Observations on Pliluwat and Gaferut, Caroline Islands With

------..----

I:O. 76

Observations on PLiLuwat and Gaferut, Caroline Islands

William A. 3Lering

with

TistoricaL axl ciimat5c illformation on Gafemt Tsl~d by Pkie-H&l&ie Sachet Observations on Puluwat and Gaferut, Caroline Islaads

illA. Niering*

Observations presented in this paper were made during the summer of 1954&n conjunction with the Kapingamaraagi Xxpedition to the Caroline Isl.ands . On the return route to Gttam, several other island groups were visited for purposes of comparison. Among these were Fuluwat and Gaferut.

I. List of plants noted on Puluwat--- Islet, hluwat Atoll The following species were recorded during a one-hour visit to the islet on Sept. 3, 1954.

Allophylus timorensis 31. Occasional in undergrowth.

Artocarpus altilis (Park. ) Fosb. Scattered in interior. Coll. no. 765.

Asplenium nidus L. Vomen o3served carrying leaves, growing plants not seen.

Ca.Lophy1lun inophyll,lm L. Large trees along lagoon shore.

9lavalia serice a Gray Frequent in forested and seni-open areas. Co!.l. no. '769.

Carica papaya 1'. Frequent around homes. . . Cassgtha filiformis L. . . Occasional on. m3ergrowth -Cocos nucifera L. Trees scattered forming open plantations, also yo~ingplantings.

Colocasia esculent2 (L. ) Scbo-Lt Importaat food species.

X- Department of sotany, Connecticut College, TCetr London, Connecticut. H Sponsored by t'ne Pacific Science Board., Nati.ona1 Academy of Sciences-- National Research Council, and sl~pportedby the Of-fice of Naval Research. Cordia subcordaa~Lm. Large specimens along lagoon beach.

sp. Profusely f l.o~reringalong pa-:hs.

Cucurbita sp. Large productive squash in living area.

Eleusine indica (L. ) Gfiertn. Frequent in open disturbed sites. Coll. no. 771.

Euphorbia ck:amissoi~is Boiss. Occasional along sandy lagoon beach. Coll. no. 770.

Ficus prolixa Forst. toll. no. 767. -F. tinctoria var. neo-ebudarun (Smn. ) Fosb. Coil. no. 766. ---Guettarda speciosa L. Occasional in interior.

!Iedyotis biflora (L.) Lam. Coil. no. 7z. -iiibiscus tiliaceus I,. Occasional in understory.

IIIwenocaliis lit-L;oralis (Jacq. ) Salisb. Coil. no. 772. -Ixora sp. Bed flor,rered ornamntal around houses.

Lelsturus repens 13. Br. Occasional in disturbed sandy areas. i4esserschmidia argentea (Lf.,\ Johnst. Mixed with Scaevola along ocean beach. Morinda --citrifolia L. Occasional as understory tree.

!m sp. Frequent around hov.ses . Nepkrolepis hirsutula (Forst. ) i'resl Dominant fern in coconut plantationr; orten forming a ciense ground cover 2-3 feet hi&.

Pandanus tectorius- Park. Frequent, often more common oceanward. Pasplum conjugatm krg. Occasional inzei~disturbed areas. Co1.l. no. 763.

Piper betle L. ? Coil. no. 768. g. fragile Eenth.? Coll. no. 764.

Plmria r- bra L. Occasional in l.iving area.

Polpodium scolonendria Bum. f. Frequent on trees, o:iten associ.a'ted wi-th 3ep:irole&,s.--.--.... -- Portulaca sp. Yellow flowering form, frequent along arid in paths.

Premna obtusifolia R. Dr. Dominant in undergrovth.

Scaevola sericea Vahl Frequent as a hordex- along beaches.

E3alassia i~em~richii-_-f___ (Ehrb. ) Asctiers. Frecpent in shallo~gsandy lagoon waters. muarea involuta p orst st. ) X. 8; S. Frequent in recently cleared areas pla;?ted t3 cocoilut. -Tkhnfetta procmnbens Forst. Occasional along ocean beach.

