a thy Hare .s S S meer foe oad Mace ae = “ pA ‘ ae See earners roars ete a an toe Oa ‘ier i. ® >; Ws * il Ps | aa be ee A {| I fad < is l ab, rl + Noy io YL 4 i gil Y, ih 14 ¢ et gil st POL IN TS. AE” iN Bae Neng OR OME Mig, EN Os \ ree yy ISIE i i H “ye u¢ & eS | - a (| i y 3 | e , 3 * , - E mH ae e Sle x Ws yt ca j cp ita = pra a a | bh i r s ETF a = i of i ¥ ‘| ees | : =U 3 So iP es \ oi } F | l i i ; | i | a a /, iF i ays Ly i tt i Be STonee Bes; eM cel IMT ss Ge my % J aS : ger iis a 3 e e e e e . e CATALOGUE OF LAND AND LAGOON PLANTS AND ANIMALS Plants . * e e e ° ® e s e ® e s e ® e s e e e Animals e e 2 e s e e s e e e « 6 e e ° e @ e s . e BIBLIOGRA? colt 103 xd CHA, BIDAR GtooDi, 2 a £ ¥ = a * = FCA tad 0 2 PREFACE During the International Geophysical Year, Scripps Institution of Oceanography, a branch of the University of California, organized several research cruises in the Pacific, as part of the world-wide prograii. Several previous cruises had called attention to Clipperton Island, and during the Doldrums Expedition of the summer of 1958 it was decided to study the biogeography of this little-known island. A croup of 13 per- sons were left on the island by the Research Vessel Spencer F. Baird on August 7, 1958 and were taken aboard again on August 26, except for four who remained until September 25 to continue their studies of sharks. I had the great good fortune to be one of the group and to study plant life on the island. In addition I made observations on the land fauna and geology, and some collections of animals, soils and rocks. It will not be easy to thank adequately the many persons who have contributed to the success of my work on Ciipperton. In the first place of course I want to express my gratitude for this unique opportunity to Seripps Institution of Oceanography and to its Director, Dr. Roger Revelle. I owe much also to Dr. Carl Hubbs, from whom the invitation was received, to Mr. John A. Knauss, leader of the Doldrums Expedition, to the late Conrad Limbaugh, chief of the Clipperton field party, to all my companions in the field and to the Master and crew of the Baird. Our visit could not have taken place without the authorization of the French Government, and without the intervention of Dr. Jean Delacour, then Director of the Los Angeles County Museum and Professor Roger Heim, Director of the Muséum National d'Histoire Naturelle in Paris, who helped procure this author- ization. The French Embassy and the Office of the Naval Attaché in Wash- ington were also very helpful. For permitting me to join the expedition and encouraging ine to work up the material, I wish to thank my superiors in the U. S. Geological Survey and in the Pacific Science Board, National Academy of Sciences-- National Research Council. The Academy also provided very welcome help in the form of a grant frou the Joseph Henry Fund. I cannot name here all the persons who have contributed identifications, analyses, and suy- gestions and to whom I am deeply grateful. They will be mentioned in the course of the paper. In assembling bibliographic material and photographs I have benefited from the facilities of many individuals and several or- ganizations: the U. S. Navy's Naval History Division, Hydrographic Office and Office of Naval Research, the Service Historique de la Marine Na- tionale of France, the U. S. Weather Bureau, the U. S. National Museum, the U. S. National Archives, the Library of Congress, the California Academy of Sciences, and the Bibliothécue Nationale of France. Miss Evelyn L. Pruitt, Head of the Geography Branch, Office of Naval Research, gave me much help in searching for photographs and documents in the U. S. Navy files. Messrs. H. C. Allison, Ted Arnow, Willard Bas- com, A. I. Cooperman, A. S. Hambly, Lester F. Hubert, W. L. Klawe, John Knauss, H. S. Ladd, Conrad Limbaugh, W. &. Malone, H. &. Maude, C. S. Ramage, Waldo Schuitt and R. E. Snodgrass gave me unpudlished information, lent me photographs and documents, or read and criticized parts or all of the manuscript. Mr. V. A. Rossi gave me advice and help with the il- lustrations, and Dr. Gilbert Corwin examined the samples of volcanic rocks. es I wish to make special mention of Mr. Obermilller, Géologue en Chef de la France d'Outre-Mer, who worked on Clipperton a few months before I went there and gave me copies of his reports even before their publi- cation, as well as samples of his rock collections. Finally I wish to express my appreciation and gratitude to Dr. F. R. Fosberg who helped and encouraged me in every step of this work. The appearance of this paper has been delayed by various circumstances, as has that of a more extensive and profusely illustrated memoir on Clip- perton (Sachet, in press). 78001 7£1 0601 Poa HET 19>, eee re cae nef ‘was Ys 4849019 Sue) is) —e2n0d np esuy ~ * Lo, Od np v a Qy \ Y7prn0d np ajuiog® O|D9S eyDWixosddy T T 00 948)8W O00! 008 (uooBo}) 83904 pebsewqns (upe20) syooy epBjD ueesB-enjq yjIM peseAcd 40 @4pq ‘y90J pub jeADJB ‘pups }0J05 UoljDjeBan snoeopqsey pexiW DW DeOWOd) seysanw ebpes SiON) = il swjod ynuos05 GN3931 GNV1SI NOLYSdd!I19 40 NOILVLS9SA « ee png eduig so 3 il e601 D9 ° 2Uld DB] ep epg c:} Se Qo , © 781001 tS = is i : ne Ey va { a> ' ai, = 2) = INTRODUCTION Geographic location, general description Clipperton Island (Fig. 1), one of the few oceanic islands in the Eastern Pacific, excites the interest and curiosity of naturalists especially because it is the only coral island in that part of the ocean The nearest atoll, Pukapuka in the Eastern Tuamotus, lies 2300 nautical Miles to the southwest. Here are some other figures that will give an idea of the isolation of Clipperton, the coordinates of which are 10°18'N and 109°13'W (adjusted position for Astro 1957, a monument placed by the U. 8. Hydrographic Office on the northeast side: 10°18' 41"N, 109°12' 34"W). ‘The nearest land is the Mexican west coast, 600 nautical miles to the north-north-east. The nearest islands to the north are the Revillagigedo Islands of which Soccoro is 530 miles from Clipperton. The Galapagos lie 1300 miles to the southeast and Easter Island 2250 miles due south. In its general form, Clipperton is a low closed ring of coral lime- stone, but the island does not exactly qualify as an atoll as a small voleanic rock rises at the end of a short peninsula in the lagoon. This type of island has been called an "“almost-~atoll," but this category is not very natural and formations have been placed in it which may have nothing in common beyond the fact that they include both coral and volcan- ic features. For the purposes of this description it seems practical to consider the island as an atoll, the only one in the eastern third of the Pacific Ocean. In area, Clipperton Rock is very small, compared to the coral ring, and observations so far seem to indicate that it has little influence on the ecology of the island. No macroscopic plants grow on the Rock, except for some lichens. The coral ring is somewhat egg-shaped and symmetrical along a northwest-southeast axis. This axis falls a little north of the Rock, which is close to the southeast coast of the island. The ring is con- tinuous and encloses a brackish lagoon. Only a small minority of atolls are closed and few of these have a rim as narrow as Clipperton's, around such a comparatively large lagoon. The greatest dimension of the atoll, along the NW-SE axis is 4 km and the circumference of the atoll ring about 12 km. The emerged land strip is widest along the northwest coast, with a maximum width of about 400 m in the west corner. The average width is under 200 m, and in narrow places it is much less. In August 1958 the northern part of the northeast coast measured only 45 m from lagoon to ocean. That area of the land was also the lowest, rising only 0.65 m above estimated mean high tide level. Generally the ground slopes up from the ocean to the tops of the beaches or boulder ridges, and gently down again toward the lagoon so that the highest point of any given ocean- lagoon section is the crest of the outer beach or boulder ridge. This crest varies in altitude, reaching a maximum of about 4m. ‘The volcanic Clipperton Rock listed on charts as 29 m high, is visible from every part of the atoll. From a small boat in the lagoon the land rim is visible all around, but appears very low. In fact one has the impression of floating ee in an immense saucer full of water and with a very low rim. Oceanward, the land rim is surrounded by a reef-flat lying at about low tide level. On the ocean side, the outline of the atoll is practically featureless and smooth. The lagoon shores are more irregular, with several small peninsulas and bays. Half way along the northeast side the small tri- angular Green Point (Pointe Verte) juts into the lagoon (see map). On the southeast side, Clipperton Rock (Rocher Clipperton) rises at the end of a small peninsula, the Isthmus (Isthme); Thumb Point (Pointe du Pouce) extends northward from the Isthmus, separated from the Rock by Thumb Cove (Anse du Pouce). Between the Isthmus and the landstrip, the lagoon forms an arm called Rock Bay (Baie du Rocher). Just west of the Rock, another peninsula, The Hook (Le Crochet), stretches northwestward into the lagoon. The abandoned quonset village and the large coconut grove on the southwest side of the atoll are located along a small bay, Pincer Bay (Baie de la Pince) formed by two peninsulas, North and South Pincers (Pince Nord and Pince Sud). Other recognizable features are the 5 Egg Islands (Iles Egg) along the northwest side, and the U. S. Hydrographic monument (Astro 1957) and small group of coconut palms marking Naturalists' Camp (Camp des Naturalistes) near the East corner of the island. Historical sketch* Clipperton Island is named after an English buccaneer, John Clip- perton, who is reported to have seen it in 1705. Clipperton, who was travelling with the famous privateer and naturalist William Dampier, deserted and stole a Spanish prize bark in which he crossed the Pacific, a remarkable feat. There is no account of this voyage during which the island is supposed to have received its name (Burney 1816), and in the description of Clipperton's second voyage (Betagh 1728) no mention is made of the earlier discovery, but the name was indicated on maps about 1730 or 1735 (Moll). Various authors (Toniolo 1919, Mexico /1911/) have suggested that Spanish navigators may have seen the island earlier, in the 16th or 17th centuries, or even that Magellan may have discovered it in 1521 (Nunn 1934). The island would then be identical with Medanos and San Pablo (Magellan). Historically, such considerations are of great interest but as far as our scientific knowledge of the island is concerned they are immaterial, as it is quite unlikely that ancient descriptions will come to light. Our knowledge of the island, and the extensive literature