War in the Pacific Archelogy and History of Guam

Before the war, there were many latte groups and other archeoelogical sites on Guam — H. G. Hornbostel estimated "more than 27 monument sites on the island" (Thompson, 1932, p. 15). The first of the accompanying maps, with distribution of latte as of 25 years or so ago shown by nachure, is from among Hornbostel's notes at the Bishop Museum in Honolulu. The areas of "dense latte" at Mt. Santa Rosa and north of Yigo are questionable, and none is shown for the northeastern coast; but however undependable some of its details, the map supports the general effect of pre-war abundance of archeology on the island. Right after the war, Douglas Osborne located and investigated 26 sites, which are indicated on Map 2, taken from Osborne's unpublished report of 1947. Several of these have since disappeared, or deteriorated further to a marked extent. A few additional sites not recorded by Osborne were also visited in 1952; there undoubtedly are quite a number of surviving latte groups which nave not been surveyed by either Osborne or myself, and considerable further work is needed to complete the archeological survey of Guam.

1. Vicinity of Agana.

Most of the latte sites which formerly existed in the general vicinity of the capital were already gone at the time of Osborne's work six years ago. Since then, others of which at least remnants survived in 1946 nave disappeared — Maina site described by Osborne has given way before a small reservoir, and the new church which was to be built on the site at Toto is now there, with no traces of the latte remaining.

There is, however, one fairly good site, in the Agana Heights area — evidently not the "badly destroyed remnants of a rather large group" mentioned by Osborne, which seems to be completely gone — on private property across the road from Commissioner Anselmo Garrido's house, little over a mile south of the Tutujan road. On the Tenorio property there is an east-west group of 5 pairs of latte (9 of the 10 halege still-standing, 6 tazas present, fallen and broken), 10 feet apart; each slab about 30" wide, 15" thick; 10" to 12' by 8" at the top; 36"-40" high. There are a few surface sherds, and it looks as if there should be so midden deposit. On the Ullca property, just to the vest, there is a broken-up group with one halege (vertical slab) still standing in a distinct row of six; the opposite row, and three more houses in an east-west are indicated only by remnants. In addition, a few scattered latte stones and a bedrock mortar are found across the street to the south, in the back yard of the first house east of Commissioner Anselmo's. If at all possible, at least the first of these groups, located right in the outskirts of town, should be acquired and preserved for study and interpretive development.

2. Maimai

On the F. N. Santos property at this locality, three miles southeast of Agana, is a site probably typical of those which formerly existed, and of those which remain to be found (the Maimai site is listed by neither Thompson nor Osborne), in the central inland portion of the island. It is in tangantanga and other dense growth near very large bamboo thickets among rolling hills. There is an extensive occupation area of black soil and widely scattered potsherds and little shells, with no very deep midden indicated at any, one spot, and at least four sets of latte: (1) disturbed, small, four pairs, east-west, only standing about 24" high, 10 feet apart each way, all heads fallen, very small; (2) three very small pairs, northeast-southwest, standing 20" high, 24" wide at the ground, 8" thick, 10 feet apart, heads fallen, very small, decayed and broken, not measurable; (3) four pairs, larger, about 10 feet apart, in very bad shape; (4) five pairs, on up the hill in very dense growth, 12 to 15 feet apart, the stones larger than in the first two sets. Very probably this area would repay excavation, but no special effort for preservation or interpretive development is indicated. There probably are other similar sites still in existence but not yet recorded in this general area of the interior, in the central "waist" of the island.

3. Tumon Bay

Except for the important Gongna Cove site, discussed separately below, the formerly rich archeology of Tumon Bay had by 1945 been destroyed by road construction and recreational development along the beach, as reported by Osborne and by J. W. Brewer, Jr., of the National Park Service. About 20 years before, the Tumon-Epau zone (not Epau or Ypao Point so much as the central section of the bay) was one of the richest areas investigated by Hornbostel (Thompson, 1932, pp. 8-11).