Vernonia cinerea (L. ) less. Weed of open areas. i?ed.elia-- biflora (L.) E. Frequent in undergrowth.

Mangrove Present along soui;n lag002 cove (senus not ietermined).

11. Description of Gaferut---- Island Gaferut Island, an isolated land mass in the northern Carolines, vas observed during a brief stop on the morning of Sept. 4, 1954*. The island is approximately 1,500 feet long and 500 feet wide, situated on a somewhat crescent-shaped reef which extends 500-750 feet outmrd from mean tide level. Gaferut is at present uninhabited, but the Japanese mined phosphatic rock there around 1935. Remnants of their buildings, clearings and exca.vations can still be seen. About an acre in the interior, although nov overgrown, sho~rsevidence of being cleared. In at least four areas excavations were observed, the most extensive being a trench about 2 feet in depth and of considerable r?idth. The other a.reas within the clearing vere quite small.

Geology and Soils

Geologically the island is composed of three types of material: phosphatic rock, coral rukble and sand. The higher interior portion is underlain by phosphatic rock which gives my to a marginal coral r-ubble border 100 feet or more in width on the east and southeast sides. Along the riestern shoreline hi& step-like beach ridges composed of rubb3.e extend 6 feet or more above mean tide level. T1ie sandy deposits form a conspicuous elongated bar extending in a northwesterly direction. ??ear the end, of the sand bar is a very large coral boulder which tras probably perched there during a severe storm. These various areas are readily discernable on the aerial photograph (~ig.1). . . . The interior phosphatic rock is ei'&er exposed, except for blaclr- ish algal or fecal coverings, or overlain with a thin rubble soil or humus layer. ' Several indhes of rubble mixedwith organic matter axe typical. However, in the 1.argest clearing rubble is absent and the rock is overlain by a. dark br0i.m friable acid nunus (pH 5) Varying from.2 to 3 inches in depth. In one area within Gut near the edge of the clearing it was found to a depth of 6 to 8 inches. giese variations indepth may be correlated with the disturbance. One saztp1.e fromthis area con- tained large quantities of small gastropod shells. Beneath the humus the underlying rock consisted of cemented coral fra@lents, sand and foraminiferal tests, whereas those from tne trenckconsisted primarzljr of larger rubble. In both cases tne material iras similar inthat it exhib- ited a brownish salt and pepper appearance'andwas relatively soft so that it coilLd be easily broken with one 's' fingers. In the trench the roclr was exposed for at least 2 feet in depth with no evidence of an unconsolidated layer beneath. lbe hunus layer aqd-underlying phosphatic rock have all the charac%eristics of the Jemo Series named and described by Fosberg (1954) and observed by other investigators (Hathemy 1953, Niering 1956, NcKee 1956). Fuxther discussion concerning the possible origin of the roclr &1 follow in a later section.

$5 The author wishes to aclfflor?leae;e the most helpful suggestions of Dr. F. Raymond Fosberg and Miss Marie-E616ne Sachet in preparation of this manuscript. Vegetation and- Associated Animal. Life.-

The general vegetational aspect is that of a low, relatively open forest, 12 to 25 feet in hei&t. The dominant tree which cha.racterizes the island is Tournefortia argentea., ra.ther than the coconut; ril~ichis so tmical of inhabited islands. Oi1l.y two mature coconwt trees (COE nucifera) 50 to 60 feet in height, probably plaited within the last 10 years, were found. The only other woody plant observed was Caesalpinia sp., and it was unixportant. On the phosphatic rock of Y~ie interior which has been least disturbed the trees are re1at;ivelg large and freqcently form a continuous but open canopy, in contrast to -the marginal rLCable borders vhere the trees are smaller and more scattered. On this coarse marginal rubble the 'Io~~nefortiafrequently exhibit a shrubbj done-like appearaxe with branches extending to the geovmd. In t!~e rubble =eas tne trees reach 6 to 10 inches in di.meter and in the interior sections attain diameters of 12 to 18 inches. The doininar~tground cover is Fleurya ruderalis which f~insa continuous layer 12 to 18 inches in height in the openings and decreases slightly in the semi-open situations. Several specimens of a cucurbit sp.) also occurred locally.

