the two is advanced, based on a study of ethnographically recorded and modern housing, as well as based on the observations of burnt-down houses.The dating is based on the fact that modern houses are built close to the ground and coral gravel is not spread underneath them. Traditional houses onthe other hand had no walls and coral gravel had been spread underneath the roofed area. Thus traditional coral gravel spreads are likely to represent acontinuous scatter of coral gravel, while modern scatters are likely show up agap where the house once stood.


By far the most common archaeological feature in the Republic of the Marshall Islands are scatters of coral gravel, commonly oval to roundish in shape. These scatters, then as well as today, signify the presence of a household. In there cent past several surveys of islands in the Marshall Islands have been conducted and such scatters have been found. Until now they could not be positively identified as historic or prehistoric unless datable material was found on them.

This study is an attempt to distinguish between prehistoric and historic coral gravel spreads. In May 1991 three guest houses burnt on Mile Island, Mile Atoll, probably as a result of arson. The charred remains of these guest houses offered the unique opportunity to study the "archaeological" imprint of modern coral gravel spreads and houses. The observations gained on this occasion seem to point a way to distinguish historic coral gravel spreads from prehistoric ones. In this study we will set out the available evidence. We will first have a look at the nature of the coral gravel spreads and where the raw material can be obtained. We will then look at the archaeological as well as the historical record and what has been recorded about these sites. In the subsequent section we will deal with the body of literature and observations on the traditional buildings as well as on the modern building patterns. A case study of some burnt houses on Mile Island follows. This study sets out the archaeological remains of such modern thatched houses and the implications for the recognition of prehistoric houses. In addition it discussed in detail the observations on the coral gravel spreads and how they pertain to the interpretation of archaeological sites in the field.

Figure1. Schematic representation of a semi-traditional to modern Marshallese household. 1- Coral gravel spread; 2 - Living/sleeping house; 3 - water cistern; 4 -cooking house; 5 - open fireplace; 6 - outhouse; 7 - rubbish area; 8 - bananapatch.

Coral Gravel Spreads - nature and origin

In early European time coral gravel spreads surrounded all Marshallese household. Today, gravel spreads surround all those household that are not surrounded by concrete (plates 1, 2). The reason for the coral gravel spread is simple. It creates an area which can be kept clean by pulling out the occasional intruding weed, and by avoiding sand and dirt from being carried into the house from the surrounding area. For the same reason coral gravel has been spread around traditional wells accessing the ground water lens (plate 3). The coral gravel also visibly demarcates the area of the household used by the inhabitant of that household as their yard. Hence mats, clothes and the like, can be aired and dried without fear of dirtying them. The area covered with gravel can be demarcated further by coral slabs set in an upright fashion as low retaining walls and borders.

Coral gravel can be collected in the supra-tidal and inter-tidal zone of most oceanward beaches, where it forms a distinct zone, sorted to sediment size by the wave action. Then, as well as today, this material is brought in from the shore and spread around the houses. As the gravel is, over time and very gradually, trodden into the ground, more clean gravel is brought in and spread about. Thus, gradually, the coral gravel spreads increase in thickness. Let us now look at the configuration of a household. Plates 27 and 28 show a coral gravel spread, lined with vertically set coral slabs, being repaved by the introduction of clean, white gravel. The photographs of this repaving process were taken in July 1992 on Ebon Atoll.

A semi-traditional to modern Marshallese house-hold (figure 1) consist sof a living/sleeping house (2) with an associated water cistern (3), an outhouse (6), possibly a separate shower cubicle, a cooking house (4) with an open fireplace outside (5). All of this is situated on a coral gravel spread(1), which at one end merges with a small-scale rubbish and food refusal dump(7).

In the archaeological record with all organic material decayed, this lay-outis reduced to the coral gravel spread and the midden area, which is commonly characterized by a high concentration of gastropod shells, namely Strombusluhuanus. With increasing land clearing using machinery, the distinction between coral gravel spreads and the midden area becomes more and more obscured.

The archaeological and present record

By far the most abundant type of site on Majuro Atoll as well as other atolls of the Marshall Islands is the coral gravel scatter upon which the houses were built. Such scatters have been reported from Ailinglaplap, Arno, Aur, Ebon, Lae, Likiep, Majuro, Maloelap, Mile, and Wotje Atolls (Rosendahl 1987; Spennemann 1990; unpubl. field notes).