4. Gongna

A large site, of a dozen groups of latte, in Gongna Cove at the north end of Tumon Bay, not separately recorded by Hornbostel, was investigated in 1946 by Osborne and Lt. E. S. Carpenter, U.S.M.C., who made test excavations, collected artifacts and potsherds, and restored the stone columns, as described in detail in Osborne's report, and briefly in his published note of 1947. The site is now completely overgrown by jungle, but easily found and in fair condition. As discussed below, it should be preserved and developed.

5. Hilaan Point

The site on the beach below (just north of) Hilaan Point, not specifically mentioned by Mrs. Thompson in her report based on Hornbostel's work, is described by Osborne. It is in the bush south of the NCS beach, close to an old truck trail. There is a single set of five pairs (one pair overthrown and disturbed in the edge of the trail) of rough slab verticals, the heads (all fallen) neat hemispherical unmodified brain corals about 30" in diameter and 18" thick. The latte are aligned approximately northeast-southwest, roughly parallel with the coast, the rows 8'8" apart and the slabs in each row 9 feet apart. The typical halege is about 36" high; 28" wide at ground level, 24" halfway up, and 10" at the top. There are potsherds and shells about, evidences of occupation to the southwest, and remnants of another latte group — two fallen slabs and one good taza; there is another scattered one, several pieces, up against the cliff base among papaya and coconut trees at an old farm, the Guillermo Flores place; remnants of still others are scattered by former house and farm sites and old roads. Potsherds are found widely over this entire area. No special measures are recommended for this site, which certainly would repay excavation.

6. Haputo Point

The large site which Osborne describes, and compares to Gongna, with five groups of latte and midden 16" deep, I did not succeed in finding. There are many scattered potsherds on the lowest slope of Haputo Point and the beaches below, and a small rock-shelter on the end of the point with deep midden — sherds and shells in black soil — in a small area of the slope beside it, and a similar concentrated little midden deposit in front of a large rock halfway around the cove just to the north and about 100 yards in. There obviously must have been considerable aboriginal occupation of the entire vicinity of Haputo Point.

7. Pugua

A small site on the beach below (cove just north of) Pugua Point was noted by Hornbostel and by Osborne; I did not visit this locality. There is no reason for it to have been disturbed since 1946.

8. Uruno Point

The distinctive latte group which Hornbostel (see Thompson, 1932, pp. 11-13) and Osborne describe as the Oruno site, with flat thin wide halege over five feet tall and with large narrow oval tazas with hollowed-out slots, evidently is not at Uruno Point but some distance south; I did not find it. It sounds particularly interesting, with its unusual latte, Hornbostel reports burials and artifacts, similar to those of Epau (Tumon Bay), from excavations there.

In the immediate vicinity of Uruno Point, and all the way north from there to Ritidian Point, occasional potsherds are found and a continuous occupation area is suggested, but only one latte site was found — about one mile north of Uruno Point, a short distance south of a fishing and farming camp — remnants of two groups of latte disturbed by trees which have grown up among them and by the rooting of pigs. There are comparatively few potsherds here, but black soil all around, extensive shallow midden. There are five pairs of halege in each set, mostly broken or overthrown; the tazas are missing or scattered. both sets of latte not quite in line with each other, the south group being "jogged" slightly inland (east), in the first (north) group, the halege are irregular, flattish, rough slabs around 27" high, 18" wide (narrowing to 10" at the top), and 8" thick, the rows 8'6" apart and the stones in each row 9 feet apart. The second group consists of similar slabs at somewhat smaller distances, 32" high, 28" wide narrowing to 14", and averaging 8" thick.