Associated with the '~ournefo_r-&iacoimmity is a luge bird population including frigate birds- minor paherstoni),-- red footed boobies (Sula sula ru3ripes), an& white teri(Gd- --al.ba ---candida). Of these the fri~ate- birds are most abundant. Their nests m.d imatu-re specimens were conspicuous in the trees. At all times the air vas filled with the din of hundreds of birds in constant flight. From iz irodachr~me tak.en over the isl.and an estimated 550 birds were counted. Vhile valki.ng through the interi.or one had to be careful of these large 'birds since one could easily have been hit as they lost altitude on the take-off. Under those trees with many nests tile stench ~msvery pronounced aad the ao- sence of Fleurya may be correlated with the concentration of guano. It was truly amazing to see such gyeat numbers of birds but presumably this is not&ypical of certain uninhabited islands. Pokak, an uninhabited atoll in the i'iiarshalls, is quite similar not only in its la-fge bird population but also in its sparse flora (Fosberg 1957). Here 9 species of vascular plants were found in contrast to 7 on Gaferut. Associated with the coconut were azure-tailed slrinks (~moia-- cyanura cyanma) and coconut crabs (w1.atro). The former were especially abundant in the trees; the latter were fonnd under the fronds on the ground as %re11as within the nearby herbaceous cover. The largest crabs, one foot or longer, were in the cavities of the rock. Althovgi~ten nuts had sprouted under the trees they mre partly chewed and the 1 to 2 foot shoots were badly damaged. :2..ether they r,rill survive is qu.estionable . &her animal life observed includ.ed hermit cra'ss and in the branches of the Tournefe orb-weaving spiders were common.

In the large clearing tomra the south end of the island dense :, .... growths of Epomoea tuba cover most of the opening and are invading the surrouding aefortia and fo~minga complete covering over the trees. Several have alrea.Gn killed a.s a result of this invasion. Infesting this viny growth was a caterpillar which develops into a whitish Lepldop- teran. Leaves not hnaged by this infestation were difficult to fincl. Another herb found locally in the clearing was Boerhavia dil'fusa, both the pink and white forms. At the north end of the island most of the sand bar is devoid of vegetation. fIowever, small Tournefortia are becoming conspicuously established at the south end of the bar. In this'sector sea turtle activity was evidenced by the many excavations in the exposed beach sand. Mso along the shore a flock of 12 to 15 turnstones (12renmia intei-pres interpres) were seen as 7iell as several plovers in flight.