These sites consist of irregularly shaped, though commonly roundish scatters o fcoral gravel of 20 to 60mm diameter. The gravel spread is anything between 0.1 and 0.3m thick and measures between 10 and 30m in diameter. In most cases seen the gravel spread does not form a visible elevation above the surrounding ground. Artificial elevation does not appear to be common, if present at all. In some agriculturally used areas of Majuro I. (Laura), the scatters have been ploughed and have been dispersed. Commonly the scatters show at one side aninter mixture with gastropod and bivalve shells, predominantly Strombusluhuanus .


The botanist Hathaway (195 3:49) excavated part of coral spread in the centre of Arno Island in order to determine its thickness in an attempt to date the occupation in the interior of Arno. Based on his local informants, a coral gravel spread increases in thickness about 2.5 cm (1 inch) every decade. Basedon this and on the observation that the gravel spread was between 12 and 19 inches thick, he assumed that the gravel spread excavated by him represented an occupation of 120-190 years duration at the time of abandonment.

Although his approach is doubtlessly a paradigm in deduction there are a few problems with his approach which may invalidate his results. Problems with this approach are mainly that vertical displacement up to 0.3m can occur by trampling in soft substrate, such as sand, which is ubiquitous on the atolls, and that the amount of coral spread depends on the efforts expended and hence on the status of the person living at the site. In addition, there is differential distribution of material after abandonment to be considered.


It appears that there is little difference in the gravel spreads dating to prehistoric times (as evidenced by 14C determinations) and thosewhich are still being used. No visual distinction could be made between either of them. In both cases the coral gravel appears to have been hand selected on the ocean shore and transported to be scattered around the house(s). The gravel permits to keep the area around the houses free of weeds and other vegetation and also permits the interior of the house to be cleaner as no or only very little dirt is carried inside. Apart from dating the sites by means of14C, there is little one can do assess their age, unless clearly historic artifacts are common in the scatters.


On some, though rather rare, occasions shell middens were encountered by themselves, that is, without evidence of a coral gravel spread in immediate vicinity. These shell middens consisted predominantly of Strombusluhuanus, with some Gafrarium pectinatum, Lambis lambis, Turbo sp., Cypraea sp., Asaphis sp. and Cerithiumsp. and an abundance of other species in very small numbers. Seen were also shell fragments of Hippoppus hippoppus, Tridacna maxima andT. squamosa, which were both used for the manufacture of tools and formed a food resource. Trochus sp. a shell very prominent in modern shell middens and rubbish dumps is conspicuously absent, indicating that it either may be of later (Japanese ?) introduction to the Marshall Islands or was not seen as favored food shell. Reasonably rare are also remains of Cassis sp., which ethnographically were used to make tools (gouges) as well as cooking vessels. Conus sp. shells, commonly used for armringsare also rare in the middens seen. Of all shells (by number), Strombusmakes up about 95%, while by shell weight it is still some 70% (Tridacna shells being particularly heavy).


Related structures are the um pits or fire pits, which are commonly only encountered by excavation, be it in trash pits, cable trenches or by purposeful testing . These pits measure between 0.3 and 1.0m in diameter, and between 0.3 and 0.6 m in depth. They are commonly filled with ash, shell debris and broken coral rubble showing evidence of heating (grey coloring). In addition, somepits are encountered in construction trenches, whose chronological affiliation cannot be determined beyond reasonable doubt, but which appear, for the lack of historic material in their fill, pre-European (figure 2).

Figure2.Profile of a pit, site MI-Mj-279

Figure3. Distribution of prehistory or historic (pre-World War II) coral gravelspreads on Majuro Atoll.


The spatial distribution of coral gravel spreads has been investigated ingreater detail only on a few atolls, one of which is Majuro (Spennemann1990). The distribution of gravel spreads (figure 3), which are clearindicators of settlement, is somewhat skewed. The majority of the spreads islocated on Majuro Island, which is predictable given the size of the island, the breadfruit and presence of a taro zone. A secondary distribution centre islocated on the larger islands on the northern shore, namely Rongrong, Calalin, Enigu etc. There is a clear absence of sites on the southern side of the atoll, which could be due to the effects of the typhoon of 1918, and a lack of sites in the D-U-D area, which could be due to modern modification. Despite the potential bias in the distribution pattern, the absence of sites on the small islands shows that permanent settlement, indicated by the construction of gravel spreads, is restricted to large islands.