9. The North Coast Ritidian Point to Tagua Point

No latte were found in the vicinity of the Casa Real or along the coast to the southwest of Ritidian Point. Potsherds were found along the base of the cliff, where there are little rock-shelters, and one small cave containing no archeological remains, as well as along the strand. Osborne speaks of "several small latte groups in a poor state of preservation" in this vicinity; they probably are on around to the east. Dr. A. H. Bryan, Jr., says that one of the chief areas of latte sites found by Hornbostel "was on both sides of Ritidian Point, from Uruno Point to east of Tarague Bay" ("Notes on the ancient culture of Guam. The latte sites," Guam Recorder XIV, 1937. Mrs. Thompson's monograph simply mentions Ritidian and Rinepsan among monument sites, p. 15).

Caves at both Ritidian Point and Rinepsan Point are mentioned by Thompson; the former is described in a manuscript of Hornbostel's at the Bishop Museum ("Notes on the cave at Ritidian Point, Guam" — skeletal remains, potsherds, stone implements, charcoal, etc., scattered in confusion), and the Rinepsan cave, with a few stick-figure pictographs, by Osborne. Drawings were also found in a cave at Mergagan Point (Thompson, 1932, p. 20).

A group of exceptionally large latte at Rinepsan is also described by Osborne. I was not able to visit the Rinepsan locality. On the one visit made to the beautiful Tarague Beach area, the only locality on the north coast east of Ritidian Point visited in January-February 1952, no latte were observed. Any sites there are likely to be disturbed unless the Air Force exercises care in road construction and recreational development.

10. The Northeast Corner of the Island

No archeological sites are recorded by either Hornbostel or Osborne from Tarague Beach to Pati Point on the east end of the north coast, or from Pati Point south through the Anao vicinity straight east of Mt. Santa Rosa, and I did not enter any of this sector. The cliff descends abruptly to the sea with no beaches along here. Sites are nevertheless to be expected, although it is one of the few parts of the coast left unmarked on the map herein reproduced as Figure 1, and the area should be surveyed.

11. Janum

At the Janum locality, on the east coast near Catalina Point and just east (northeast) of Sagua Bay, evidences of intensive occupation — basalt mortars, slingstones, bones, abundant potsherds — were noted by Osborne; the locality is not mentioned by Mrs. Thompson and evidently was not investigated by Hornbostel.

Two sets of latte were found near the Taitano place at Janum, about or over 100 yards apart. One is of six and the other of eight medium-sized halege, about 30" tall, mostly standing but the caps mostly disappeared. Nearby, on the Taitano farm, there are several stone (basalt boulder) mortars, a few pebble pestles, and an abundance of potsherds.

12. Pagat Point

Here, also not previously recorded, is an extensive and important site, much disturbed by natural forces but still interesting and promising. At least some of the latte were standing up to about 10 years ago in good condition. There is now an extensive group of scattered fallen columns, with two stones standing in each of two rows and other sets evident. There were at least three distinct latte groups, Slabs (halege) now standing run 30" to 36" high, 8" to 10" thick and 18" wide at the top, about 24" by 14" at maximum width and thickness. The fallen stones strewn about include similar slabs; very rough, more or less, hemispherical, tazas, about 30" high by 60" diameter; and miscellaneous big rocks. There are two basaltic boulder mortars, and an abundance of potsherds, midden deposit over a foot deep. The entire site is about 200 feet along the coast by 75 feet from the strand to a rock point standing just inland. The site is in dense growth, with many pu-ting and chopag trees, just above the narrow beach. Pagat Point is reached from a trail which goes down the cliff southeast of the Marbo Poste, exchange and the Air Force housing area. Ownership was not determined, but there is no modern occupation of this section of the coast by Guamanians.

13. The Northern Interior

Very few archeological sites are on record or to be expected on the plateau of the northern half of Guam, in spite of the abundance of remains all around the coast below the cliffs. Any that formerly existed, furthermore, are likely to have been destroyed by airfields and other modern construction except in the central part, where there seems to have been no occupation at all, prehistoric or modern.

