Niku III (1997)
"Once and For All."
- Preliminary planning.
- Summary report.
- "Hell and High Water."
- "Completing the Puzzle."
- "I Saw Pieces of an Airplane."
With the water catcher on the windward side seemingly eliminated as a likely Earhart associated site, attention focused during the 1997 expedition on Aukaraime South, the lagoon, and specific sites in the village.
The Aukaraime South Site obviously merited further investigation as the site of the 1991 discoveries of the shoe parts and other possible Earhart-related artifacts, and because Bevington had identified it as the site where he and Maude had seen signs of some sort of occupation. During preparation for the 1997 fieldwork we inquired of Harry Maude about his own recollections. Though he did not identify a specific site, he confirmed that Bevington had shown him a site where he recalled seeing piles of debris that he associated with Arundel's workers.
There is nothing very obvious to recommend Aukaraime South as a camping place. Like most of Nikumaroro, it is flat, heavily wooded, with no distinguishing geographic features, lying about two meters above the level of the lagoon. In is not far from Baureke Passage, however, and in reviewing airphotos of the area we noted that between the site and the passage, there is a linear area that is relatively clear of vegetation. Historical photos indicated that this area has been fairly clear since at least the late 1930s, apparently as a result of frequent salt-water overwash during storm events (Fig. N-30). We speculated that it might have been an attractive landing site for Earhart and Noonan. The 1938 New Zealand aerodrome survey maps of 1938 (Fig. N-31) showed the clear area bordered by Buka trees; these had been cleared for coconut planting by the time of the 19_DATE_ airphotos. A typical Buka tree, as we measured in the field, has a trunk-to-limbtip radius of six to eight meters. A Lockheed Electra, 11.7 meters long and 16.6 meters across the wings, would not be very visible to the Colorado pilots if landed on the cleared area and parked under a tree to escape the fierce tropical sun.
But if Earhart and Noonan had landed on the cleared area and camped at Aukaraime south, why did Bevington, Maude, and their colleagues not see the airplane? It seemed plausible that the same forces that kept the clear area clear had cleared it of the airplane -- that at some point before the Maude-Bevington visit, storm waves had swept the cleared area and carried the airplane into the lagoon.
Once in the lagoon, assuming it was afloat, the airplane could have gone almost anyplace, but it seemed most likely that it would have sunk somewhere not too far from the northeastern end of the cleared area and the inner mouth of Baureke Passage. To check this possibility, the 1997 expedition was equipped to conduct a detailed underwater search in the lagoon, using both divers and electromagnetic sensors.
Finally, the village was chosen for further investigation simply because it was, after all, where we had found all the aircraft fragments during the previous expeditions. Wherever the airplane was, it appeared likely that the colonists had been salvaging pieces from it and taking them to the village. It was possible, then, that we might find the "smoking gun" artifact in the village -- the fragment with a definitive serial number or other identifier linking it unquestionably to the Earhart airplane. More realistically, a larger sample of airplane debris from the village might help us understand what airplanes were producing the pieces the colonists used, and the transformation processes that led such pieces to be part of the village's archeological record. Understanding these processes, we hoped, might give us clues to the original location of the wreckage.
The expedition team of ten, under the direction of Gillespie, departed Suva, Fiji on February 22, 1997, aboard the Nai'a, a 110' motorsailer owned and operated by Nai'a Cruises, Inc. We were accompanied by a three-man documentary crew from ABC Television, under the direction of Producer Howie Masters. We arrived off Nikumaroro on February 27, and after the usual preliminaries began work.
This expedition was equipped with Trimble GPS units, and a (DON ELABORATE??) base station that was established near the landing site. With this equipment we hoped both to record the locations of specific sites and features accurately, and to locate Nikumaroro itself more precisely than it had been in the past. Unfortunately, the base station required at least xxx satellite readings to produce an entirely accurate location. After only xxx readings, the onset of Cyclone Hina began to flood the base station and it had to be quickly relocated. Nevertheless, the base station (DON???????)xxxx
Relative locations of artifacts and features on sites were plotted using a Canon(??RIC??) "Total Station" pulse laser, mounted on a tripod over established datum points at sites where intensive work was done, and in hand-held mode when mapping long transects.