--Discussion The origin of the phosphatic rock is of considerable interest in that it resembles the Jemo Series. The R (htunus) and B (rock layer) horizons are comparable but the less consolidated C horizon was not found. Although the greatest depth reported elseiihere for the cemented B layer is 1%feet, from Jemo Island, this may merely indicate that cemnei-.tation on Gaferut has talcen place Lo a greater depth. To be associated wit11 the Jemo Series infers a dual-biotic relationship only one component of which currently exists on Gaferu'c - namely, a large bird popula%ion. The other facet - a ~isoniaforest, is wanting. Could there have fieen' a PisonLa forest in the past - tile soil forming processes operative for a long period res~fitingin the formation of the phosphatic rock and then tile forest clestroyed by a typhoon? (See section on climate below). 1)e- nudation of islands resultin from typhoons has also been reported from Ailinginae and ULirik Atolls in the northern hlarshalls (Fosberg 1956) and most recently by Blwnenstock (1956) and Fosberg (1961) on Jalu-it Atoll in the Marshalls. Tile small very intensive typhoon 'i,rhich hit Jaluit completely removed .the vegetation on the narrower parts of certain isletc. Many trees were uprooted or snapped off and i,&shed away. Large pisonia-.-- trees were u.prooted, mcl others still standing had many ,braaches blown off and were greatly defoliated, During the hei&t of the stonn wave surges 6 feet in height and locally more than 8 feet swept over the is- lets accompanied by winds approaching 125 lino-ts lumenst stock 195G). From these observations it is not unreasonable to ass-une that a Pisonia forest could have been destroyed on Gaferut (see note on p.13). Surely the very soft nat~ireof tFe wood wouid lend itself 'to tremendous storm damage. Although persisting root systems riould tead to prod-me vigorous root suckers prolonged salt water' inundation right well.. have killed a?g remaining root systems rIrliich ?.rere not compJ.etely'washed away. IJ. small isolated island such as this one would be partic:!.larly vulnerable to heavj. damage since it Gould get the 'full impact of the storm regardless of .the direction from which i-t came. That a severe stomkas hi$ Gaferut is evidence& by a large coral bo.ilder off the northendof tile isla.nd. :liens (1959) observed large blocks of this t-ype on Jzluit which were washed LOO- 300 feet during the Qphoon. In addition, the extensively developed inar- ginal rubble border along the east and south sides of the isl.and and the high beach ridges oiithe west may well have been laid down during such a storm. The fact that..pbosphatic rock fo'mtion does 'not appear to be occumbg at present and has never been reported under Townefortia sw- gests that it occurred under a different vegetation type presumably Pisonia grandis.

Although the idea that a severe storm may have destroyed t:le ---Pisonia forest is most tenable, its removal by lnan in clearing the land to facil- itate the mining of phosphate is another possibility. Celeterious effects of the existing bird population were not strikingly evident as has been observed by :Pt.the77ay (1955) on Canton Island, where he found dead and dying Tournef~~tia,preslmal?ly associated witl~the concentrated guano deGosi%; this is probably correlated ~riththe drier clkte of Canton Island in cont1:ast to Gzferut. Gn Gaferlxt the larger trees heavily used by birrls ?.rere not as vigorom in appearance as those in other areas but no dead trees >reye noted resulting from this factor. IIovever, as mentioned above, the sparsity of herbaceous cover and seedling reprod~dcionmay be correlated with excessfire &az.o. Perfodic visits by nativss may also play some role in reducing the bird popA.ation, but their present density vould suggest that this influence is negligible.

In interpreting future tre~dswithin the existing vegetation it appears that Tournefortia will probably pe-mist as the typical vegetation for some time since there are no other com3etitive species available vhich might replace i4;such as Fzo'a or --Ochrosia ssitifolia.--- Wuether or not adequate reprodv.ction will occw to replace mature .trees in the future is questionable at this point. Tfle two coconut paLm that wre presumably planted do not appew to be spreading. In fact, the poor vigor of the sprouted nuts as a result of coconurt crab activity would suggest that the pa3m ~~i~lprobably not become an important part of the vegetation. Tbis may well be a limiting factor in the establishment of coconut, even if introduced, on uninhshited islands or those ~rLeretke coconut crab population is not kept in ckec!:. Since the crab is comid- ered such a delicacy by the natives it presents no serious threa-t on inhabited islands. Another speci.es, Iponioea tuba, must also be considered since i.t has already enadfed and killed seve;?al ---.-To~urnefor-Lia. It is not impossible to visucllize this very aggressive and drougilt-resisCant species as eventually becoming dominant over an increasingPy extensive portion of the interior.

1. Gaferut, an island 1.11 the northerr. Carolines, is dominated by a low forest of Townefortia ar~entea.- Associated with this comriiunity is a l'wge bird population, primarily frigate birds and red footed 3oobj.e~.

2. The underlying phosphtic roc:: resem5les %ha5 of tie Jemo Series. Its presence suggests that formerly a Pisonia forest existed on the islan?. and has since been destroyed perhaps by a typhoon. A large coral boulder perched on the reef is intiicative of a severe storm in tlie past.