Figure4. Frequency distribution of prehistoric and early historic site types onMajuro Atoll.

Traditionaland modern house construction

Let us now look at the traditional and the modern way of constructing houses in the Marshall Islands. The traditional house construction has been described by a number of ethnographic sources, such as Erdland (1906; 1914); Finsch (1893); Krämer (1906); Krämer & Nevermann (1938). The following has been extracted from these sources and unless so specified there is little disagreement between the sources. Figure 5 shows most of the individual parts referred to in the discussion. Table 1 gives a comparative glossary of all Marshallese terms recorded in conjunction with the individual parts of the houses, as well as the modern spelling according to Abo et al. (1976), as adopted by the Marshallese Language Commission.

Figure5. Marshallese house. Construction details (for Marshallese terms refer totable 1).

Table 1 Glossary of Marshallese terms for parts of the house.

                       Steinbach-Gr     Erdland      Krämer & Nevermann   Abo et al.                              össer                                                         House Part                 1902        1906; 1914           1938             1976       Coral spread                --             --             edjeman             Lã        Cross battens for           --                         djädädje, ruwe         --        loft                                                                                    Cross beam                 tur             jä             durr djä            tur       Cross beams for loft                   lon in iem            ra               --        Diagonale Latten         aunwölle       aninwólä             --            añinwolã     Floor of Loft              bwo            böo                bo               po        Frame                       --             --                --            kãdikdik     Gable                     djabbo     kijmen, räman           --               --        Gable batten              tortor         dordor           tortar,           tortot                                                              darag,tarak                     Kingpost                  ruling         rullin           druleng             --        Loft                        --             --           im kidjerik           --        Longitudinal battens        --                            wádädje             --        for loft                                                                                Longitudinal beams          --                          moe & moerik          --        for loft                                                                                Open part below roof        --             --               lóau              --        Post                       jur            jor               djur             joor       Rafter                      --             --          katal (keerer        jekpad,                                                                [Rtk])           kattal      Ridge batten                --         lajogemen         lädjókemen           --        Ridgepole                 burwoj         borwaj           borowadj          bõrwaj      Roof                        --             --                --          maltu, tõrak   Roof batten              djekeber       jekeber           djokeber         jinniboor                                                          (kädilmak [Rtk])                  Roof plate                  --             --              kaelep             --        Saddle roof             älik in im    älig in iem            --               --        Stick in thatch unit        --             --             keinadj             --        Thatch unit                 --             --           adj, adj, ad          aj        Tile slat                   --          kärikrik         kerikerik            --                                                             (kedillemak [Rtk])                 Wall                       jojö           jojö             djädji             --        Window flaps                          rönel bällok                            --        

Table 2 Glossary of terms for different types of buildings

Type of building               Steinbach-Grö    Erdland       Krämer &      Abo et al.                                      sser                       Nevermann                  Boat house                          --             --            --             --       Chief's house                       --             --         imalablab         --       Church                              --             --            --          mõn jar     Cook house                        bellak         belak      bellak, iman     mõn kuk                                                                     kemát                    Hospital                            --             --            --         mõn taktõ    House for the demented              --             --            --         mõn bwebwe   Hut used to dry Pandanus            bui           bui            bui            --       preserves                                                                                Latrine/Toilet building             --             --            --        mõn bwidej,                                                                              mõn kõppojak  Light house                         --             --            --           miade      Living house                        iem            im            iem            em       Meeting house                       --             --         imalablab    mõn kweilok   Men's House                         --             --            --             ja       Menstruation house                 juken         jugin         djuken           --       Pandanus scraping house             --             --       iem an kilok        --       Restaurant building                 --             --            --          mõn mõnã    School house                        --             --            --         mõn jikuul   Store House/Ware house              --             --            --            joko      Tattooing house                     --             --         imalablab         --       Temporary House                     --             --            --        imon kõppãd   Town Council House                  --             --            --        mõn kweilok   Trade store                         --             --            --          mõn wia     Two Storey House                    --             --            --           nikai      Whore House                         --             --            --         mõn utlam    

Krämer classifies the Marshallese house as a transitory form between a house built on level ground and a house built on stilts. The traditional house consists only of a roof structure which is set on four main posts, which protrude between 1.3 and 2m above the level ground.. Walls are commonly absent and the area underneath the roof is utilized as the living area. It is covered with a coral gravel spread in a manner like the area outside the house. To make the area more comfortable, it is then covered with mats.