Datum points used in the 1991 grave excavation and shoe search were relocated, and a permanent datum point established, marked after excavation by a subsurface circle of bottles around an easy-to-find metalic core. From this point two loci were laid out for intensive surface inspection. The "Shoe Locus" included but went well beyond the original shoe discovery site, while the "Psychrometer Locus" encompassed the area where the psychrometer and medicine bottle lid had been found. Both areas were then cleared of coarse surface litter (a considerable undertaking), and blocked off in four-meter squares. Each of these was then carefully inspected on hands and knees, sorting through the fine surface debris with trowels and fingers, and was swept with metal detectors. While this work was underway, both areas were also probed with an electromagnetic sensor, revealing a single apparent anomaly. A 1x2 meter test pit was excavated on this anomaly, in 10 cm. levels, passing the soil through 1/8 inch screen and washing a sample. A second identical unit was then excavated adjacent to the first, with screening reduced to a sample. A series of shovel test pits were then excavated in each of several grids distributed across the area, one of which was expanded to a 1x2 meter test unit when it revealed a concentration of wood ash and charcoal (Fig. N-32). All excavations were backfilled at the close of the project, after being marked with cans and bottles to facilitate their relocation.
Figure N-32 -- Aukaraime South: Areas Intensively Investigated 1997
The area between these two loci and the clear area along the shore of Baureke Passage was also inspected, both along the lagoon shore and for about one hundred meters toward the ocean. Aside from scattered bottles and boards, nothing was found in the interior. Along the shore a series of five short coral "piers" were noted. The first was about forty meters east of our landing place at the lagoon shore of the "shoe site." The next was about forth meters west of the landing place, and the next about forty meters farther west. The fourth was about thirty meters from the third, and the fifth and last was roughly 100 meters from the fourth. Each was about six meters long and a meter wide, made up of coral chunks. Among other possibilities, these may represent fish traps, sand traps to build and protect the shoreline, or walkways to overwater latrines.
An initial search box, approximately 115 meters on a side, was corner-marked with weighted buoys located using the clear area along the east side of Baureke Passage as a visual reference. Additional boxes of various sizes were laid out from the first, eventually forming a gridwork of twelve boxes (see Figure N-34). After the boxes were laid out, the southeast corner of the southeast box was tied into two benchmarks on the lagoon shore using the total station. The total station was then used to relate all the adjacent boxes to one another.
All twelve of the boxes shown on Figure N-34 were surveyed in their entirety using the electromagnetic sensor (EM-31) and a submersible magnetometer. GPS was used track the movements of the boat containing the sensors and to locate some of the box corners and calculate box areas. The total area inspected amounts to about 4 percent of the lagoon area. In addition to the electronic sensing, divers were towed on "manta boards" behind the lagoon boat and inspected the area visually. The few "hits" with the sensing devices were subjected to detailed inspection by divers, following circular search patterns centered on each "hit."
Manybarrels Site Because of the plexiglas and other aircraft-related debris found at Manybarrels' in 1996, this site was a major focus of attention. Located in fairly dense forest southeast of the Government Station, it was hard to locate precisely, but a long point-to-point transect with the Total Station enabled us to plot its location with fair accuracy, as shown in Figure N-33. The site itself was cleared of coarse surface debris, visually scanned on hands and knees, and swept with metal detectors. Metal detector hits were marked with painted tongue depressors and then trowel-excavated where the artifacts responsible were not visible on the surface. Artifacts and features were described and photographed in place, and collected where they appeared to be of possible interest -- either as aircraft associations or in order to understand the site as a whole.