3. Tovrnefortia will probably persist os the dorninant vesetation for some time. Thecoconv.t -uahs. , -nresw!uXbl~r " nlailted. do not amear- to be spreading. Ipomoea t~ba-which has aireddy ki.lled severel To'~li^~iefortia may become increasingly important in tne futlxe. Blumenstock, David I. at Jaluit Atoll in the Marshall Islands.

Fosberg, F. R. Soils of the Xor-them iVarsha11 Atolls, with special reference to the Jemo Series. Soil Science 78(2): 99-107, 1954

Military geography of the Northern Marshalls. Engirieer Intell.igence Dossier, Strategic Study-Mzrshall, Srlbfile 19-Analysis of the iJatu.ra1 Enviromxat, 1-320, 1956.

Lonely Polralr. The Living Wilderness 22(62): 1-4, 1957.

Plo-ra aid vegetation, in: Blw~enstock, D., ed., A report on typhoon effects u$on Jaluit Atoll. Atoll Research Bull. 75: 51-55, 1961.

Hatheway, 1. I;. The land vegetation of Arno Atoll IvZ3rs!m.ll Islands, Atoll Research Bull. 16: 1-68, 1953.

The natwal vegetation of Canton Islaad an equatoria.1. Pacific at~ll. Atoll Research Bull. 43: 1-9, 1955.

McKee, E. Geology of Kapingamaraagi Atoll, Caroline IslanGs. Atoll Research Bull. 50: 1-38, 1956.

Niering, W. A. Sioecology of Kapinga~narangiAtoll, Caroline Isl.ands, terrestiiial aspects. Atoll Research BuJ.1. 49: 1-32, 1956.

Wiens, Herold J. Atoll develonment and morphology. Annals of the Assoc. uf .b. Geographers 49: 31-54., 1959. Fig. 1 Aerial view of Gaferut Is?.and taken April. 1944, 10 yews prior to the present ohsema.tions. Light way vegetation dominant throughout is Tournefortia argentea. Rubble areas are evident, demarcated by the scattered specimens of Tournefortia, especially on the eastern and southern sides. On the phosphatic rock of the interior, Tournefortia forms a somewhat circular patfern around a uniform darker area rrhichrepresents the section cleared for phosphate digging. In 1954 the cleared area was dominated by Ipomoea tuba. The sand bar bvil6ing nor.th~.restwardi.s devoid. of vegetation. If the coconuts were present at the time the photo vas taken, tiley were too sinall to he detected. There appears to have been relatively llttle cha.nge in the vegetational pattern in the 10 year period.

Photo courtesy of U. S. Navy.

Historical.and.Clhatic information on Gaferut Island

. Mazie-HeL6ne Sachet

In view of the scarcity of information on Gaferut, it seemed worth- while to utilize the library resources of Washington in order to supple- ment the valuable observations made by Dr. Riering.

The real native name of Gaferut is Faiau, Pallao (Spanish spelling), or Fayo (which means stone or rock in the Woleai language of nearby islands). The name Gaferut is never included in the old lists of islands of the Carolines, and, according to Smith (l.951, p. 30), is never used by locaL people. Chamisso (1821, p. 115) relates how Caroliniaus from Woleai, Lamotrek and other atolls went every spring (~~~il)to Guam, stopping on the way for several days at "Fayo, the desert island,;' and returning in May or June by the same route. Riesenberg and 1laneshiro(l960, p. 285) identify this stopping place as Gaferut. Chamisso also discusses some of the early descriptions of these central Caroline atolls and in- cludes a chart modified from Cantova (1728), in riliich our Gaferut is obviously the island called Fauheu.