The roof has a loft, which can be accessed through a hatch from below. The loftwas used as a sleeping area and to store the precious things such a good mats and the like. Since it was sat on posts with a very smooth surface it was almost rat proof (hence the name: im kidjerik--rat house). As mentioned, traditionally there were no walls. With the arrival of the Europeans and the advent of Christianity, the houses were increasingly furnished with walls, either only on the wind (weather)-side or all around

The roof rested on a frame made from two roof plates and two cross-beams which have been tied to the plates. This frame forms both the base of the roof and the frame with the floor of the loft. In the middle of each of the crossbeams aking post is placed, which supports the ridge pole.

The rafters are set out in pairs, and meet at the ridge in such a fashion that they pass underneath ridge pole. the thus partially supporting the weight of the ridge pole. Across the rafters tile battens are tied horizontally onto which slats are attached to which the individual Pandanus thatch unitsare fastened. The a central ridge batten was attached on top of the pole. The ends of the roof at the gables were competed with gable battens. In some cases the roof structure could be reinforced by diagonal battens. This type of roof structure is identical to that in some parts of Western Polynesia, particularly Tonga.

The gable itself was also constructed by rafters running from the ridge pole to the cross beam, by batten attached perpendicular to them and by tile slats tied on top. The roof thatch units consisted of brown, fallen Pandanus leaves which had been wrapped around a "backbone" of a wooden stick and tied with the central ribs of coconut leaves.. The thatching begins at the bottom and by tying successive layers of thatch units to the tile slats the ridge is reached. The ridge itself is covered with old mats or woven coconut fronts. Such a roof is said to be watertight for 1-2 years, but longer on the northern islands which have less rainfall. The attic or loft rests on the roof frame. Several cross-beams have been tied onto the longitudinal beams (roof plate)

In the 1910s some older, but European-contact period houses (on Lae) had their posts not buried in the ground but sat on large stones which had been partially buried in the ground. The living area underneath the roof was also bordered by stones. At the turn of the century more recent houses had bottom plates with a floor made of thin planks and round wood.


There were a number of different house types at the turn of the century. The main houses were the living houses (iem). Apart from these we know of cooking houses, commonly covered with coconut thatch and without a loft, houses to scrape Pandanus and houses for menstruating women. Tattooing houses, meeting houses and boats houses were rare. To dry Pandanus preserves andprotect them against rats small houses were built on for high posts which consisted of two open platforms and a saddle roof. Small houses were built fo rthe spirits of the deceased, in which magicians and sorcerers kept some belongings. Table 2 provides a glossary of the terms for the types of buildings. It should be noted that several terms given by Abo et al. (1976) are of Japanese origin, such as "nikãi"--two-storey house, a termderived from the Japanese word nikai "second floor".


The four main posts supporting the roof are made from ironwood (Pisoniagrandis; Kanal). Alternatively, wood from a number of trees, among them the mangrove Bruguiera gymnorrhiza served as timber. The wall covering for the houses was either made from split aerial roots of the Pandanus tree (P.tectorius), or covered with roof thatch. Table 3 sets out what could be compiled on wood and material use. The list is by no means complete.

Figure6. Traditional Marshallese house seen on Maloelap in 1910 (after Krämer& Nevermann 1938:Plate 14b).

Figure7. Traditional Marshallese house seen on Majuro in 1910 (after Krämer& Nevermann 1938:Plate 14a).

Figure8. Traditional Marshallese house of deviating construction seen on Majuro in1910 (after Krämer & Nevermann 1938:167).

Figure9. Traditional Marshallese house of deviating construction seen on Majuro in1910 (after Krämer & Nevermann 1938:168).

Figure10. Frame structure of previous house (after Krämer & Nevermann1938:168).