Sam's Site, Kent's Site, Gallagher Highway On one of the first days ashore, while filming the team at work, ABC videographer Sam Painter discovered several pieces of aircraft aluminum in a complicated residential site slightly north of the trail from the landing to the lagoon. Promptly designated "Sam's Site," this site was not thoroughly cleared, but was mapped and inspected as closely as time allowed.
Early in the work, team member Kenton Spading located several pieces of aircraft aluminum not far from the Cooperative Store (where Artifact 2-1, the Navigator's Bookcase, had been found in 1989). "Kent's Site" was also mapped and inspected, though not intensively.
Because we continued to find aluminum and other interesting objects each time we traversed what we had come to call the "Gallagher Highway" -- the trail from the landing site to the lagoon -- we mapped the "highway," describing its cultural features and collecting artifacts that appeared to be possible Earhart associations.
We had hoped to undertake detailed surface inspection and excavations at Site 17 in the Government Station, the "Carpenter's Shop," but the approach of Cyclone Hina forced us to cut the fieldwork short and flee, eventually landing in Funafuti, Tuvalu. The last several days of work in the village were conducted under conditions of heavy rain and dangerous surf at the landing site.
Aukaraime South Surface inspection of the vicinity of the "shoe site" on Aukaraime South was remarkably unproductive. No more shoe parts were found, with the possible exception of artifact 2-4-G-xx, a small washer described in Section xx. Fragments of rusted ferrous metal were noted here and there, almost certainly the remains of fuel tanks from the colonial period. A concentration of roofing nails and a pair of gloves were found, the residue of TIGHAR's 1991 work. Scattered flecks of charcoal were noted.
Initially, excavations were similarly unproductive. Whatever the anomaly was that was detected by the EM-31, it was not visible in the ground. A shovel test placed at the exact site where the shoe was discovered in 1991, however, revealed a concentration of wood ash and charcoal, in an irregularly circular area about fifty centimeters across and five to ten centimeters below the surface. The surroundings of this feature were excavated and screened, revealing a scrap of paper can label given artifact number 2-4-G-xx (RIC???) and described at xxx.
The feature itself was removed in its entirety and rturned to the U.S. for analysis. A small sample from each quadrant of the feature was first scanned with a scanning microscope. A one-half liter sample from each quadrant was retained for possible future analysis. The remainder of each quadrant, about three liters of soil, was divided into fine, medium, and coarse fractions through water flotation separation by Cultural Resource Analysts of Lexington, Kentucky. All fractions were retained, and inspected under low-power magnification. No evident cultural material was found. The microscopic scan indicated the presence of a few nodules of a material that might have been melted plastic, and the flotation recovered about 25 milliliters of wood charcoal. Examination of this material by tropical botanist Rachel King of the University of Miami indicated that it was most likely from a monocot such as coconut palm.
The can label initially appeared to be of considerable interest, but then was found to contain a fragment of a grocery bar code. We concluded that the label, and hence also probably the fire that produced the feature, represent the leavings of the 1978 Republic of Kiribati survey of the island, or that of some other relatively recent visitor.
Lagoon The lagoon area shown in Figure N-34 was inspected as described above, with entirely negative results. The only cultural object noted was a length of anchor chain.
Manybarrels' Site Laxton describes a typical housesite on Nikumaroro, and elsewhere in Kiribati, as follows:
"A Gilbertese village has three buildings to each bata or household. The sleeping and living quarter fronts the village street; behind it is the eating room, about twelve feet away, and behind again the cookhouse. It would be a poor village indeed which was not scrupulously clean, and Nikumaroro prides itself, and is as good as the best. Forty yards away are the village cone sheds, each household owning at least one of the beautifully made canoes"
Similarly, Knudson reports that:
"The house site comprises a minimum of three buildings: a sleeping house about 15 feet by 18 feet with a floor raised about three or four feet from the ground, a small cookhouse behind the sleeping house and on ground level, and a canoe shed"
Figure N-35 illustrates the spatial organization of the bata represented by the Manybarrels Site. An "L" shaped alignment of coral slabs apparently represents the base of a house that either never had, or has lost, the west and south sides of its platform walls. The house would have been somewhat under four by six meters in size in order to fit within the platform walls. Doubtless, like other Nikumaroro houses, it consisted of four or more upright poles supporting a pitched thatched roof, with woven pandanus frond walls under a meter high. The house fronted on Sir Harry Luke Avenue, some eight meters to the southwest. About the same distance to the southeast, near a large coral head outcrop, the cookhouse was represented by a dense concentration of charcoal and wood ash, with a number of calcined large animal bones, apparently representing pig and turtle.