Of this island, Chamisso says also (p. 1211) that it is mi.nhabited, without fruit trees or fresh water, which only collects in pj.ts after rain, and that the inhabitants of near-by atolls visit it to collect turtles and birds. Farther on (p. 196) he says that the god of the desert island of Fajo is called >ag< and (P. 205) he ~riteg:"CbYne desert island of Fajo, as at Bygar / Bikar ,in the lvlarshalld, fresh water is conjured into the water There is a species of blag-bird hrobably the frigate-bird, which was sacred on Sorol and Ful.uwat/ which is under divine protection on this island, and not permi4;ted to-be eaten." Chamisso had this information from his friend and informant, Ih3.1, a native of Woioleai.

:.hat then was the real Gaferut'? It was either an imaginary magic island, or perhaps a former islet of a reef now devoid of dry land (cf. legend of Ngaruangl in Gressitt 1953, p.2). According to Krher (1937, foot-note p. 346) Gaferudj is a name for @pen, a fomer atoll near Yap; XU(17, p. 304) mentions that Sepin is a sunken magic island, cul- turally linked to l?~~~uzzg(northernmost island of Yap). All this might explain why Senfft (1906, p. 284) was told that Gaferut was a devil's island and never visited by the Caroline people, ~ihowere afraid of it. The report then would apply to the real Gaferut, -rather than to the island we now know und.er that name, and which Sei1ff.t visited (see below).

The confusion of, names vas recently clarif&edby Smith (1951, p.28) who calls our Gaferut Fayaew (in his own systemof phonetic spelling), and adds (p. 29): "'Gaferut' in turn is a bastardieation of a.Yapese name, even though Yap has onljr very remote concerns for that island." Fayaew belongs to the Faraulep people. When the name Gaferut was first applied (or misapplied) to the coral speck in 9'14'~) 145°23'~has not been ascertained. The island is also called Grimes (Gurimesu-to in Japanese) from Captain Grimes of the sht~ J* (Findlay 1e70, p. 766), who discovered an . island 5.n lat. 9°16'~., Long. 145O43'E. !Ie described it as hi& and well-vooded, of 6 miles in circumference, so there is possibly some confusion with some other island, perhaps Pais. Findlay adds:" It has since been announced as -IIigh =, at lat. .gO1l'N., long. 145°h5 'E. .. ." Tilese names are generally considered as synonyms of Gafer~rt.

One of the few references to Geferut in the literature is an account of a visit by the German District Administrator of Yap, A. Senfft, in kc. 1905. He described it (1906) as a flat sand bank, only locally reaching a height of 2 meters. The only vegetation noted was "a species of mangrove." Countiess seabirds >rere nesting in the trees or on the flat ground. Coconut crabs were also observed, and tracks of large sea-turtles. Senfft also noted that a violent storm must have recently hit Gaferut, as most trees had broken branches and some very large ones were completely uprooted, Later, a Gemexpedition exploring for phosphate deposits is said to have discovered phosphate on Gaferut (Aso 1946, p. 117). According to German sources (Sapper 1910) this expedition took place in 1907, not in 1903 as reported in the translati.on of Aso. As a result of their dis- coveries, the Germans started exploiting phosphate on Angaur, PeliLiu and later Fais, but never on Gaferut. Horfever, the Japanese did mine phosphate there, inspite of great transyostation diff'iculties, starting about 1937.

In addition to' the phosphate workers, other Japanese visited Gaferut, among them Yata Haneda, a mycologist interested in lknous fungi, vho mentions Gaferut a.nd its nhosphate in an account of his i937 travels i.n Micronesia (1939). !t'he distinguished geologist Eisaburc Tayama apparently also visited Gaferut, and he incluaed a description of it, maps, sections and a photograph in his volumes on coral reefs of Micronesia (1952). IIe wrote (p. 262): "The table reef of Gaferut is a cmde half circle with the convex side facing east. The length of the arc is 1.1 km. Gaferut is the only island on the reef. The shape of tine island corresponds roughly to that of the reef. The northern half of the island is chiefly sandy and the southern half primarily gravelly. Recent limestone (~ig.1011.) emergent at low tide, is best exposed, toward the western end vhere it strikes northwest and dips 5 to 10 degrees to the southrfest. Four recent limestone ridges may be discriminated neax the southern coast. The inner ridges are of foramf.nifera1 sandstone and the outer of coral conglomerate; they strike East-Vest 'and dip 5 degrees. The central part of the island is flat-toppedaad rises 5 meters abo-~ethe reef-flat. The upper surface is level and built of coral limestone (~ig.205). This limestone is alter- ing to phosphateore; it conformably over!.ies a brown clay; and the brown clay, in turn, conformably overlies a foraminiferal sandstone, and the sandstone, the coral gravel and foraminiferal sand bed.