Two interior sites are described by Osborne, only one with latte, at Mogfog — this group apparently has been removed since 1945 by military construction, as none could be found in the vicinity of Mogfog and around where oldtimers stated the latte to be. Osborne also refers to evidences of prehistoric occupation east of Yigo (no latte) and scattered potsherds throughout the vicinity. A large former latte site at Dededo had already been destroyed in 1946 by airfield construction.

14. Pago

The former settlement of Pago, occupied 1681-1857, on the northwest side of Pago Bay, in cultivated fields on the Perez property, is marked by abundant potsherds, including much Chinese and/or Japanese porcelain as well as the locally made coarse thick brown Marianas Plainware, but no latte or other visible remains of structures. Investigation of this site would be a particularly interesting study.

15. Yona and Pulantat

The marine encampment close to Yona cleared away a number of latte; remnants of at least two groups were seen in February 1952, about 1/4 mile northwest of the village. The large Madog site near Pulantat is reported to be still there undamaged; it was described by Osborne in his 1947 published paper as follows: "In a dense tongue of jungle above Pulantat, near the radio station, is one of the most spectacular sites on Guam. It has not been previously mentioned * * * One of the several latte groups is in an excellent state of preservation. The supports are well shaped and smoothed, and the caps approach the type previously recorded for Fena (Laura Thompson, 1932,p. 14). Another set has the most massive supports yet seen on Guam. The single standing stone is nearly 7-1/2' high and 3-1/2' in diameter. With the cap in place it must have been 10-1/2 or 11 feet * * * many smaller contiguous latte. Unusual and asymmetrical shapes and sizes are common. All of these are buried in nearly impenetrable jungle and limonchina thickets. * * * Other smaller sites exist nearby." (American Anthropologist 49-3, p. 522). Additional detail is given in Osborne's manuscript report. I did not visit this important site.

16. The Ylig-Togcha Sector

A latte site in the Ylig vicinity is mentioned by Thompson (1932) and described by Osborne, and a similar one in the Togcha vicinity is mentioned by Osborne. Neither was found in 1952, but they may well still exist. This area of the east coast, between Talafofo Bay and the Yona-Pago Bay area, looks promising and has never been intensively surveyed; it would repay investigation, along the coast itself and up the little streams and the entire Ylig River drainage, which should contain latte sites comparable to those of the Fena River basin. There are remnants of at least two destroyed latte sites in farms alongside the road about halfway between Togcha Bay and Camp Dealey.

17. Talafofo Vicinity

The other 50% of the Talafofo River site at Asoiroga has disappeared since Osborne's brief description of it as half destroyed by the road (Highway 4) in 1946. Potsherds were collected below the hill by M/Sgt. A. L. Aitken, USAF, in the last few years. No archeological site appears to survive in the immediate vicinity of the town of Talafofo. There seem to be no pictographs in the Talafofo caves; those referred to by Mrs. Thompson and others (e. g., Lt. F. A. Stephenson, "Talafofo cave writing," Guam Recorder XIII, 1936) must either be intended for Inarajan (see No. 19 below) or else are imaginary interpretations of natural streakings or else have been covered by dripstone and lime incrustations — which would have long since concealed any very ancient pictographs in at least the smaller caves here and in much of the largest one.

18. Dandan

At this locality, a mile north. of Malojlo and hallway between Talafofo and Inarajan, just west of (inland from) the highway, there is an extensive occupation site with two groups of latte and scattered remnants of others, on the farm of Mr. Jose San Nicolas and in the Martinez pasture. These are large stones, 3-1/2 to 4 feet high but very thick, in groups oriented roughly NW-SE and NE-SW. Potsherds are abundant. No sites around Malojlo are mentioned in Thompson, 1932, but Hornbostel is said to have dug here and found human bones. Other sites are reported in the general vicinity and north toward Talafofo.

19.Inarajan and the South Coast

The cave on the north side of Inarajan say in which there are several rather small stick-figure pictographs in white is a little wave cut cleft just above the water, with no floor; I do not see how it could have contained potsherds and broken stone implements (Thompson, 1932, p. 20) of any considerable antiquity or in any quantity.