Figure N-35 -- Manybarrels Site
In this case, then, in contrast with Laxton's and Knudson's perhaps somewhat idealized description, the cookhouse was not behind the house but to the right of it as one faced the house from the road. This placement may be the result of the prevailing wind, which would tend to blow smoke into the sleeping house from a cookhouse placed to the northeast. The eating area probably was behind the house, however, represented by the substantial scatter of artifacts that we recorded there (Fig. N-36). The placement of the two 55-gallon drums included in this cluster, four to five meters apart and aligned with the house platform, suggest that a roofed structure stood here with rain barrels at two of its corners -- probably an open-sided shed where household work could be done protected from rain and sun.
Artifacts in the cluster shown in Figure N-36 included a wide range of household items -- a plate, a bucket handle, a tablespoon -- as well as brass and ferrous pipes, flashlight reflectors, an eyeglass frame, and lead weights, probably from fishnets. Most interesting to us was a tangle of cable identical with that found in 1996 -- apparently aircraft control cable -- and two clusters of artifacts near the small rock outcrop. One cluster included a large piece of stainless steel, a flashlight reflector, a copper tube, a battery cable, and a dense rectilinear mass of copper wire identified as the winding off a transformer or electric motor. The second included two large slabs of pearl shell, a red glass bead, and a small rectangular piece of aluminum, apparently Alclad. Nine additional pieces of aluminum were found, most clustered toward the edges of the site. All the aluminum pieces were small and obviously deliberately cut; in essence they appear to be "blanks" cut from larger pieces into convenient sizes for transport and storage until needed in some craft application. It appears that some kind of handicraft production was among the activities carried out in the eating area of the Manybarrels Site.
At the edge of Sir Harry Luke Avenue, eighty meters "down the road" to the southeast of the Manybarrels house site is a steel pipe driven into the ground and set in concrete. A standing coral slab adjoins the pipe perpendicular to the road alignment, with patches of concrete on either side and a loose piece of concrete that has fallen into the road. The numeral "16" is on the northwestern patch and on the loose piece, while the number "17" is inscribed in the southeast patch (Figure N-37).
Figure N-37 -- Boundary Marker
"Next day commenced the erection of the boundary marks. We alloted some spoilt cement and damaged piping and old paint from the U.S. radio site stores, title in which had passed to the British government. Old Kirata and assistants cut the pipe into four-foot lengths; the cement was mixed, pits dug under each peg, part filled with clean rubble, the length of pipe driven in erect and its foot bound with cement. A number was given to each land and engraved in the wet cement. Later they returned and filled the engarved numbers with pitch, painted the projecting pipes, topping them with scarlet for gay effect. The completion of this merited another picnic, during which the lines of the plots were carried from lagoon to sea, marked with stones and small boulders."
Figure N-38 shows Laxton's sketch-map of land divisions on Ritiati, together with part of his list of landowners. If the Manybarrels' Site was the land parcel numbered sixteen, it would have been the bata of Teng Maraki and Nei Kantaraa. If -- as seems likely given the distance from the house to the marker -- it was parcel fifteen, it was assigned to Teng Banibai and Nei Tebea.