"The 'eef -flat is extremely vide on the northrest side. The inner zone of the reef-flat is not exposed a.t low tide, and is dotted with shallow pools about 0;5 meters deep, Seaweeds axe graving over the reef floor, and mushroom rocks, 2.5 to 3 meters high, are standing here aild there. " Tayama's fig. 103 (9. 116) is a small-scale sketch map with bathy- metric contours, and giving 2.7 m as the height of a rock on the reef. Fig. 104 (also 146) is a profile of the east coast, from the reef front to the bedded phosphatic rock of the interior, showing a mushroom rock, beach-rock ("recent limestone"), beach sand and gravel betmen the reef and the phosphate platform. 12x2 height of the latter is shown as 5 m above low tide, Fig. 105 is a "Colmmar section of beds exposed in p4t on Gaferut Island, Gaferut 'Table Reef, " including: "a. Surface soil-Blackish br07~(20 cm) b. Coral limestone-Phosphatic and include abundant Yricbcna (35 cm) c. Brown clay (25 cm) d. Foraminifers limestone-Somewhat phosphatic (65 crn)'' lying on "e. Foraminifera and coral sand." An analysis of the "phosphatic reddish brown clay intercalated in the cay sandstone" is even on p. 265:

3.28 48.97 3.17 2.50 19.35% Elsewhere (1942) Tayama had rerflar~edon the anazingly high percentages of silica, iron and aluminum oxides. In appendix I of the 1952 ~iorlris a photo (fig. 48, p. 35) of Gaferlvt Island. Concerning this F. R. Fosberg says (personal comunication):"Tke presence of bedded phosphate roc!: on Gaferut suggests that a vegetation of Pisonia grandis my have existed on the islmd, and the presence of humus on the surface indicates that his must have been in the very recent past. The comple4te absence of Pisonia now, as indicated by Xering is indeed remarkable. Tayama's photo suggests that Pisonia may possibly have persisted until at least the dzte of the photo, as the much tal!.er forest on the right eide of the picture has the aspect of Pisonia, though the reproduction is so poor that this can be regarded only as an impression rather than a certai.nty."

There are almost no descriptions of Gaferut in the literature, other than the brief German and Japanese texts. Tile Sailing Directions (U. S. Hydrographic Office 1938) described it briefly as fol?.ows:"lorr, thickly covered with trees, and encircled'by a reef, mere are no coconut palms and no inhabitants, but natives from Faraulep Islands visit it to catch birds, vhich are numerous on it." In a later edition (l952), the i.nfor- mation was emended to yead:"Gaferut is low and covered with trees. Some of the cocorwl; palms attain a height of 65 feet. Tnere were no permanent inhabitant;? in 1935, but since. that time phosphate mining has been reported. Numcrous birds exist on the island, but no food or fresh water.

"The =em high-water interval at the island is PI. 30m. Keen high- water spring tides rise 2: feet." --Climate The climate of Gaferlt is a humid tropical one, with litt2.e seasonal change. There are no records from the island iLself, but in this area of the ocean, temperatures vary little around %he year, the mean average air temperature being about 820F, vith daily variations probably not esceed- ing 15O, and usually less. Atmospheric ilimidity is always high. It rains throu&?ut the yeas, with probably higher rainfall in the sumer. The to.ta1 amount of rain must be in the nei@ibo;.hood of 100 inches (less than l'ruk to the sw~.thand more than Guam to the north). At Lamotrek, somewhat to tile south, the rainfall vas 1.04 incnes per yea, based on 4 years of record.