concerning it, then begin in 1711. On April 3, Good Friday, two small French vessels, the Princesse and the Découverte, who had left Brest together in 1708, met with an unknown island which was named Ile de la Passion. The Captain of the Découverte, Michel du Bocage, and a passenger in the Princesse, Mr. de Prudhomme described the new island, the one in his log and the other * For a more detailed treatment see Sachet 1960. -5 - in his diary. Their discovery was first mentioned in print in 1725 (La Barbinais Le Gentil), while their accounts were published in extenso in a French report (France 1912). The original documents are kept in in the French National Archives. Mr. du Bocage described Clipperton as "A large Rock, cragged and jagged, at the south point of a very flat island..." The northeast side, “sandy with some brush and a dried-up tree on the north-east point, was but a very narrow tongue of land. The center of the island was a large lake reaching from one side to the other. The west side appeared to have some low brush with soil and some rock but very low, although a little higher than the east side." Mr. de Prudhomme also described the very low sandy island, without any trace of inhabitants, unwooded excerpt for some very low bushes, and with some dead trees on the sea shore "as if they had been thrown up by the currents." These brief accounts are remarkable in that they describe Clipperton much as it appears today from aboard ship, except for the recently added coconut trees and various traces of human activity. There can be no doubt that Ile de la Passion is the island now called Clipperton. The coordinates given in 1711 were: Du Bocage 10°28 'N, 263°50! (11303'W Greenwich ) De Prudhomme 10°18-19'N, 268°11' (108° 27'W) (From the latest observations, the coordinates are 10°18'N, 109°13'W). Over a hundred years later (1832), an American sea-captain, Morrell, described the island as he saw it in August 1825: "It is low all around near the water, but a high rock rises in the centre, which may be seen at the distance of six leagues..." As far as known, Morrell and his men were the first to land on Clipperton. Except for a small sketch map, poorly oriented, which Mr. de Prudhomme mentions and which is kept with his diary and revroduced in the French re- port (1912), Clipperton Island was first mapped by Sir Edward Belcher and the map published by the British Admiralty in 1849. In May 1839, when Belcher visited it, the island was (Belcher 1843): "a very dangerous low lagoon island, Peel he of trees, with a high rock on its southern edge, which may be mistaken for a sail ..»/The belt of land/ literally consti-~ tutes two islands, formed by its two openings ..." (See fig.4,facing p.69). A few years later, a French ship owner, Mr. Lockhart, arranged to take possession of the unclaimed island for France and to exploit its phosphate deposits. Thus it came about that in 1858, one of Mr. Lock- hart's merchant vessels, L'Amiral, Captain Detaille, Master, arrived at Clipperton. Lt. Victor Le Coat de Kerveguen, who had received a special commission for the purpose, took possession of the island for the Second Empire on November 17, 1858. The Amiral then proceeded to Honolulu wheres the necessary papers were filed to announce to the world that the island was now a French possession. Le Coat de Kerveguen took notes on what was observed on Clipperton and his manuscripts, sketches and maps are reproduced = ole in the French report (1912). In 1858 the coral rim was closed, the lagoon salty, there was no vegetation, but great numbers of sea birds. Some soil samples were collected, but exploitation of the phosphate deposits did not seem worth while and was not then undertaken. In August 1861 (Pease 1868) a young American, Lt. Griswold, who later lost his life at Antietam, visited Clipperton Island and found it uninhabited, covered with birds but devoid of vegetation. The lagoon was closed, its water fresh and full of a water plant which he collected. This first botanical specimen from Clipperton was sent to the California Academy of Sciences in San Francisco and lost in the 1906 fire. Just who discovered phosphate deposits on Clipperton Island is obscure. They were known to Lockhart in 1857, and Griswold was on a phosphate explor ation trip. In 1892 Frederic W. Permien made several trips to Clipperton to survey the deposits and in 1893 the Oceanic Phosphate Company of San Francisco sent two men, Jensen and Hall, to survey the island and begin exploitation (Anon. 1893). The observations of Jensen, and his excellent sketch-map aze reproduced by Agassiz (1894, pp. 174-175). Jensen had collected a piece of the volcanic rock, which was identified as a trachyte. At the same time, A. Churchill Fisher made observations on tides (U. S. National Archives). W. C. Erratt, of the Schooner Anna of San Francisco, surveyed the phosphatic deposits probably in 1897, and his map is available in the Archives. Also in 1897, P. J. Hennig (Anon 1897) prepared a map, which he forwarded to the U. S. Hydrographic Office. It was promptly published as an H. 0. Chart (no. 1680), replacing the charts based on Bel- cher's sketch, and remaining in use by all navigators until the French survey of 1935. Exploitation of the Clipperton phosphate proved difficult and not very profitable, and several companies Were to attempt it one after the other before the island was abandoned during World War I. When the French war ship Duguay-Trouin arrived at Clipperton in Nov. 1897, a small camp was discovered with three employees of the Oceanic Phosphate Co., who hoist- ed the U. S. flag. France protested, the United States announced that they had no claim on the island and Mexico joined the excitement. The Duguay- Trouin was hardly out of sight when a Mexican gunboat, the Democrata, arrived (Dec. 13, 1897). Here started a diplomatic conflict which was to last until 1931, when King Victor-Emmanuel III of Italy awarded the island to France, and which has become a text-book case for students of diplomacy and International Lew. France and Mexico had agreed to submit the case to the monarch's arbitration in 1909, and both countries publish- ed historical reports (Mexico /1911/, France 1912) to present their cases. They are valuable documents, the French one especially, as it includes all the texts describing the island to date, some of them unpublished until then. During this period, Mexico continued to lease the phosphate deposits of the island to various companies including the Pacific Islands Company Ltd. of John Arundel. Also during that time, several naturalists visited the island and made geological and zoological collections, among them John Arundel in 1897 (Wharton 1898, Garman 1899), Snodgrass and Heller in 1898, R. H. Beck in 1901 (Beck 1907), and the California Academy of -7- Sciences expedition in 1905 (Slevin 1931). Casual visitors added — specimens, especially shells, to the collections of Clipperton animals. For instance a group of shells in the U. S. National Museum was received in 1897 from Mr. Arnheim, a ship chandler who had obtained them from sailors. In 1906 a light was built on top of the Rock, but during World War I, the island wasabandoned in a dramatic chapter of its history, and was forgotten. There are quite a few accounts of it for the period 1893-1906, but hardly any information between 1917 and 1935. According to a letter in the U. S. National Archives, several men from the Schooner Ethel M. Sterling from San Pedro landed on Clipperton on January 5, 1929, after several days of bad weather kept them on board ship off the island, and they collected samples of guano and lagoon water. The research vessel Velero III (Fraser 1943) stopped at Clipperton on January 6, 1934, but landing was impossible, although some biological specimens were dredged a short distance off shore. In January 1935, the French training ship Jeanne d'Are after several visits when landing was impossible, came back to Clipperton, and a group of officers and midshipmen succeeded in making a landing on January 26. They drew a map, wrote descriptions of the is- land, and collected some plant and rock samples (Lacroix 1939, Gauthier 1949). The ship's seaplane took some photographs. A bronze plaque sealed on the east face of the Rock commemorated the visit and established French ownership. The next well-known visit was that of President F. D. Roosevelt in July 1938, during a cruise on the USS Houston. Dr. Waldo Schmitt of the U. S. National Museum, who had been invited to travel with the President as expedition naturalist, landed on Clipperton for a few hours and made some very valuable collections of plants and animals which were described in a series of papers (Smithsonian Misc. Coll., 1939-1942). During the second World War, Clipperton is said to have been visited by Japanese submarines; in 1943 and 1944, the U. S. Navy made several reconnaissances of the island and the famous Australian pilot Captain Sir P. G. Taylor landed his seaplane in the lagoon. Such visits resulted in valuable descriptions (Byrd 1943, Taylor 1948) and photographs. In Dec. 1944, a small U. S. Weather Station was established. The ship (LST 563) bringing most of the material for its construction struck the reef, and its great rusted carcass, battered and much dismantled was still a con- Spicuous landmark near the landing point on the northeast side in August 1958, and served as a breakwater for small boats effecting a landing. Rusting landing craft, fuel tanks and ammunition were scattered nearby across the land strip. The weather station lasted until October 1945, and its ruins are still recognizable as an abandoned quonset village in the southwest coconut grove, the roads marked by tracked vehicles during that period can still be followed most of the way around the island. Since the end of World War II, French Navy ships have been visiting the island regularly (Gova 1952, Bourgau 1954) and when landing turned out to be possible, markers have been attached to the base of the Rock. In 1958, there were 4 such commemorative tablets. During the fifties, scien- tific groups from various institutions stopped at Clipperton and made Za. observations or collections: the U. 5. Navy Electronics Laboratory in May 1952 (Hertlein and Emerson 1953), Scripps Institution of Oceanography in Dec. 1954, and Oct. 1956 (Dawson 1957, Hertlein and Emerson 1957), and U. S. Hydrographic Office, Nov. 1957. This last group was making astron- omical and geophysical observations, and determining the exact position of the island. A French Navy officer and Mr. A. G. ObermUller, chief geol- ogist of the France d'Outre-Mer accompanied the American geophysicists. The geological collections have been described in an important report (Obermilller 1959). Another valuable result of this expedition is a col- lection of photographs taken from a helicopter. In