Early on the morning of 2 February the island came up on the ship’s radar and by 0700 the first team members were ashore at the landing and clearing a trail across to the lagoon shore. By fortunate coincidence, the tide was high which permitted a launch to be walked through the main passage into the lagoon without delay. The lagoon boat was in place and the trail from the landing completed at 0835. The team then embarked for the trip down the lagoon, arriving at the search area at 0900.

Although appearing relatively open in early aerial photography, the area was found to be now solidly overgrown with the tangled underbrush known in Gilbertese as te mao (Scaevola frutescens). This made it particularly difficult to navigate to the precise location where debris appears in the photos. Without calibrated GPS (Global Positioning System) information, the only sure method was to cut and physically measure transects across the island from lagoon to ocean. Only by finding the area where the width of the land matched the scaled distance in the photo could we be certain that we were in the right place. This was a frustrating and labor-intensive procedure which involved many hours of machete work in temperatures averaging 106° F. By the end of the day two transects had been cut and measured thus making it possible to define the areas to be searched. The team departed the site at 1635 and was back aboard ship by 1800.

The team was on site at 0755 and began building and searching boxes of terrain. From a known point on a transect a line was cut 90° into the bush 10 meters in length. From this point, another 10 meter line was cut paralleling the original transect, then back again to form a box. Surveyor’s flagging was used to create a physical boundary, thus permitting a thorough visual inspection of the boxed area despite the nearly impenetrable vegetation.

At 1012 cultural (man-made) debris was encountered in an area 38.2 meters northeast of the northern transect. Detailed examination of the site disclosed the presence of a variety of objects comprising a small shelter or campsite constructed of materials which clearly originated in the Gilbertese settlement at the western end of the atoll. An unused roll of tar paper roofing material suggests the possibility that the structure was never actually completed. Of particular note was a relatively large (roughly 1 meter square) steel tank identical to others seen in the abandoned village and which appears to have served as a cistern. An M-1 carbine shell casing found nearby testifies to the site having been visited by U.S. Coastguardsmen in 1944 or ’45. It seems logical that this is the ”water collection device” reported and sketched by USCG veteran Richard Evans.

The remains of a steel barrel or drum was found in a location which matches reflections from a large metallic object, designated “Candidate #1” in the 1941 photo.

While part of the team examined the shelter site, others continued to progressively search designated sections of bush. By the end of the day no new cultural sites had been found but strips of vegetation-free coral had been encountered and mapped. These appeared to match in location and orientation, although not in overall dimensions, the ”cleared” strips visible in the early aerial photography. All team members were back aboard ship by 1750.

Once more on-site by 0755, part of the team began the process of photographing, documenting and mapping the shelter site while the rest of the team took up the search for “Candidate #2.” Having resolved our on-the-ground location with relation to the early aerial photographs, it was a relatively simple matter to navigate to, box off, and examine the suspect area. An exhaustive search turned up no cultural debris. In the spot most closely matching that of Candidate #2, the team encountered a very old buka tree (Pisonia grandis) – the only one in that particular area and, in all probability, the anomaly seen in the photo.

At 1530 the work on this part of the island was judged to be completed and a decision was made to use the remainder of the day to correct an oversight from the Niku II expedition. The map location of the gravesite excavated in 1991, near which the remains of shoes believed to be those belonging to Earhart and Noonan were found, had never been accurately established. The team therefore relocated that site and measured its azimuth and distance from landmarks identifiable on the map. The site exhibited considerably more ground vegetation (specifically, networks of light vines) than had been present in 1991, and looked very much as it had when first noticed in 1989 during Niku I. A severely oxidized ferrous fitting with what appears to be a brass cap was collected near the site in the hope that it will provide some clue about the various types of activity the area has seen over the years. The team departed the area at 1635 and everyone was back aboard the ship an hour later.