Table3. House parts and materials utilized

 House part                 King   Cross  Roof                        Wall   Roof   Lash-  Unspe                       Posts  posts  beams  plate  Rafte  Batte  Slats  cover  cover  ing    cifie                                                    rs    ns                                 d      Bruguiera, wood        n                         n                                         n      Callophyllum, wood                               n      n      n                                  Coconut fronds                                                        n      n                    -- husk (sennit)                                                                    n             --trunks               n    n      n                                                              Hernandia, wood        n    n      n             n                                                Pandanus, aerial                                                      n                           roots                                                                                             -- leaves                                                             n      n                    Pisonia, wood          n    n      n      n                                                       

Table4 Timeline of the development of Marshallese houses

Date    Foundatio     Floor     Walls        Loft     Posts         Roof            Wall                   n                                                                   construction   <1800     None        Coral     none          yes     Set on       Thatch           none                            gravel                           stone?                                   1817      None        Coral     none          yes     Set on       Thatch           none                            gravel                           stone?                                   1850s     None                  none,                              Thatch         poles w.                                     Thatch                                            thatch?      1870s     None        Coral     Thatch       yes?     Set on       Thatch        poles with                         gravel                           stone?                       thatch      1890s     None        Coral     Thatch       yes?     Sat on       Thatch        poles with                         gravel                           stones                       thatch      1910s     None        Coral       Thatch,    None      None        Thatch        poles with                         gravel,       wood                                           thatch,                           pounded                                                    Wooden frame                         sand                                                                     1930s     None,       None,        Wood      None      None     Thatch, Wood    Wooden frame           Concrete    Concrete                                                                   1950s     None,       None,        Wood      None      None        Thatch,      Wooden frame           Concrete    Concrete                                        Wood,                                                                                       Corrugated                                                                                        iron                       1970s   Concrete    Concrete      Plywood    None      None      Corrugated     Wooden frame                                                                       iron                       1990s   Concrete    Concrete       Wood,     None      None      Corrugated    Wooden frame,                                    Concrete                           iron          concrete                                                                                        bricks      

5.The burnt houses on Mile Atoll

The descriptions as well as the pictures available for the traditional Marshallese houses clearly indicate that the entire area underneath the roof had been covered with the coral gravel spread. Thus, in the archaeological record we would have to expect a roundish or oval-shaped coral gravel spread without any gravel-free areas in the middle, the small gravel areas of the posts having become obliterated over time.

In May 1991 three guest houses burnt on Mile Island, Mile Atoll, probably as a result of arson. The charred remains of these guest houses, however, offered the opportunity to study the "archaeological" imprint of modern coral gravel spreads and houses. In the following we will set out the available evidence.


The guest house complex was located on Mile island, Mile Atoll, some 100m from the lagoonal shore. The complex consisted of three guest cottages set in a row running east-west, which were serviced by a store and administration building erected on the remains of the World War II power station.

The three guest houses were connected by a walkway made of coral gravel (figure11).. The walkway was lined with vertically set coral slabs of varying size. For each house a small connecting track was laid out, which ended in a stepping stone of coral. The area to the north of the walkway was covered by grass, while the area to the south, where the houses stood, was covered by coral gravel, gradually giving way to sandy surface. Some five metres behind the houses a row of Pandanus trees and a patch of banana trees were planted.


The three buildings were built to the same specifications. Figure 12 provides a schematic lay-out of the building. Plate 15 shows such a building when the guest house was still operational. The rectangular building was erected on ten major posts (plate 15, full circles), which had further floor supports set in regular intervals (shaded circles). The northwestern corer of the building had a small verandah taking up about 1 sixth of the building space. A glass-fiberor plastic portable toilet-cum-shower unit was set into the southeastern cornerof the building and a mattress was--usually-- located at the northwestern corner.

The building was made of locally available materials and imported wood. The wall supports, as well as the floor supports were made of local timber of unknown type, while the floor consisted of plywood. The walls, with the exception of the verandah walls, were made of Pandanus thatch, reinforced by a few 2x4 inch planks. The roof also consisted of Pandanus thatch throughout, reinforced by battens and rafters made of locally available stickwood. The building had three windows hinged at the top, consisting ofPandanus thatch reinforced by 2x4's.