May 1958, Mr. W. L. Klawe, of the Inter-American Tropical Tuna Commission visited Clipperton for a few hours and collected some plants. Finally in August and Septem- ber 1958, the group of naturalists from Scripps Institution of Oceanography camped on the island and made observations as complete as possible on the flora, fauna, weather and otheraspects. The group included four biologists using skin-diving equipment to study the marine fauna, three icthyologists, two entomologists, an ornithologist, two radio operators, and myself as botanist. The marine biologists working on the reef and in the surround- ing ocean, collected marine animals and algae and made ecological obser- vations and I concentrated my efforts on the land flora and terrestrial and lagoon habitats. However, to give a complete view of the atoll, I shall include brief descriptions of certain marine aspects from my limited observations completed from information communicated to me by my compan- ions and extracted from the literature. A list of the recorded land flora and fauna, the latter largely compiled from the literature, will also be included, in-so-far as necessary to give an account of the land ecology. Extensive accounts of the zoology are being prepared by the zoologists of the expedition, and other specialists. -9 - WEATHER AND CLIMATE The weather and climate of Clipperton Island are very poorly known. The only weather records, outside of brief mentions by visitors, were collected from January to October 1945 when a U. S. Navy station operated on the island, and during August and September 1958 when Messrs. Limbaugh and Chess of the Scripps group made daily observations. The Navy records have never been used and are not available; the others will be utilized below (and see table 1). The U. S. Hydrographic Office publishes climate and weather data for ocean areas from information transmitted by ships. However the ocean area where Clipperton Island lies is not included: The Weather summary for Central America (H.O. 531, U. S. Hydrographic Office 1948) has no information for areas west of 100° W long. so the 5-degree square for Clipperton is not discussed. Summaries of Pacific Ocean weather often do not extend far enough to the east to include the region of Clipperton. One reason for this state of affairs is probably the small amount of information available in an area where there are no recording stations and few ships records. According to the World Meteorological Organization (Anon. 1957) this is one of the poorly known ocean areas of the world, from which more reports are much needed. Information on Clipperton must be gained from general sources such as the Weather Bureau's Atlas of climatic charts of the ocean (McDonald 1938) and the U. S. Navy Atlas (1956). The climate of Clipperton Island is an oceanic tropical one, with little variation in temperature but with seasonal rainfall and storms. The seasonal variation is correlated with tropical cyclone activity which reaches a maximum in the eastern North Pacific in August, September and October. In these months the island frequently experiences winds from a southwest quarter, probably generated by tropical cyclones to the north. Temperature According to the most recent compilation (U. S. Navy 1956) the mean monthly air temperature is never below 80°F (26.7°9C) and only in June is it equal to or above 820F (27.8°C). From the same source, the mean air-sea temperature difference for 3-month periods is of the order of -1°F or at most -1° to -3° (air cooler than sea). In the period August - September 1958, the lowest recorded air temperature was 75°F (23.9°C) at 1300 PDT (Pacific Daylight Time, which is the correct time for Clipperton Island) taken during or just after a rain storm. The highest recorded temperature was 87°F (30.6°C) also at 1300. On sunny days it Was quite hot in the sun, but never unbearably so. Night temperatures were not recorded but seemed only slightly below day time temperatures. SO Atmospheric pressure Because Clipperton lies close to the equator and its belt of low pressures, the atmospheric pressure there is always low, with little yearly variation and with probably regular and slight daily variation in calm undisturbed weather. At Colon, Panama (U. S. Hydrographic Office 1948) the yearly variation of mean monthly pressure is only of 2 mb, the highest mean in February (1011.18 mb) and the lowest in June (1009.14 mb). In the same area of Panama, the total daily range at Cristobal averages 2.9 mb for the year, being about 3 mb for October to March and 2 mb for June to August. The highest pressures there occur at 1000 and 2200 local time and the lowest at about O400 and 1600. From a large sampling of ships' reports, the maxima over the ocean at 10° lat. occur at the same times, and the diurnal range is 2.8 mb. From the short record available (August - september 1958) the daily variation at Clipperton seems to fit this pattern, with a usual daily range of the order of 2 mb. During the period of record, the total range of recorded pressures was 8 mb (1002 to 1010). The low values coincided with bad weather and probable cyclone activity in the area. Since the records are only for August and September nothing can be said of any yearly pattern of atmospheric pressure variations. Winds Surface winds at Clipperton Island vary greatly during the year, In the winter the dominant winds are the northeast trades, although they are far from exhibiting the type of constancy experienced on islands farther west. In the summer the winds change considerably and from August to October they blow principally from the SW quarter. During that time, as tropical cyclones dominate in the area, winds are extremely var- lable and their direction can change rapidly during the day. Table 2 is from MacDonald's atlas. Wind roses for the 5-degree squares are included in the Hydrographic Office's Pilot charts of the North Pacific Ocean and those for the square of Clipperton are reproduced from recent monthly charts in table 4. Nothing more is Inown of the wind regime for the island beyond this generalized information except for the August - September 1958 data and the observations of Taylor in September - October 1944 (P. G. Taylor, 1948. In 1958, the recorded observations of wind direction (see table 1) indicate a predominance of winds from the southwest quarter, especially SSW to SW in August and SW to SWS in September. These obser- vations fit well with the generalized data of McDonald's atlas (see table 2). Recorded speeds were up to 26 knots, but the majority of observations in August indicated speeds of 10 Imots or less and in September a majority of observetions below 20 knots. In August, at least, the strongest winds were not recorded. There is only one record of calm and indeed the al- most constant wind, whatever its direction, made living on the island very comfortable. Rains almost always came from the southwest; from the northeast land strip one could see the dark rain clouds engulf the coconut grove E. 4 og i Tait se di Ie Wed, ‘asewuley 10ossajoig Aq peszedaid yIeUD = *WI0}}0q ay} 3e serey 429M GUO Spua pue dsUTT 34} Jo do} dy} 3e SuUTISaq aoUeNnbas ayy “SuoT (UH OOFZ) SeTIW OOST ATeyeWIxOIdde st sul] STYUL ‘pueTST ay} ye SpuTM PaAZesgo aie suUT] STU} UO SMOLIB OUT ‘“SUTT YSTe1}S 94} SuoTe oI] UlSy} 0} SATIETII PULTST ay} JO SUOTITSOd aATSSadoNS 9Y} pue MNM 10 MN PieMmoy ATTeL38Ues saow (auoToOAOTWUe YW ‘auoToOAD D) suteayshs ayy, ‘sjouy ut petjeqe] aue (SeuT] UaYOIg) SYde,OST 94, “MOTJ-PUTM JO UOT}DAIIpP ay} JUeSerdaz SaAaNd SUTI-pI[OS au, ‘puUeTST uojtaddt[D ye suotyearasqo (gP76I[) S,1OTAC], But}}1J Qouaenbas y21eyo Yoe}OST/PUT[WIeAI4S pajonajsuooday z ‘stg \ PQ AL aN aad § a, ete on the opposite side of the atoll, which promptly disappeared from view, cross the lagoon and eventually reach the northeast side, drenching it with a downpour of incredible suddenness and violence. In October 1956 (Limbaugh, personal commmication) the winds were generally from the southwest, with velocities up to 13 knots. During his first short stay on Clipperton (from September 9th, 191+) Taylor seems to have encountered reasonably good weather, sunny and clear much of the time, and he gives little information on it in his book (1948). During the second visit (September 21 - October 14) the weather was much worse, culminating in a hurricane (see p. 19). Strong winds from the southwest were common, with squalls up to 35 or 45 knots (estimated) and frequent and rapid changes of surface wind directions were noted. After spending some time on the island, Taylor summarized the gener- al pattern of winds and weather as follows (p. 194): "From observations from the time we had been at the island, it seemed that there was a fairly regular weather cycle, with sometimes one period of the cycle more prominent than others. "We had arrived the second time in the bad sector, when the wind was in the south-west and at a time when this phase was exaggerated, with stormy conditions. "As the wind swig into north-west the weather had improved, and by the time it had come round to north the sky was clean with scattered cumulus, and the surface wind was seldom more than fifteen knots. Fine weather with light winds and little cloud then prevailed for several days as the breeze worked round east to south and back for south-west, when a high overcast would come over and the underlying cumulus would build up to cumulo nimbus with lightning and storms and the air was uneasy and wild till the wind shifted again toward north. The weather went round this cycle in about a week, and always there was rain from the south-west or west, and sometimes a shower from a local build-up of cumulus in the fine periods.” This sequence fits well the storm tracks commonly observed in the Clipperton region, and has been represented in a streamline/isotach reconstruction (fig. 2) prepared by Professor Ramage (personal communi- cation 1960): "The cycle starts with a tropical cyclone (C), moving NW or WNW to the north of the station /Clipperton/,. Winds veer to W and NW and the weather improves as the eastern portion of a small high pressure cell (A) follows the storm circulation across the station. Winds de- crease further but weather stays fair as the divergent portion of a col preceding the next cyclone moves over. The cycle is complete when strongly convergent SW winds associated with the next cyclone set in." This recon- struction also compares well with the sequences recognized from the 1958 data (p. 16). =nieh = Tropical storms and hurricanes While not as numerous and destructive as they are in the Caribbean area or the China Sea, tropical storms and hurricanes are common in the southeast North Pacific, with an average of at least 6 storms every year, two of which reach hurricane strength. Hurd (1929 p. 45, 1948) assembles the cyclonic storms of this region