While we cannot be certain that the site was not occupied earlier, land parcels fifteen and sixteen were apparently parts of the "New" Ritiati Village created as part of Laxton's reorganization of the colony in 1949. They were apparently assigned to settlers already on the island, however, not set aside as leasehold land for the new settlers Laxton intended to bring in from Manra. The small pieces of aluminum were probably exchanged among families engaged in craft work, however, so there is no guarantee that only immigrants from Manra would have aluminum from the known wreck on that island, or that only "old" families would have aluminum from any older wreck that might have been found on Nikumaroro. In addition, of course, travel between Nikumaroro and Kanton Islands provides another source of aircraft aluminum. None of the aluminum pieces on the Manybarrels Site is distinctive enough to be assigned to any particular airplane.
In surveying a transect to tie the location of Manybarrels into known points along the Gallagher Highway, we recorded one other house site, a substantial stone structure resembling the "pigpens" located in the southern part of the New Village in 1989, and a well. This must represent either land parcel seventeen or parcel eighteen, the batas of either Teng Abara and Nei Marenga or Teng Teibi and Nei Taiana according to Laxton's map and table.
 Sam's Site/Gallagher Highway North Figure N-39 shows the spatial organization of "Sam's Site" and the adjacent northern Gallagher Highway. What we call the "Highway" is not a historical track, though it more or less parallels the road Laxton mentions between the landing and the lagoon. It is simply the way we found to cross the island from landing site to lagoon with the least inconvenience and environmental impact, so it represents a more or less randomly selected wandering transect across Ritiati at this point.
The northeastern end of the "Highway" is a stone structure on the lagoon beach. The purpose of this structure is unknown. Immediately to the southwest, the land becomes quite swampy, and there are no structures. Then the path rises somewhat, and hence becomes more dry, as it passes to the southwest. It crosses the remains of at least four houses, three other structures, several long walls, a well (home of a coconut crab when we arrived), and another well or small babae pit. "Sam's Site," which extends off to the northwest with no real boundary from the "Highway," contains more linear walls and a wide range of artifacts -- a sewing machine, bicycle parts, the casing of a barometer or chronometer, large rivets, clamps, and a good deal of aluminum.
We know from both air photos and Laxton's account that there was a road from the landing to the lagoon somewhere in the vicinity of the Gallagher Highway. Some of the long walls shown on Figure N-39 -- each made up of aligned coral slabs -- may represent the edges of this road. Others may represent property boundary markers, or the perimeters of public facilities. Although Laxton's hand-numbered map is hard to read at this point, it appears that the land just northwest of the road to the lagoon was Ritiati Parcel 24, assigned to the London Missionary Society, while the parcel immediately southeast of the road was Noriti Parcel 1, assigned to Teng Banibai and Nei Tebea.
Gallagher Highway South/Kent's Site Figure N-40 shows features and artifacts along the southern part of the "Gallagher Highway," including "Kent's Site, and the adjacent Cooperative Store with the associated house sites mapped in 1989 (in one of which the Navigator's Bookcase, Artifact 2-1-V-1, was found). "Kent's Site," a poorly defined house site containing planks, a bed frame, bottles, and a number of aluminum pieces, lies immediately north of the 1989 house cluster. The Gallagher Highway ends at the base of the now-destroyed landing monument, and for purposes of the 1997 survey, at the nearby GPS base station.
- John Clauss, TIGHAR #0142CE
- Veryl Fenlason, TIGHAR #0053EC
- Richard E. Gillespie, Executive Director, TIGHAR
- Van Hunn, TIGHAR #1459EC
- Thomas F. King, Ph.D., TIGHAR #0391EC
- Tommy L. Love, D.O., Col. USAF, TIGHAR #0457EC
- Gary F. Quigg, TIGHAR #1025EC
- Carolyn J. Schorer, TIGHAR #1376EC
- Kenton Spading, TIGHAR #1382CE
- Kristin Tague, TIGHAR #0905CE
- Tonganibeia Tamoa
- Senior Examining Officer
- Customs Division, Republic of Kiribati
- Patricia R. Thrasher, President, TIGHAR
- Donald Widdoes, TIGHAR #1033ECB
- Howie Masters, ABC producer/director
- Sam Painter, ABC cameraman