An inspection of the buildings was conducted in June 1991, approximately 6 weeks after the fire. This assessment resulted in the following observations (figure 13 and plates 16-20): The former locations of the three buildings were well visible due to the remaining walkway connection, the stepping stones (see plate 20 left foreground) and three well defined piles of debris (plates 17; 18& 19). Of the walls and roof only has and few charred pieces of wood survived. The toilet unit had burned into a chunky ash which crumbled upon touch. The metal springs supporting the mattress structure had survived, and so had several nails and the hinges of the doors and windows. Several were seen, but are not plotted in figure 13. The wall posts had burned to the ground but charred stumps were protruding from the ground.

The most striking observation was that the central part of the former house area was free of coral gravel (plates 17-20). The boundaries of the gravel tono-gravel areas were irregular, an indication of how the gravel had been applied after the building had been erected.

The effects of the fire on the vegetation can be seen in plates 16 to 20. It consisted mainly of scorched leaves and branches. None of the plants seemed to be damaged beyond recovery.

In February 1992 an opportunity existed to repeat the assessment (plate 21-26). Almost all of the smaller wood particles and ash had blown away, leaving behind only a few larger pieces of charred timbers. Only a small amount of charred wood had become embedded in the surface. The tops of the charred wall posts were well visible (plate 23, 24) and so was the boundary between the coral gravel spread and the surrounding sand (plate 25), although increasing growth of vegetation over the site (plate 21) began to obscured some of the information. Macro remains, such as the coral lining of the walkway and the stepping stones were well visible. The scorched vegetation had recovered and showed little evidence of the fire.


The site on Mile offers the opportunity to document, over time, the development of the site, unless it is being redeveloped. We can expect that the boundaries between the gravel and no-gravel areas will continue to become blurred and that at one point in time the tops of the charred wall posts no longer will protrude making the recognition of the exact site as a house site more complicated.

The macro evidence of coral lining of the walkway and the stepping stones are likely to remain for some time until they are quarried and re-used at another site.

Figure11.Schematic lay-out of the three guest houses on Mile before (left) and after(right) the fire.

Figure12.Schematic detailed lay-out of a guest house on Mile before the fire.

Figure13.Schematic detailed lay-out of a guest house on Mile after the fire.


The descriptions as well as the pictures available for the traditional Marshallese houses clearly indicate that the entire area underneath the roof had been covered with the coral gravel spread. Thus, in the archaeological record we would have to expect a roundish or oval-shaped coral gravel spread without any gravel-free areas in the middle, the small gravel areas of the posts having become obliterated over time.

The observation on Mile have shown that the modern houses may not be set on the ground and thus may not have the characteristic coral gravel spread throughout the area. Observation on Tokowa, Mile, however have shown that there are a few modern building, shed, which do have a coral gravel floor on the inside. Thus we can state that a coral gravel spread with a coral free area(s) in its centre is likely to be of historic or modern date. Coral gravel spreads without such gravel-free areas, however, still remain undatable as far as morphological charateristics are concerend.

It is interest to note that only a small part of initially available ash and charred wood remains had become embedded in the surface of the site. Most of the ash and charcoal had been blown off-site by the tradewinds. This has implications on the recognizability of burnt houses from the archaeological record. The amount of ash embedded may be substantially smaller then anticipated.


I am indebted to the following, whi assisted, one way or the other, with this study: Hemley Benjamin (Historic Preservation Office, 1992); Carmen M. Bigler(Secretary for Internal Affairs and R.M.I. Historic Preservation Officer, Majuro); Jane Downing (Majuro); Newton Lajuan (Historic Preservation Office, 1989/90); David W. Look, (Chief, Preservation Assistance Branch, WesternRegional Office, U.S. National Parks Service, San Francisco); and Steve Simon(Nationwide Independent Radiological Survey, Majuro).



1976 Marshallese-English Dictionary. PALI language Texts: Micronesia.Honolulu: University of Hawaii Press.


1910 Reise um die Welt mit der Romanzoffschen Endeckungs-expedition in den Jahren 1815-1818 auf der Brig Rurik, Capitaen Otto v. Kotzebue Zweiter Theil:Bemerkungen und Ansichten. Chamisso's Werke Vierter Theil. Berlin: G.Hempel.[For the benefit of the anglophone readers I have quoted as much as posible the English translation rather than the German original (1st edition) of 1836. However, the translation of Notes and Opinions is not complete].

1986 A voyage around the world with the Romanzov exloring expedition in the years 1815-1818 in the Brig Rurick, Captain Otto von Kotzebue. (translated by H.Kratz). Honolulu: University of Hawaii Press.