into 4 classes: 1. The coastwise storms, that run parallel to the Central American and Mexican coast, 2. those that strike perpendicularly upon the Mexican coast, 3. cyclones of the Revillagigedo Islands and 4. cyclones west of the 125th meridian. The last need not concern us here. The storms of the first class origin- ate somewhere along the American coast, often in the Gulf of Tehuantepec or south of it, or sometimes perhaps as far south as the Gulf of Darien (Ives 1952). They travel usually northwestward in a course parallel to the coast; some swing inland over the Mexican west coast, some travel far into the Gulf of California; a few cross over to the Caribbean area, and conversely, some storms originating in the Caribbean travel to the Pacific side. The storms of this class are far to the east of Clipperton and probably never affect it. The storms of the 2nd class occur farther west and travel in a north- ward direction. Those of class 3 travel in a west-north-west direction and are mostly observed north of 15°N. Some of the storms of these 2 classes may form near Clipperton, or pass by it. There are also storms in this area which do not readily fit in the 4 classes, such as some which travel in a westerly course or sometimes even somewhat south of west. The occurrence of tropical storms in the southeast North Pacific is seasonal, and the monthly distribution is given by Visher (1925) as follows: Dec. to Feb. 1% each March to May 0 June 5% July 11% Aug. 15% Sept. Qe% Oct. aha, Nov. 5% Because Clipperton is removed from shipping lanes and has only been inhabited occasionally and for short periods of time, historical records of storms and hurricanes are poor. However, some of the storms which are reported west of 110°W and travelling in a northwest direction may have originated near the island or passed it, either as storms oy rarely, as full-size hurricanes. A tropical storm, it must be remembered, need not score a direct hit, or even have reached its full development, to affect an atoll lying only a few meters above sea level. Even storms travelling several hundred kilometers away may generate storm waves high enough to batter the island and even to flood the lower areas of the land strip. Ae The older records of storms in the southeastern North Pacific are listed by Redfield (1856), the Deutsche Seewarte (1895), Visher (1922, 1925) and Hurd (1929, 1948). In recent years, tracks of storms have been plotted by the Weather Bureau in the National Summary of Climato- logical Data and the Mariners Weather Log, and by the Hydrographic Office on the Pilot Charts. From these records the list on the following pages has been extracted. Some of the storms may have passed Clipperton Island close enough to affect it, or may have formed near it. Those marked with a * were reported very near the island, or experienced at it. SeTseses Jieqoy n°446e% U4 uo [Boytedet 197 eroyy4 Loyouv 40 Tossea v fq poosuetiedxs oxen sttunbs AT Zeqysonygnos su07149, BoyyoO10g OPTLL Spy pue *viog FFOUST SUS Teqrzuce S ‘m ‘Nn Sq sputm f4aeTTey UOTALUIOS UT utogs fSpTag Sucaves MS ‘mh SMM Sm SGN sputm fcTaeyy SpnazzTey (ZeAoge se Wto4s sacs) SITaug pue §sueuud UTSSe mS pue M °N SH “9S “MS SpuTM :TOZ0dy S ‘m ‘N Sputm fou009 sojzou pue *sutyaoced drys Sc6T TOUSTA 876T PInHy 6cé6T PINy 6c6T Pany u u ul u ep,IenS9g °C PLOEF POU o4tunoss *q ub uN it PLOT IPOY DOAN sCCoLCL 1970601 Ret pent oLOT o8TT of TT Qe0T ofl o0tt TE, 9TT BeuL oSTI 10S 59TT oLTT [Ay OT o9TT 167oLT 1L€00T ovT out oSTt 10£,€T 10Zo7T NiSS,ST pojdo° oil SovTtd €z *ony Ti=8 °9des Té-L2 “any T *490 -6% *ydos 2 any Te Arne € Arne ce Arne o¢ oun 6-8 “eny Atng 6T-9T “Tor F-E °900 § *sny ec=Cc Suny = 15) = ee MNM U0UQ Mm BuTAo S§yaeg Teavay M SUTAOT ae ‘oO TTT ‘oL°EL FE poqeutyse soqueo fooyedery MN SuTAOU £52°60T "ae EL qe pozreyo r9eqU0D fWeISYSTy UCTTeney N Sutaou §USzT4TO UCTTeney *d ‘mozyeq ess eusoTamy oedaeyT AoA pues osuaqut freA y AUTO Tsussseg SdoTT TED duvo uvotxeyy pofoaysep uz0qg Sojou puc BS6L Yous $307 Isy}ve *1ey] pus il u AtewrIng TCyN 87ET tOTABI, mM 876T pany 876T Painy pue Gzél LOuSTA SC6L TOUSTA ST6T TTCQUTY 7E6T STIcopy Oo dtlOG of °60T oe OL ollt ore aioe Lelia! 09°71 uoqzeddt{t9 oULt ocL 9901 oct 3 62o ETT 1VEo9T QALt oST WZ7oETT 1 CT 9 9T Mo8TT NoVT 107,60T 1070ST uoqzeddt{T9 peyioded SdeTq 9e~7T Ane 6THLT Ae OT-2 sume g-¢ SunL tL °4°0 TI=9 °920 Ve-etz *4cesg 62-42 °3°0 9T=ST °9°0 QT-ET °9°0 OL=6 *4deag Te Arne 6-7 *y4das gount *LS61% 6G 6T *GS6T *SS6T “7761 ‘O76L ‘OV6T *O€6L ‘8261 *eZ6L "C261 *t26l *STO6T “ST6L* Syeq *U%CLE 944 UO IND COST +e M S°Oc~ ‘N 9°ZT 3 S90UN G7 GN poqtodez IsTUNOL NVIIVMVH tuys eyy Wey pezBooT 4SaTy sem yoTYM IOVS oy} ST STYyZ ATUTBZIE0 YSOUTY Sy, *u0TZe4S 943 FO yQaou ysnf ButTssed roqueo eyy SMNIM PICANOF PeAou tit0ys TeoTdosy to uoTsseidep yy 3 (UOTZCOTIMMLIOS Teuosied) eseuey Losseyorg Aq pejyonaysuooced axel seouenbes suTMOT TOS ey, °ezeysi =OSTO IOJeVT PeATOsqo susoys yourqyovg dyrey oste (1 Stqe4 ees) *ydeg-"sny UT PpaqgoeT[OO Bqep SUT ‘pues fq pereaco useq ATsnotasad pey 4yvuy pesodxe yoor -Youveg pus ‘reuz09 ysenyyzAOU ey} 4 peqyTsodeper ' puv peaouer Teasers yonu ‘4nd ueeq pey ezoys ee, (Cyup peysttqndun ‘ysnequry) pue{st eyy Jo septs 1. 4Sonyqgnos ey4 UO PpoATesqo oq PTNOD syzooTjo sqt ‘LT °90eg ug “aequeqgdeg Jo eT ppTu ey UT HI09S @ ZuTpNTOuT *7eyyvon peq sos sen oTeyy Seqdeg=*sny UT puerTSsT ey. FO ASAINS oYyy SUTIMG 16 *SOABA TIJOVS S9T Jo qooyye oy} petytsuequt Aoyy 4eYyy YONS e190M SOpty pue S4UeTIMS USA sUTZ BV Ye PeTmMoo0 savy AGU pue fesamoo jo fa90uTqaN{STP Teoo, LO JOUTT ATOATYLTOI © U9Eq SAVY ACU YT “*PeTITAUEPT 9q qouued WI0%S Styq 3eyuy SuTqQUTOddestp so0yye4 ST 41 “SG6T AP pus LOET “AON UeENgeq PULTST ou? YIU aAey 4Snu YOTuA yey 0% puodsaz10004 suees (ecé6T ‘Soy teyyeoy *1eWy) PAooer uo UsO4S ONees ei-Q/.°dd seg *~T JO opts 4ysvoyyaoU IOAO PTCT Yoous ToAvas pus pakorzsop uotTzeqzose, uozreddtT 9 uozaeddtt9 pe, 10dsz sdeTg g ‘ony (®)n 8S61Lx 8S6T=156Tx oqTed *sqouy OL SPUTA WNUTXGU §pIengsenyZcou SuTAOU USSTy ey SuvoTIINyY u oLOT oll vy Ator *sqjouy O09 1eJUSeD AeeU SPUTM SYYTT T9O6L *9deg ‘*soT oy. uo §pzengsem SuTAOU euuvor Utoys TeoTtdouy TOYZ * IVY Mo€TT No9T OT AIne ‘T96T *pueTs~ uoqzeddt{9 pez -O9Fju oAvy 04 ATOYTT suses *TO6T Youu SsoT TsyQvey SIOUTIG UT O961 JOJ poysoder stitoys ey Fo suoN uozzeddtt9 O96T *pucTs~t uojszeddt{9 poyoosse aaey 07 ATEYTT suees *QO6T Yours SST JoyQvey SIOUTIVY UT 6G66[ TOF poysodesr stlz0ys oy} Jo ouoy uo4zoddtt 9 6S6T ' LOS6T YoreW TOF Boy reyyvey Ls SIOUTIVY UL Ppopslodet ote peuoTyUs suzo4qs ’ oud/ °29zUEd 1104S 9Yy4 Ieeu pedoToAep pry S}OUN OOLT 1OAO JO sputMn teqyyseertey4 AT ZA0Ys Jo four, styy Ag °4y9 O44 UO JIND OOZI 32 SyOUy OY AS OF PEsvetoUT (MCTT ‘N6L) BPT IAC eooy JO PUGTST 944 UO SPUTA SYM peqvoo0T 4SITJ SUTOQ MN WOU. PUe N GAOT 04 PoNUTAUOD 47 “PLE oy} UO UOTALAS eYy FO N S44 OF SUTAOU ATT CULTS vote uojzszeddt{9 oyy ut AFTsueqUT ATMoTS pue dotToasp of peazvedde uotsseidep y, € *ydeg ~ Le *any (P) _ “MN 849 04 ez qou petyTsuequt ATayssod uy. mN 0% WS WoT uoTAeT4Ss eYyy ssozoe pessed ms04s [Teotdos4 [Les ev JO soqueD oUT, G@-LT *ony (2) *u0T9e4s ey4 JO yAsou ysnf es.moo fTIZE4SeN B Tr *sny (c) uo peAol Usoys [Tvotdo7z, Io uotsseadep [Tous yy uoqaeddtt9 Bs oLx Seenied Opud Cremer > Ba qa0der drys Sooanog Po,dccod ooUTd Seyou puc o/tl occ 9F *sqouy QQ SpuTm ‘prengsomyytou SuTAoU vpTIg us0qs Teotdogy, iu oOIT oll g=9 °90eg STO6T BOT *SqOUY OG-O€ SpuTa ‘TuowN wtoys TeoTdouy, TOYyZLOM IVE GOST STT o8l Gey *sny *T96T *sqouy GG ToyUed Teou spuTH Spreiqsen sutaou §(€¢z *d oes) ouTTopey uasoqs Teotdogy, Ph o60T oat oe ATue *S4ZOUY 09-05 AGEU SPUTA UNUTXe *pICAQSOMYyZAOU-94S0MN BSUTAOW eZTtT h o8Lt 002 8T=ST Atne 04 *sqouy OZ, 1e4UeD Oy AvoU SpuTM SezZTYT SUBOTIINY a ot ail +r Atne * 104U90 ey} Teau SpuTmM 4youy Og *4SOM SUTAOU UeSTYZey i ecEL 091 ot Arne TO6T °3.des #307 useTY4 ey Toyysom STeyy o60L oOL ¢t Arne *TO6T Sojou puv *“surjaoded drys Soounosg Pojaoded SduTg o7ed gus The following is condensed from Taylor's book (1948, pp. 224- | 254), the only eye-witness account of a Clipperton Island hurricane. October 12, 1944. Soon after dawn, fresh breeze from the north, sky dark, heavily overcast, nimbus hiding the tops of cumulus, and fast moving clouds below. Altimeter of aircraft set at zero read 29.95 (1010.8 mb), low on the scale of earlier daily readings, but no lower than had already been recorded. At breakfast time, the fresh puffs on the lagoon had changed to squalls, the wind swinging around as its force increased soon reaching 50 knots. Rain came in blinding sheets, drenching the camp. The squall passed, but the wind hardened into a steady blow, and within an hour of coming ashore for breakfast, little remained of the camp. (This camp was on the west lagoon shore, just north of North Pincer). Suddenly the wind eased to 30 kmots, but soon after swung into east and became worse. Half an hour later there was another slight easing of the wind, but soon it swung more towards south and became more violent. "About two hours after the first squalls the wind swung nearer to south... The Rock was blotted out and the lagoon was sweeping towards us in a wall of water like spray from many hoses, reaching from shore to shore and leaving no definition between sky and lagoon. It was a fantastic sight..." The wind was still blowing with hurricane force, estimated as between 80 and 100 knots. A little before noon it swung to south and the lagoon waves abated somewhat, but danger from the ocean became greater. In the coconut grove: "the heads of the palms were streaming back in windblown fronds like the hair of a girl in the wind. The taller palms were bent like tight-strung bows, and held till it seemed that their spines must break. Their fronds were awash in the blast of air... And now, instead of the horizon dark with cloud against the flat rim of the land, there was a cold white stream of breaking surf, visible above the land, and roaring madly at the island... Already only the reef was stopping the rollers, which ... were higher than the land.’ If the hurricane were to swing into southwest the ocean would flood the strip of land where the camp was. Trying to reach the coconut grove only 200 yards away, two men went first oceanward to seek some shelter on the ocean side of the beach ridge, but the sea was already far up the beach, and kept rising. "It was roaring by now not two feet below thetop of the bank. The surge was sending little rivers trickling through the stones over the top... Now there were no rollers breaking on the reef. The whole ocean was tearing by in a roaring flood of water, clawing at the island." The pigs and some birds had taken shelter in the coconut grove. About 3 p.m., there was a definite lightening in the sky and a Slight easing of the wind, which definitely slacked up a few minutes later. Back in the aircraft, the altimeter showed steadily rising pres- sure. By 5 pem., the wind had dropped to a hard blow of about 40 knots, and by night, eased to about 30 knots. The next day was sunny, with some showers and a light breeze from the south. Two Catalina flying-boats were anchored in the lagoon during the storm and barely survived it. On the lth, with fine weather and a light southerly wind, Captain Taylor took off in his flying-boat the Frigate Bird for the Marquesas and Bora Bora. Rainfall The Pacific coast of Central America and the ocean area between it and Clipperton Island have a decidedly seasonal pattern of rainfall, with the months from December or January to April or May drier than the summer and fall. This pattern apparently is still valid westward as far as Clipperton Island. The Sailing Directions (U. S. Hydrographic Office, 1951) indicate that there is a "dry" season in December to May, but just how dry is not mow. This information is probably based, in part at least, on observations furnished by the guano workers who lived on the island during the last years of the 19th and the first years of the 20th centuries, as it was already included in the 1902 edition of the Sailing Directions. For instance, P. J. Hennig kept notes on the weather during his stay as “keeper” of Clipperton and is said to have forwarded them to the Hydrographic Office with his map (cf. pe 6). I could not find a record of his data, but a newspaper account (Anon. 1897) quotes excerpts from them: "Mr. Hennig, who is a man of ability as well as patriotism, kept an accurate log of the island from October 1, 1896 to August 5, 1897. He has presented the log to the hydrographic office, and it is the first valuable record of the Clipperton conditions. "October had almost daily rains, with southerly and southwesterly winds. There was about as much rain in November, several heavy thunder storms and a few hard squalls. December was sultry, with many rainy days. Little rain fell in January, and the weather was hot. The word ‘pleasant' occurs frequently in the February records. ‘Severe thunder storms were again noted in March, April and May. The record of the Kin- korea disaster is as follows: "Friday, April 30th--Wind N. N. S. Moderate breeze with squalls in first part; then gentle breeze with squalls and light rein, but mostly fair, hot, sultry weather. Heavy surf. At 8 a.m., stranger in W. S. W. in sight. "Saturday, May lst--Wind N. E. by E. Moderate weather; heavy surf. Stranger coming around east side of island, flying British colors. Proved to be ship Kinkora, Belfast, water-logged, and bad steering. . Beached at 5 o'clock. All hands saved. "Easterly and northerly winds prevailed from the beginning of Novem- ber until June; then the south and southwest winds returned. "June and July were hot months, with occasional rains and squalls." In addition, the manuscript diary of John Arundel for July and August 1897 (which I have seen by courtesy of Mr. H. E. Maude, and which is quoted here with the permission of Mrs. Sydney Aris) includes this observation of one of the workmen, who "says it rains nearly every day - showers - though you may have 2 or 3 days without - but the winter months are as a rule the driest - though he has never known a month with- cue -e. In the winter dry season - 3 or 4 mos. you have North East 8. es ee The only recent observation bearing on this "dry" season is the fact that the vegetation on Clipperton Island in August 1958, appeared to be recovering from a dry spell: many plants had dead twigs together with young new growth. In May 1958 (Klawe, personal communication) the dry condition of the plants was striking. Perhaps even more significant are the thick "trunks" (see p. 78 and Sachet 1962) of the beach morning- glory (Ipomoea pes-caprae) which suggest dying back of the vines during an unfavorable season. However, it must be kept in mind that with a very porous substratum such as is found on coral islands generally, effective drought may occur even with a rather high total precipitation if the latter is irregular. Certainly the dry season at Clipperton cannot compare with the intense droughts occurring on some desert atolls of the central equa- torial Pacific where occasionally there is no rain at all for months. The drier season is indicated in the world atlases (McDonald, U.S. Navy) by a smaller number of observations reporting rainfall (table 3). While there are still many observations reporting rain during the winter and spring, they may correspond to light showers very different from the down- pours of the summer months. The maps in the same atlases, however, indicate that an area of very abundant rainfall lies to the west of Clipperton Island. In the Navy atlas, the January isogram for 40% of observations reporting rain forms a sort of ellipse to the southwest of Clipperton, between latitudes 5° and 9°N and longitudes 114° and 127°W. ‘The size and position of this ellipse vary somewhat during the year and while it is formed by the 40% isogram from November to February, it is enclosed by the 30% or 35% isograms for the rest of the year. What the frequencies of observations reporting rain represent in terms of height of rainfall is not mentioned. However, Miller (1951) published quarterly maps of rainfall based in part on the frequency of rainfall figures of McDonald's atlas, which are comparable to those of the Navy atlas, but based on fewer observations and compiled only from Greenwich noon observations (5 a.m. Clipperton time). Corresponding to the wet region of the Navy atlas, Méller shows for the quarter December - February the isohyet for 400 mm forming a wide strip in the space 0° to 10°N and stretching east to about 97°W. Clip- perton Island is halfway between the 400 and 200 mm lines. In the March to May quarter, Clipperton is still between those lines, and the 400 m tongue scarcely reaches east of 100°W, but is wider than before. On the June - August map it extends eastward all the way to the Atlantic side of Central America, and Clipperton Island is in the middle of it. During that quarter, according to the map, the Gulf of Panama has over 600 mn of rain and the area of the isthmus over 1000 mm. In the last quarter, the 400 mm isohyet still extends to the Atlantic side of Central America, but it crosses the meridian of Clipperton only a short distance north of the island which presumably receives something over 400 mm of rain in that quarter. Such calculated figures cannot give an idea of the exact pattern or amount of rainfall, but are useful as indications of general trends. Clipperton Island is situated between an area of year-round high rainfall to the southwest and one of marked seasonal rainfall to the east, and re- tains in attenuated form the seasonal pattern of the latter, with lower - 22 - rainfall in the winter and spring. The rains during that period "probably result from persistent low-level wind convergence but in the summer... cyclonic depressions or storms seem to be the cause" (Ramage, personal communication). In August and September 1958, exact measurements of all rainfall could not be taken, but an ordinary tin can was kept in the open during the heavy rainstorms which occurred with southwesterly squalls. As much as 125 and 150 mm of rain were recorded during 24 hour periods of violent rain. The total amount recorded for the period August 11 - September 17 was 890 mm. No one who is not familiar with the wet tropics can imagine this type of rain. There is warning from the approaching dark cloud, but when the rain starts, it is with its full intensity and in two seconds one is completely drenched. Protecting notebooks and films is a haunting worry. Because of the strong winds which bring the rain squalls, there is much of what we called “horizontal rain" which seems wetter and more able to get into shelters than the usual kind. ‘Some such downpours last for hours. In October 1956 (Limbaugh, personal communication) it rained almost every afternoon or evening. The amount of rain was quite variable, prob- ably never exceeding an inch (25 mm) in any one day. The measurements made by Limbaugh in 1958, his observations in 1956 and those of Taylor's indicate that much more rain falls in the Clipperton area than Mdller deduces (about 300 + 350 +600 + 450 = 1700 mm). Profes- sor Ramage therefore has suggested another hypothetical calculation which may give a better idea of possible rainfall distribution on Clipperton. "Palmyra Island, also a coral speck is located at 5°53' N, 162°05' W in the western part of the heavy rainfall region which also affects Clip- perton. Between 3 and 5 years' rainfall measurements at Palmyra give an annual mean of 4445 mm with little apparent seasonal variations in rainfall intensities. We might assume that average rainfall intensities are about the same for Clipperton and that the mean rainfall in both areas is propor- tional to the average number of observations of rain recorded in the Navy atlas. The atlas shows that over the year 19.7% of observations in the Palmyra area and 22.2% of observations in the Clipperton area recorded rain. We might then estimate the annual rainfall at Clipperton as Hubs x 22.2 — and then by prorating among the months in accordance with their rain obser- vation frequencies, come up with this hypothetical distribution: i F M A M Ji i A S 0 N Dye! WT 192). QLUowd 86) © Oko 30: FOIm Cb, LOT t6SLl wS5o001'539 It must be pointed out that the difference in the vegetation of Palmyra and Clipperton is much more striking than would be expected from these figures, and that even the somewhat greater seasonality indicated = 85 = for Clipperton would scarcely account for it. Although an absolute correlation between vegetation and rainfall is not expected where such small floras are involved, and considering that tree species may never have reached Clipperton, still this discrepancy suggests that caution should be observed in accepting these hypothetical figures. A new source of information has recently become available in the photographs taken by the weather satellite Tiros III. The following paragraphs describing this type of information were contributed by Mr. Lester F. Hubert, Meteorological Satellite Laboratory, U. S. Weather Bureau. "Meteorological observations made on Clipperton Island have been inadequate to provide a reliable climatology. A small sample of ob- servations is especially vulnerable at this location because several years in succession may display a single type of meteorological regime, followed by a year of radically different weather. This is due to the fact that Clipperton Island is situated on the boundary of the Pacific dry zone--a tongue of desert-like weather lying just north of the Equator. Approxi- mately each decade, during the El Nino” years, this zone is disturbed, that is, becomes more rainy and shifts its position. "The meteorological satellite now provides a means of obtaining the