No latte have been reported in the vicinity of Inarajan, or anywhere along the southeast coast, from Malojlo (two miles north of Inarajan) around to Merizo, but there surely must be archeological remains of prehistoric and historic Chamorro occupation in this area. A find of potsherds at a depth of 13 feet on the Merizo-Umatac road near the Toguan River in 1926 is reported (Thompson, 1932, p. 31). Many shell gorges and fishhooks have been found on Cocos Island.

Latte sites just inland from Merizo and extending up as far, apparently, as Jalaoian have been heard of but not visited. There is none surviving or known to nave existed at Merizo or close to the coast in the vicinity, or on up the southwest coast past Umatac and Ceti Bay to Facpi Point. A small site with a few latte stones is reported on Facpi Point by M/Sgt. A. L. Aitken, USAF.

Potsherds are common along the strand at Umatac, Ceti Bay (as mentioned by Osborne, ms.), and Seja Bay, but there seem to be no latte whatever. No sites in this part of Guam are mentioned by Mrs. Thompson. Osborne, however, found one good taza near the beach at Ceti Bay, and thinks there may have been a latte group there; he collected a particularly interesting range of potsherds here.

20. inland from Umatac

A partly overthrown latte group on the right side of the road on a hillside about two miles from Umatac toward Mt. Lamlam and Agat is mentioned by Osborne (1947, p. 522, and ms. report). No sites in this vicinity are mentioned by Mrs. Thompson; but this one is said locally to have been dug into by collectors for items to sell to Hornbostel in the 1920's. it has not been noticeably disturbed further since 1946. Three slabs in one row are still in place; heaped at one end of the set are three more halege and a half-dozen capstones, at the other end are two more tazas. No potsherds were found here (Osborne collected several).

There is another set of latte just off the Umatac road, below it on the north on a slope, a few hundred yards southwest of the first site above: five pairs of good heavy 3-1/2-foot vertical columns, spaced 12 feet apart in two rows 15 feet apart, all heads present but all fallen and some broken.

Several other latte sites are known to exist inland from Umatac; mainly in the Sologna Valley, at locations within view from the first Umatac Road site — one well up in the valley and almost clear across it, at N. 70° E. from the Umatac Road site, one down and to the close to a little knoll 35° S. of E. from the Umatac Road site. Another, the Pajon or Pajung Valley site, is across the Sologna Valley the east, and beyond it is reported the Fanji site, a group of 12 columns 3-1/2 feet high "standing in a square" (presumably 6 pairs in 2 rows), according to Mr. Joe Quinata of Umatac.

21. The Talafofo River Basin

Only in the comparatively extensive drainage of the Talafofo River have latte sites been recorded in any number in the interior of Guam (Thompson, 1932; H. G. Hornbostel, "Notes on inland latte," ms. at Bishop Museum; Osborne, 1947, and ma.). Several of importance have been separately described:

Acapulco (Osborne, ms.) — a large site on a tributary stream safely below the Fena dam, believed to survive undamaged (not visited in 1952).

Fena (Thompson, 1932, pp. 14-15; Osborne, ms.) — has been destroyed, and entirely submerged by the Fena Reservoir.

Mepo (Thompson, 1932, p. 14; Osborne, ms.) — removed in the construction of the entrance spur for Magazine 173 in the Naval Ammunition Depot, the very large stones heaped on either side of the entrance.

Chandija (Osborne, ms.) — an important site, largely destroyed by a NAD road.

Bona (Osborne, ms.) — a small site, evidently has disappeared except for the stones which have been set up near the NAD entrance.

San Isidro (Osborne, ms.) — could not be located, and may have been removed in the course of NAD construction.

Other sites have been reported to exist in the basin, and those downstream from Fena toward Talafofo presumably survive. Very possibly there are important sites to be found along the Ugum River, major tributary from the south. There seems to be none on the grassy uplands on the north of the Fena basin crossed by the Santa Rita-Talafofo road.