1906 Wörterbuch und Grammatik der Marshallsprache nebst ethnographischen Erläuterungen und kurzen Sprachübungen. Archiv Deutscher Kolonialsprachen. Vol.IV. Berlin: Georg Reimer.

1914 Die Marshall Insulaner. Leben und Sitte, Sinn und Religion eines Südsee-volkes. Anthropos Bibliothek. Internationale Sammlung Ethnologischer Monographien, Vol.2(1). Münster: Aschendorffsche Verlagsbuchhandlung.


1893 Ethnographische Erfahrungen und Belegstücke aus der Südsee. Annalen des K. & K. Naturhistorischen Hofmuseums 7. Vienna: K. &K. Naturhistorisches Hofmuseum.


1953 The Land Vegetation of Arno Atoll, Marshall Islands. Atoll Research Bulletin 16. Washington: Pacific Science Board, National Research Council.

1957 Agricultural notes on the Southern Marshall Islands. Atoll Research Bulletin 55. Washington: Pacific Science Board, National Research Council.


1821 A voyage of discovery into the South Sea and Beering's Straits. for the purpose of exploring a north-east passage undertaken in the years 1815-1818, at the expense of His Highness the Chancellor of the Empire, Count Romanzoff in the ship Rurick, under the command of the Lieutenant in the Russian Imperial Navy, Otto von Kotzebue. 3 vols. London: Longman, Hurst, Rees, Orme and Brown.

1830 A new voyage around the world in the years 1823-1826. 2.vols. London: H.Colbourn & R.Bentley.


1905 Der Haus- und Bootsbau auf den Marschallinseln(Ralik-Ratak-Inseln). Archiv für Anthropologie N.F. III Braunschweig: Friedrich Vieweg & Sohn.


1938 Ralik-Ratak (Marschall Inseln). In: G.Thilenius (ed.),Ergebnisse der Südsee-Expedition 1908-1910. II. Ethnographie, B:Mikronesien. Vol. 11: Hamburg: Friedrichsen & de Gruyter.


1987 Archaeology in Eastern Micronesia: reconnaissance survey in the Marshall Islands. In: T.Dye (ed.), Marshall Islands Archaeology. Pacific Anthropological Records 38. Honolulu: Department of Anthropology, Bernice P. Bishop Museum. Pp. 17-168.


1990 Cultural Resource Managment Plan for Majuro Atoll, Republic of the Marshall Islands. 2 Vols. Report prepared in fulfillment of U.S. Department of Interior, Office of Territorial and International Affairs TechnicalAssistance Grant MAR-42. Report submitted to the Historic Preservation Office, Majuro, Republic of the Marshall Islands. August 1990.


Bibliographic citation for this document

Spennemann, Dirk H.R., 1992, Coral Gravel Spreads. Observations On TheArchaeological Differentiation Between Prehistoric And Historic House Sites InThe Marshall Islands . HPO-Report 1992/7. Majuro Atoll, Republic of theMarshall Islands: Historic Preservation Office.
URL: http://marshall.csu.edu.au/Marshalls/html/archaeology/arc.html

Dirk H.R. Spennemann, Institute of Land, Water and Society, Charles Sturt University, P.O.Box 789, Albury NSW 2640, Australia.
e-mail: dspennemann@csu.edu.au

select from the following...
World War II

Digital Micronesia-An Electronic Library & Archive is provided free of charge as an advertising-free information service for the world community. It is being maintained by Dirk HR Spennemann, Associate Professor in Cultural Heritage Management, Institute of Land, Water and Society and School of Environmental & Information Sciences, Charles Sturt University, Albury, Australia. The server space and technical support are provided by Charles Sturt University as part of its commitment to regional engagement. Environmental SciencesInformation Sciences

© Dirk Spennemann 1999– 2005
Marshall Islands Kosrae CNMI Commonwealth of the Northern Mariana Islands Guam Wake Pohnpei FSM Federated States of Micronesia Yap Chuuk Marshall Islands politics public health environment culture WWII history literature XXX Cultural Heritage Management Micronesian Journal of the Humanities and Social Sciences CNMI German Colonial Sources Mariana Islands Historic Preservation Spennemann Dirk Spennemann Dirk HR Spennemann Murray Time Louis Becke Jane Downing Downing