information necessary to reveal the rainfall regime, and the exceptional cases that are sometimes more important than the "average." TIROS pic- tures have demonstrated the potential of detecting and tracking cyclonic disturbances near Clipperton. They show that many more cyclonic storms occur in the area than were previously known from inadequate conventional weather observations. The picture obtained by TIROS III on July 22, 1961, from an altitude of approximately 450 miles, shows the curvature of the earth, the approximate location of Clipperton Island, and that of hurricane Madeline, with bright curved cloud bands that are believed to produce precipitation. Infrared measurements from the satellite can be used to approximate cloud heights, thereby distinguishing between the thick shower- producing clouds reaching up to 15-18000 ft. and higher, even up to the tropopause, and the shallow clouds reaching to 4-5000 ft that usually yield no significant precipitation. During the anomalous (EL Nino*) years, we may find that disturbances (cyclonic storms) pass over Clipperton Is- land. We are now in a position to commence accumulating data for this neglected part of the ocean.” The photograph of Madeline represents a typical summer cyclone situation in that part of the world. The orientation of the cloud bands seems to indicate that the winds at Clipperton were probably from the south west (cf. pp. 10-11). * El Nitto is a warm tropical current from the north, that flows along the north coast of South America (Ecuador, northern Peru) down to about 6°S, during the southern summer. Some years, at more or less regular intervals, it reaches farther south, pushing away from the coast the cold waters that usually flow along it. At the same time, important atmospheric disturbances take place, especially unusually abundant rains that may affect vast regions. me Ol = Atmospheric humidity As could be expected in a wet tropical climate, atmospheric humidity is very high. The lowest relative humidity recorded in August - September 1958, as calculated from the depression of the wet bulb was 69%. Cloudiness and visibility The U. S. Navy Atlas presents data on cloudiness over the oceans on monthly charts including two series of isograms, one representing "% frequency of low cloud amounts 6/10 or more," i.e. cloudy, the other repre- senting "% frequency of total cloud amount 2/10 or less," i.e. relatively clear. Extraction of positions of Clipperton Island in relation to these lines gives, for low clouds, monthly figures ranging from between 20 and 30% in November, in the neighborhood of 30% for most months, and approach- ing or exceeding 40% in July, September and October. The same extraction of positions for total cloudiness gives figures less than 20% for the months from May through December, and ranging between 20 and 30% for the remainder. ‘Thus both sets of isograms indicate a high degree of cloudiness throughout the year, with the summer months, from April to October, even more cloudy than the winter. On the basis of 85 observations between August 9 and September 16, 1958, the sky was completely cloud-covered over 3/4 of the time and half or less covered less than 1/5 of the time. The clouds were generally cumiu'us or cumulonimbus, occasionally stratus or altostratus, with cirrus usueily visible when there were breaks in the lower cloud layers. The bases of the principal clouds Were usually estimated at between 600 and 2000 m altitude. Visiblity is generally always good at Clipperton Island except during rain storms. : ssey) °J pue ysnequtTy °9 Aq peqoeTToo eqep worgy *gc6T ‘puel[s][ uoyweddITO ye oumsseud oTueydsowye pue spuTmM °T eTQey oN MSS MSS MSS S MSS MS MSS Sites TOcasol Iiless0c WOeeSOl | W0ceoLL silc=sOc.| "ce-~Sic ite 9.0c OOTe 0060 OOZT 0060 ST OTOL: qu emssead oOTJeydsowyzy sqouy peods purty wots UOT}OSITP PUTH (Ldd) uty, (8S6T °sny) yyuoy] ay4 Jo Aeq [ee ee ae ee SE SS icon [Bees a ES ES eee, eS SF ee SS SO: ee a ee a ee See a OOT een ee ee We ee s00r x oe oe be et ee OLOL: qu einsseid 0 Lieydsouzy 8 6 wal val eT OT 8 8 9 OT | sqouy peeds puty 64 MSS MS MSS MSs MSS MSS MS MAM MN wots 606 9fee-STe | We-STe 9 70e-G6T WOe-S6T | VOC-S6T Yee-STe We-STe ‘78e-Ske | WE-STE | xWOTJOSITP put OOLT 0060 OOT? OOLT 0060 OOT¢e OOLT OOET 0060 OOT?2 (Ldd) Sty (8S6T *3ny) ime OT ; 6 ° 8 | yquoy euy jo Aeq ee oe ee ee ee ee le z Pee ee ee el = if Or G 0g 4 S € 21 8 aN Msi HSS MSS MS ass MS ‘SM MSS USSG || MMSE? || WESSEL WWESCS)) WISE-SE6 || ISUISSINL | WEG || I.E 3° WISE OOTT 0080 | 000T | 0040 | 0040 | ote 0060 | OTe 0060 TE o€ 6e= |= ee Le 9% Go Ez pe ee i ee et ee SE ee ee eee ff ee eke 2 ee ae ee ee eS ee ee eee eee ee | re ee ee ee es [ee |e Sa ee ae ee ee a ae ee ee eee eee (i Se ae ees Ss | Sane eae a ea ee ee eee | (es |S a ee ee ee ee ee Se eee eee) S 6 G ¢ 8 Ol OT i 8 S G Msi ci MSM MSS ANM fst MS asa SNE EN eaNN MiCaSecal= (once | Scale nic SOc | iecaGUc Semcon ecasca| iclacinn < ol9=SS - Sei Guitlmaae 0060 | 0060 | oote 0060 | 0060 | cote | 0060] core OLN O0CL = 00c0 ae we IP 0g or Oe ak Geese eee 91 @ -d “To Taen €OOT TOOT SCOT 900T LOOT BOOT 600T OLOL?: qu ainsseid dTIeyasowyy sqjouy peeds put}; wooly uoTJOSaTp PUT/K (ldd) ewty (2S6T *sny) YyuoW] sy fo Aeq OOT OOT OOT 900T LOOT 800T 6001 OTOL:qu emmssoid ITIsydsowyzy sqouy peeds puTM wots UOT }OeATP FUTN (Ladd) ewty (8S6T *sny) yquoy, ayy jo Aeq € *d *[ spqey ae ee, | ES ee SO hg i a Ba ee ee OO ZS fis Bee eee x x x SOOT x es Sait RE SS GS Be RS EES FASE CO SRR Es Se ee aS Se SS Oo fa Ree eea 6001 OTOL: qu einssead dTJeydsowyzy oT 8T ST €Z 6T 9T 8T (ord 6T (ers tz sqouy peeds put MSs MSS [MS MS lass Ms MS MSS MSS MSS [KS WOLF "e-S0e ee-STe 9‘*tee7-STe | WEe-See 9WIe-Soz| WEe-Se~ «Wee-STe «NIe-Soz = W1e-So7 *MeZ-Soe §4Ze-Stz uoTJIeITp PUT) 00%z OOOT 00L0 OO8T 0040 (10) 42 OOTT OOOT 0080 00L0 0090 (Lad) euty (8S6T °4deS) 9 : G : q yoy] ayy Jo seq Pet A ie ee eX ae COO Ree ae ee ee ero ele ee i ee es ee ee ee x SOST = Pe yet ere ee ae ee 9001 =a Eo ee ee OL areas sen OLOL:qQu sunsseud od Tleudsow zy 92 He 1G OT q ie ul sjyouy peeds put MSM MSM Ms MS M ASH "ASS wots ° 19¢-SSe «(TWie-Ste §=(ee-ST?| Wee-See = Wee-STe §=6TLT-SOT| 4ST-STT UuOoTJOSITp PUTA 0060 0080 0090 OOLT 0040 0050 0040 (Ldd) Sut (8S6T °>des). ‘ € : z ee yquoy, ayy jo Aeq Ted TeSraen €OOL HOOT COT x 900T LOOT x x x 800T x x x x 600L xX OTOL:qu ainssoid dTIsey dsowyy 40 SO LO SO 90 90 90 90 90 sqjouy peeds Purtpi IASM MSM MSM MSM MSM MSM MSM MSM MSM wots HGe-Stre uSe-Sie %Se-Sie 19¢-SSe% 9e-GS6% 9Se-Stie 9 ‘Te-SEe ‘Te-SEe =WIe-SEe UOTJOSAITP PUT! OOT? 0002 OO6T OO9T OOET OOTT 0060 0080 0040 (Ldd) ety (gS6T °4de: Reh) 3 Free ee a age, ce a " ee ;) <> tae 0 Ka PAL meet t Se . ip: POSE. Oa ey : J iy bh Sytctiiaen na ST, ee ee ae Ole | 4 : 7 " ee ee ek. a mgd ; is H ts a ; mes Fei ae | huh) ¢ a NS Foie y PERN orm : . , fi j Dy Zi { r = } ne ‘ sa & 2 a w ; j ‘ ; ; : Se hig We Vas i @ - : P aD cl a, pgs * salted cammieits band vinairae ae patent dant ype earals yy ercigrenegiend 5 Spat ity 0 oe f ane a. , } ; ; r sy ; : z =) er P > 5 ’ _ ni A = — “” Bony We! } - : t % , zoom Re ee nai en le ae eenctons | ee *< F; | Be se ed aye i i i AB A : ie | ¥ y i Nr a } x 3 2 ; A oe | ee | ; : ' 7 ! wie dy Fi a f vifirbo ones eng een en oi vane palit sienna i <4 E r f : } j x 2 “s i \ a ~ ‘ ‘ —, Ni f ™ n vs a 2 & i ey lax Ay i A i WH 3 , x i Ae i 2 7 i : ; Ba fi a a * ib : Ket ; b ; Se ete ee ee scabies Bechet wenhow iva ale 2 Fite ed ae hye ( Nive per i Betertcsortes ate sagni nsotey Satw ott ols 94 dtnnel of! .Sobe adi dtiv Eft evworw iran’ seu patw band no Sforrs ‘ect to ebiatun tnewiag aid serving ebao2 ipottapsss ne tp rein “aot 10 aoxt mold sed Batw stoi xg tonnes Yorte 3 to dignel Lin Pimo by onus earorie at pat KAO TET exis bas movoru rer) myeceva oil ‘ewode exedtest ‘to todmyn ect Ogio add Th “etceo Suit nb, ferompa ed |. oieoe iL fac! .ebaty oldstwey bas .erke vetged "i wales ‘on ~ ere exit ddtion OT evorts 186709 : sbabrgost” pee ones Table 4. Wind roses for Clipperton Island's 5-degree square (10-15°N, 105-110°W) from Pilot charts The wind percentages are concentrated upon 16 points of the compass. The arrows fly with the wind. ‘The length of the arrow, measured from the outside of the circle on the wind rose, when compared with the attached scale gives the percent of the total number of observations in which the wind has blown from or near the given direction. In some instances the full length of the arrow cannot be shown. In such cases the arrow is broken and the percentage is shown numerically, between the broken lines. The number of feathers shows the average force of the wind on the Beaufort scale. The numeral in the center of the circle gives the percentage of calms, light airs, and variable winds. The numeral in the lower right corner shows for the month the percentage of ship reports in which gales have been recorded. 2060),¥03)0¢ pulm jo 0/095 os oP os oz or tC) i \ ik i ies 2 ey) é t i p } ta tiep Toe i daem rae i as , a) vs Vian es : Y = 250— HYDROGRAPHY Surface currents Clipperton Island is located in an ocean area of variable surface currents. In the eastern part of the tropical Pacific ocean, the North Equatorial Current flows in a general east-west direction around latitude 10° N. The South Equatorial Current flows in the same general direction, straddling the equator and with a large part of its water moving in the northern hemisphere. The Equatorial Countercurrent, flows from west to east between the other two and always in the northern hemisphere. ‘The exact position of these currents varies seasonally. Information on the currents around Clipperton can only be derived from very general charts and summaries applying to the eastern part of the Pacific and the central American region. Most of the available data of this sort are compiled from ships' reports and published for the use of ships (U. S. Hydrographic Office 1947). They have also been summarized by Schott (1935, maps 29, 30). Very little of such information has been collected near Clipperton Island because it is removed from shipping lanes. According to Schott, the North Equatorial Current flows westward around Clipperton during the winter months, and according to data in the U. S. Hydrographic Office (Mr. W. E. Malone, personal communication) at speeds from 0.5 to 1.0 knot. At this season the Equatorial Counter- current is well to the south of the island and, according to some authori- ties, it is weak and at times nonexistent as a well-defined current. During the summer months the Countercurrent is better developed and according to Malone (perscnal communication) the island lies in the shear zone between it and the North Equatorial Current. Schott's map would indicate that the Countercurrent itself always flows past Clipperton in the summer but this generalization is probably only occasionally valid. In the summer of 1958 (Knauss and Pepin 1959) the