The wind regime is probably the climatic component most affected by seasonal.ity. The north-east trade winds are rather steady in the mkter and spring, and northeast, north and east 7rind.s prevail. From June or July to October, the winds we more variable, with often a strolie; component from southvest, vest or south. . .

Every year some tropical storms or typhoons originate in an axea between Gaferut and Truk., and mmy of 'them, .travelllne; north~.restt~ard to~rasdGuam, must pass new Gafemt. A direct hit is probably not too frequent, but very strong winds and high w.ves mst occur rather often. !Illat Senfft observed the results of the passage of such a storm in 1905 is itlily Even Oyhelia I, ~?hichwas so damaging in Jaluit, passed not far from Gaferut toyard the end of its destmc+,ive career, and so must have Ophelia 11 (P!ov. 30, 1.960) which dcvasta'ced Ulitki. Such storms can occur any month in the year i.n this part of Yne j?~~cffical- . . though they are more fi'eqi~entin the sunner and fall. . .

Bibliography-.

: . Aso,Y. .' . : Phosphates. . '. 1-209, 1946 (mimeographed translation Ro. 399 by U. S. ? available in Geological SurVey Gi1brary). . . Cantima, J. -A. . . ' Lettre .. . au li. P. Gvillaimie Daube~~ton. ... Lettres 6diriantes et cwieuscs 18: 188-247, 1720.(niap tipped in betwen pp. 188 and I.&). . . Chamisso, A. von Remarks and oainions . .. of tne inatwalist of tine e;q)edition: in, Kotzebue, A voyage of discovery.. . 3: 1-36, 436-4142, 1821. Findlay, A. G. A directory for ,the navigation of the north Paci:fic Ocean. 2nd. ed. 1-1007, London,. 1870......

Gressitt,. J. L. Notes on Ngarumgl and Kayangel Atolls, Palau Islands. Atoll Research Bull.. 21: 1-5, 1953. Haneda, Y. Lx trip of the outlying is!.ands of ~apjT Collecting an& Sreeding ~(9):419-430, 1339 (not seen, amotated in Foster, B. L., .&nnotakd Bibliography of geologic and soils literature of iqestern Xurth Pacific Islands, 1956).

Kriiaer, R. Zentralkarolinen 1 (Lamtrek Gmppe, Oleai, Feis ) : in, G. Thilenius, Ergebnj.sse der ~GdseeExpedition 1708-1910, I1 R LO, 1: 1-413, 1937.

~:!cller, W. Yap: in, G. Thileniils, Ergebnisse der $idsee Expedition 1908-1.910, 11 B 2, 1: 1-380, 191'7.

Riesenberg, S. H. and Xmeshiro, S. A Caroline Islands script. Bureau of Am. Ethn. Bull. 173: 269-333? 1.960 (Anthrop, Pap, No. 60).

Sapper, K. Dr. Friedericis Sfidsee-Xeisen 1909. D. Kolonialbl. 21: 141-142, 1910.

Senfft. A. Bericht aoer seine R~u~dreisedilrch die Westkarolinen und Falau-Inseh. D. *Kolonialbl. 3.7: 281-204, 3.906.

Smith, A. G. Gamvoelhaelhi ishiljl Pleleeya (Guide to Woleai spelling). 1-51, Gum, 1951 (mimeographed 'by Depastmext of Education, BICOMTXR~)ACIS).

Tayama, R. Types of phosphate deposits and their distribxtion in the South Sea Islands. SangyE Kenliyu.sho 111; 0: 1-28, 1942 (translated, 1940, by Geological SIWV~~).

Tayama, R. Coral reefs in the South Seas. Hydrogr. Bull. 11: 1-292, 2 vols. of appendices, 1952.

U. S. Hydrographic Office Sailing directions for the Pacific Island I. 11. 0. 165: 1-632, 1938 (later ed. 165.4: 1952)