With the need to begin the return voyage to Fiji that evening, this was to be the last day of work on the island. The team was ashore by 0700 and the decision was made to spend the available time in re-examining locations and features in the village where airplane debris had been found in the past. By 0807 the “carpenter’s shop” had been re-located along the shore of Taziman Passage. It was near this spot that Artifact 2-18, the ”dado,” had been found in 1989. Although one wall and some shelving had been standing then, the site was now leveled by subsequent storm activity and identifiable only by the presence of massive objects (the iron wheel and frame of a cart, a coil of heavy cable). Smaller artifacts were either swept inland or buried under up to 5cm of sand. While some of the team began a partial excavation of the carpenter’s shop, others attempted to re-locate another site along the shoreline where a sheet of aircraft-grade aluminum, cataloged as Artifact 2-2, had been found in 1989. Bordered by poles set in the ground which were notched at the top to support cross beams, the site had been littered with glass bottles and other debris prompting the Niku I team to dub this site “Noonan’s Tavern.” Efforts to re-locate this site were, however, unsuccessful and it is feared that it has been virtually obliterated by storms.

At approximately 1030 the excavation of the carpenter’s shop site produced two lengths of shielded electrical cable. On the end of each cable was a single-pin connector surrounded by a knurled tightening ring. The cables were very unlike the other objects which had been found during the excavation (mostly heavy ferrous tools and machine parts) and, while badly deteriorated, appeared to have most of their component parts intact. On-site evaluation was that these were consistent with cables and connectors for an American radio of less than 100 watts output. Because the island radio station had British equipment, and the U.S. Coast Guard station at the other end of the atoll would likely have had a communications radio of more than 100 watts, the cables were judged to be of sufficient interest to merit their collection for further analysis. They were recovered as Artifact 2-3-V-1 (TIGHAR project #2 , expedition #3, Village site, object #1). Nothing further was collected from this site.

In the afternoon an effort was made to re-locate a former dwelling site where several aircraft parts were recovered in 1991. By 1340 a spot had been located which was suspected of being that same location. (Later mapping and comparison to 1991 Field Notes, however, established that identification to be incorrect.) The site is sufficiently far inland from the shore to show little or no sign of storm damage. A close examination of the ground surface revealed the presence of several small objects and scraps of material apparently left over from projects of handiwork. Among these were items which appeared consistent with aircraft materials. These included:

• A roughly 6cm x 12cm (2.5 in. x 4.5 in.) sheet of uncolored transparent plastic 3mm (1/8 or .125 in.) in thickness from which rectangular pieces had been cut. A smaller shard of the same material found nearby fits a break in the bigger piece. Both pieces exhibit a slight but uniform arc over their surface and were apparently once part of a larger sheet. These were collected as Artifact 2-3-V-2. (See “Part #40552.”)
• A 15cm (6 in.) length of what appears to be thin-gauge high-grade stainless steel wire twisted together in a manner consistent with aircraft safety wire. This was collected as Artifact 2-3-V-3.
• A rectangular object 4.5cm x 4cm (1.75 in. x 1.5 in.) made of non-ferrous metal (lead?) and giving the appearance of being a cast cover plate with indentations in the back. The front features a circular logo with the word “STURDEE.” This was collected as Artifact 2-3-V-4.
• A roughly 50cm x 30cm (1.5 ft. x 1 ft.) sheet of apparent stainless steel estimated to be as much as .060 in. in thickness from which rectangular pieces had been cut. Designated Artifact 2-3-V-5, this object was not collected but was left in situ.
• An electrical “cannon plug.” This was collected as Artifact 2-3-V-6
• A very small ferrous object, possibly a fuse holder, collected as Artifact 2-3-V-7.

Work on the island was concluded at 1600 and all were back aboard ship by 1620 at which time Mr. Teuatabo approved the export of the artifacts for research purposes. The return voyage to Suva, Fiji was accomplished in five days and, on February 10, 1996 the TIGHAR team returned by air to Los Angeles. Mr. Teuatabo returned by air to Tarawa on February 11, 1996.