Countercurrent was 250- 300 km wide. During the IGY there was a great increase in the work on ocean currents and particularly on those of the Eastern Pacific. Indeed the main pur- pose of the Doldrums Expedition was to study the Equatorial Countercurrent. Most of the data from such recent surveys are not yet available but John Knauss, leader of the Doldrums Expedition, considers (personal communi- cation) that the Countercurrent runs about 50 miles south of Clipperton most of the time and probably does not move as far north as Schott has it, in fact probably does not even move as far north as Clipperton every year. The Countercurrent is irregular, can change its direction or speed, or can even disappear within a very short period (Knauss 1960). Besides the seasonal variation in the surface currents around Clipperton, the situa- tion may be complicated by the occurrence of transverse flow between the North Equatorial Current and the Countercurrent. These variable currents during the summer show a slight predominance of east and northeast sets with speeds of 0.5 knot or less (Malone, personal communication). With regard to the transport of propagules of plants and animals by surface currents it must be kept in mind of course that when the Counter- current is well developed, much of its water mass is deflected northwestWard Sle as it reaches central America and flows back toward the west with the North Equatorial Current, therefore it is not impossible that some Indo- Pacific organisms might reach Clipperton Island by this roundabout way rather than by direct eastward flow. Waves There are many ways in which ocean waves of various types affect oceanic islands and coral atolls especially. For example, the very growth of the living reef is linked to the amount of surf and consequent varying oxygenation of the surface water; wave erosion and deposition are major factors in the arrangement of materials that form atolls. The reverse influence of an island as an obstruction to swells is also well known theoretically and has been demonstrated for some (Arthur 1951). In a detailed study of an atoll, therefore, a knowledge of ocean wave patterns is necessary, but it has not generally been feasible, so far, to obtain such knowledge. Only for a few atolls has information on waves been available to be correlated with data on the morphology or ecology (Munk and Sargent, 1954, Guilcher, 1956). In the case of Clipperton, hardly anything is known of the pattern of waves and their influence on the atoll. Yet, besides the more general effects alluded to above, waves have one very practical role on Clipperton, that of entirely preventing landing on the atoll more often than not. The Sailing Directions (U. S. Hydrographic Office 1951) as well as most accounts of visits to the island are replete with warnings of the dangers and difficulties that may be encountered, and make it clear that it is often impractical to attempt a landing: "On the reef, the sea breaks heavily and continually; the surf is terrific and at times covers the whole island" (this last observation is exaggerated, although storm waves can cover the whole width of narrow parts of the atoll ring and pour into the lagoon). There are times, however, when sea and wind are calm and when it is possible to land easily with a small boat over the shallow reef. This was the case when I landed on Clipperton and when I left the island. Waves approaching the island, because the reef front falls off very rapidly, suddenly find themselves over a very shallow area, and break with great violence, the continuous line of such breakers forming the surf (for a general discussion of this and related phenomena, see Bascom 1959). Goua (1952) considers that the swells coming from the southwest to southeast 9 months of the year are refracted around the reef and break on the north part of the atoll, while in February, March and April the north- east trade wind prevails and the southern swells diminish, spring there- fore being the best time to effect landings. This is also rather over- generalized and simplified. Except for these remarks, and for some observations made from shore by Limbaugh in 1958 (ined.) the only available information on waves appli- cable to Clipperton is that generalized from ships' observations in the form of roses for the 5° square (10°-15° N, 105°-110° W) in the U. Ss. Hydrographic Office Charts of Seas and Swellls (1944), and reproduced here = 27 as table 5. Seas may be conveniently defined as reasonably large waves, in the area of their generation, and are characterized by steep sides and sharp crests. Their energy content is generally being augmented by the wind. Waves that have left the area of generation--no new energy being added by the wind--are called swell. They have less steep sides, rounded erests and longer periods than are general for seas. They are gradually losing energy while travelling over great distances, often thousands of kilometers from their area of generation. In any oceanic area the wave pattern is extremely complex, resulting from seas generated locally and swells arriving simultaneously from sever- al directions and points of origin. The effects on atoll morphology are correspondingly difficult to unravel, but are nevertheless very important. No attempt has been made to establish any correlation on Clipperton except to point out that beach ridges, resulting from wave action, occur around most of the island, as do reef-front grooves and spurs (all around), also correlated with wave directions (Munk and Sargent 1954). No recent published information could be found on tides at Clip- perton Island so an inquiry was made at the U. S. Hydrographic Office. The summary of information compiled by Mr. W. E. Malone is quoted below. "Tidal data for Clipperton Island are not given in either British or U. S. tide tables indicating a lack of recent reliable observations. H. O. Publication 84 does contain the notation regarding Clipperton Is- land that high water occurs 8 hours and 40 minutes after an oe or lower lunar transit of the local meridian and that the spring rise is yaad feet [1.30 m/ while the neap rise is 2-2/3 feet /0.80 m/. The basis ae this information is obscure. However, extrapolation from tidal data for surrounding areas indicates that high water at Clipperton Island should occur about 8 to 9 hours after an upper or lower lunar transit of the local meridian and that mean spring tidal range should be between 3 and 4 feet [6. 90 and 1.20 m/. It appears then that the information contained in H. O. Publication 84 is fairly accurate. Tidal currents are reported to flood eastward and ebb westward, however, tidal currents are probably dominated or obscured by the general ocean currents." The information in H. 0. 84 (U. S. Hydrographic Office 1951) is the same as that given on H. O. Chart 1680, Clipperton Island. This chart, based on the 1897 survey of P. J. Hennig, Master Mariner, bears the information: "H.W.F. &C. VIII h. 40m. Springs rise 4t feet. Neaps rise 2 2/3 feet.’ The chart also bears arrows indicating pace tion of ebb and flow curreats, ebb arrow pointing westward, flood arrow eastward. Whether this information was collected by Hennig himself, compiled from his obser- vations and those of others such as Fisher (cf.p. 6) or calculated, is not known. Since that time various visitors to the island have made casual observations on the tides, but their results are not available. In 1956, a tide gavge was installed by a party from a Scripps Institution vessel but the gauge functioned only a very short time before it was removed for repairs. In 1958, its emplacement was found to have been destroyed during the 1957-1958 storm. =o oes Ocean temperatures Schott (1935) presents surface temperatures in maps XX-XXTITI, by means of isograms. For Clipperton, the February temperature is over O7oCr that for May over 28°C, that for August over 27°C and that for Nov- ember 27 C. Agassiz (1906) includes information on water temperature at stations a little distance from Clipperton_(see p. 29). At station 4543 the sur- face temperature was 79.5°F (26.4°C), the bottom temperature (at 2058 fathoms, 3779 m), 34.7 F (1.5°C). At station 4544, surface temperature 80°F (26.9°C), bottom temperature 34.4°F (1.1°9C) at 1955 fathoms (3570m). Water temperatures are also listed in the results of the Swedish Deep-Sea Expedition (see p. 30) and at station 70 surface temperature was 26.93°C while at 905 meters depth it was 4.83°C (bottom at 3690-3860 m). These figures are included in a table of analyses of sea water at varying depths (Bruneau et al., 1953, appendix, p. VII). In August-September 1958, the sea temperature measured in a bucket of surface water collected on the reef ranged from 79°F (26.1°C) at 0900 to 84.9°F (29.4°C) at 1100. cf oneay, P32 winety e) bao ‘Mel: seq ‘one at ede at. pes tak -vaokbevessio to redsua ad? a Sianabiaeraedal soot AOR eule> to Sac) oc Bie sete eft “to mater g@wota st Lewes wih agoitarrado Yo xm vont | tN . ‘videgamel eaudd ‘wl ee to..dnaotey,. ody foa.t [ pKGerreh 2: mn Tiles —- SHOTS salt ar act yd Satwonenye* eEn shinee \ B9%7Snr L .gworrs onkt yrds adit yd ees ignerget coxa anobt bbeo Lfewe eit ome odf .) (eri soot ab anottosiih to diepieg och nody Awoce at tel off seven eave Yo ates ott dobar Swot anvidoerhh odd at if _ iatece fredondts edd ac Danale proc pier torhey oar mort beswERaon Wore (S 2 Gi eemlt to sodaea aif ovip wworrs aif to Chas oat Is Levee ood: f Fane ie) MOsT sai gerclinatil eyed ‘oiduwa cnt saoe ons deus beuraiinads | #4 sae hog- cis re \ abst Hote thaws he sem go. OL ui. noltoaxth 6. timag sy oct rte iy to egataosiss ot VOPR |, e+ Je sely odd poole mwede one ciitionkh exit oyewiloe at totner set Boat memwit 4 salt OAT his akiowe to aaok gurthbem Bix 7 eeateecih. mit cide: wtlowe: ie ayoa sgiivte.anstveg iT sot So Hasta ‘eael ei mabresth ‘to. Bago weg: eat gods .eQesnss nes] ont 6 spbubdses ont al sie mratia ois Goktoonks one’ Aish anotiifines olf 2 aadh oem Sud OE sade . ebized pute "9 bbe qwatl ~lrarninchetd wean q 4 a, wl wedrel exj | aifown. Sas vaus, Te) *aliobtebaco get gahioacil tot Hf oaks mtusntteny, south Bt bd raveLiot as agin Sis go sw2la sigso tddtw “ah y mothe wel) . to seods aL fave, tO 80 a Dem. ‘2 ne i, Pisvome to veo? pillows doe saag: Cede mont gous _ehhare hes gee. ols we £ atone i he i Table 5. Sea and swell roses for Clipperton's 5-degree square The number of observations for sea is shown in the upper left hand corner of the area and the percent of calms for those observations in the upper right hand corner. ‘The number of observations for swell is shown in the lower left hand corner and the percent of calms for those observa- tions in the lower right hand corner. The sea conditions are represented by the light line arrows, while the swell conditions are represented by the heavy line arrows. No arrow is shown when the percent of directions is less than 7. The arrows point in the directions toward which the seas or swells move. ‘The length of the arrow measured from the center mark, when placed on the attached scale, and the numeral at the tail of the arrow, give the number of times in each 100 observations that the seas or swells have been moving from or near the given point. When the percent of direction is 15 or over, the condition within the direction are shown along the shaft of the arrow in percentage of low and medium seas or swells, the first figure from the center is always the percent of low. The percent of high seas or swells within the direction is the remainder of the percentage. when the percent of direction is less than 15 but more than 6, the conditions within the direction are shown by the letter L, M, or H (meaning predominantly low, medium or high) beside the percentage figures for direction. The conditions of seas and swells (low, medium and high) within each direction are defined as follows: low seas and swells, those of amount 1 and 2; medium seas or swells, those of amounts 3 and 4; high seas and swells, those of amounts 5 and above. Scole of seo or swell percentoges . 5 ae pte a pete i Sa oe eetay vy 4a werk te Nae ae tains Y - 2 = SUBMARINE TOPOGRAPHY Ocean floor, Clipperton Ridge The Pacific Ocean between Clipperton Island and the central American coast shows regional depths not exceeding 2000 or 2100 fathoms (3660 to 3640 m) except for some deeper basins and trenches reaching (Middle America Trench) 3000 fathoms (5500 m). The topegraphy of this area has been studied in detail by Menard and Fisher (1958). The same paper describes the Clipperton Fracture Zone, which extends roughly east-west for 3,300 miles (5,300 km) between longitudes 96° and 127°W, and shows for the first time that Clipperton Island is one of the peaks of a ridge, called Clipperton Ridge, 8,000 to 10,000 ft (2500 to 3000 m) high above the ocean floor, and which represents the fracture zone between TO7> and 113°W. The roughly east west ridge is about 60 miles (95 km) wide and 330 miles (530 km) long and is bordered along part of its north side by a narrow deep trough. Clipperton is the only feature reaching the ocean surface, but there are at least two other seamounts on the ridge, one to the northwest, the other to the southeast, of the atoll. To the north- northwest of Clipperton another group of seamounts and deep troughs called the Mathematicians Seamount Range occurs between latitudes 13° and 16°N. The paper by Menard and Fisher includes many maps and sections and should be consulted for further detail and general interpretation. Recent explor- ation of the Clipperton area thus reveals that the atoll is part of an east-west ridge, but remains isolated from other emerged areas, in par- ticular is well separated from the nearest group of islands, the Revilla- gigedo Is., which belong to a different system, and from the American continent. The features of the ocean bottom and the location of the frac- ture systems in the East Pacific are very well. shown in the maps accom- panying a recent article on the East Pacific Rise (Menard foci Ocean bottom Some information on the character of the ocean bottom in the general area of Clipperton can be derived from the records of the Albatross East- ern Tropical Pacific Expedition of 1904-1905 (Agassiz 1906). On the return part of the voyage the Albatross sailed just to the east of Clipperton. Stations 4543 at 8°52'2"N and 108°54'W and 4544 at 10938', 106°47'6" are the nearest to the island. At station 4543 at 2058 fathoms (3775 m) the bottom sample was described as "rd. cly. dk. choc. br. stky. M., few glob., many sm. blk. Mang. part., few Rad., Sponge spic., transp. min. crys. mass of yel. floc. mat." At station 4544 at 1955 fathoms (3570 m) the bottom was described as "stky. dk. choc. br. M., many blk. Mang. part." In the general discussion and in plate 3 (character of the bottom) the Clipperton area is at the limit of a zone to the south and west, where manganese nodules are very abundant, and one to the northeast of "brown and green mud." It is also beyond the northern edge of the area where radiolarians are found, and north of the zone where diatoms are abundant, which zone corresponds well with the cold waters of the Humboldt current (p. 6). 80) & More recently (1947-1948), the Swedish Deep-Sea Expedition explored the Eastern Tropical Pacific with another Albatross and collected infor- mation, including soundings of various types and bottom cores, near Clipperton (Pettersson, pp. 50, 117 and pl. I, 1957). Among the many papers collected in the Expedition's Reports, that of Arrhenius (1952) concerns the ocean bottom of the Eastern Tropical Pacific and the Clip- perton area in particular (station 70, Sept. 20, 1947, core 47). From his very detailed studies, Arrhenius recognizes (pp. 189-190) various environmental subdivisions in the East Equatorial Pacific, represented in a map included in each part of his work (Figs. 1.0, 2.0, etc.) Clip- perton falls in the North Equatorial Carbonate Facies of the East Pacific Ridge, with a calcium carbonate content of the surface sediments between 50 and 75% (map, fig. 1.1.3.1. p.» 14). Core 47 (9°14'N, 109°39'W), from the top of a seamount near Clipperton Island, is described in detail pp. 133-136 and in pl. 2.47. While the core was unsatisfactory for Arrhenius! general correlations of geology and paleobiology of the East Tropical Pacific, it is interesting in the study of Clipperton Island. Arrhenius summarizes his study of core 47 as follows: "Core 47 consists of chalk ooze /Globigerina/ with a few marl ooze layers. The topography of the surroundings, the lack of normal stratification, and the faunal composi- tion make probable that redeposition strongly influences the sequence. The redeposited material appears partly to be of shallow water origin. This is understandable as high peaks rise from the East Pacific Ridge in this area, some of them extending above or close to the surface of the sea (Clipperton Island and Germaine Bank). "The manganese present is to an unusually high degree concentrated to macroscopic nodules. As a result the blackness of the deposit is comparatively low and the hue yellowish.” Undersea mountain No information is available on the mountain upon which Clipperton Island is perched beyond the data given by Menard and Fisher (1958). Their fig. 3 shows the 1400 fathom (2560 m) contour surrounding the base of the Clipperton mountain and that of another seamount to the northwest. The Clipperton mountain is elongate in a NW-SE direction which is also that of the longer axis of the atoll on top of it. According to the same figure, the emerged atoll is located in the southeast part of the mountain top. This had been suggested by earlier fragmentary observations: for instance in 1942, the USS Atlanta reported the 500 fathom (900m) curve as being about one mile off the southern half of the atoll and twice that off the northern half. The existence of submerged peaks to the north- northwest was also reported by the 1957 U. S. Hydrographic Office survey (Obermilller 1959, p. 48). The general slopes of the mountain form the asymptotic curves commonly found on atolls, but nothing is known of their aspect, detailed relief or composition, as there has been no dredging or detailed sounding along them. The only undersea mountains that have been explored in some detail are in the northern Marshalls (Emery et al. 1954). - 31 - Upper slopes and terrace Some information on the upper part of the underwater slopes of Clipperton was collected during the 1958 survey by the marine biologists equipped for diving (Allison 1959a, 1959b, and personal communications). The intertidal reef flat slopes off at its outer edge to about 6 m, forming the reef-front (also called fore-slope). This feature is cut by deep grooves or channels. At the base of this steeper part the slope flattens out somewhat and a surface a few hundred meters wide extends outward and down to 12 or 18 m depth. This has been termed the submarine terrace. It was observed all around the island, but varied in width, becoming narrower on the north side, to the west of the northeast sandwash. At the edge of the submarine terrace, the slopes become more precipitous, and have been estimated as between 25° to 40° or even 60°. The divers were able to make some observations down to 40-45 m along this steep outer slope. Its surface is very rough with channels and crevices around the loosely-piled coral heads and boulders. Some boulders are loosely cemen- ted to the slope and slides can be started by loosening the cement. The marine biologists found that, while some coral and algal growth occurs on the reef front and submarine terrace, it is at the edge of the latter and on the outer slope just beyond that organisms are most abundant and luxuriant. That area was reported as 100% covered with living corals. It is very unfortunate that these extremely valuable observations could not have been extended and documented by soundings and by dredging, and it is to be hoped that such work will be included in future studies of Clipperton. The discovery by the marine biologists of a submarine terrace around Clipperton is of great interest in view of the fact that similar shallow terraces have been well mapped at Bikini (Emery et al. 1954, p. 68 and pl. 68), Eniwetok (p. 95) and Rongelap (pp. 109, 191, pl. 71). The average depth of their edges are of the order of 45 ft (14 m) (Bikini), 40 ft (12 m) (Rongelap) or deeper (8-12 fathoms) (15-18 m) at Eniwetok, where the terrace has been called 10-fathom terrace. The Bikini and Rongelap terraces have a maximum width of 360 m, that at Eniwetok reach- es 3500 m between some of the reef projections. A 10-fathom terrace has also been reported, although not described in detail, in Newell's study of Raroia atoll in the Tuamotus (1956, pp. 334, 341), and is apparently found around other atolls, such as Ifaluk in the Carolines (Bates and Abbott 1958, Tracey et al. 1961), and along other coral reefs. In the northern Marshalls the distribution of live reef forming organisms on the slopes has been described by Wells (1954, p. 398). Little is known of growth on the reef-front, but apparently few corals grow in this zone, while beyond the 10-fathom terrace growth is more luxuriant. Growth on the terrace is poor probably because it is an area of deposition of debris. Tracey et al. (1948) suggested this explanation for Bikini. These observations cannot be compared in detail with those made at Clip- perton but similarities are evident. Interpretation In the northern Marshalls, the 10-fathom terrace, which occurs also along lagoon shores, has generally been interpreted as represen- ting an ancient erosion surface, formed during a glacial epoch when the ocean level had been lowered, and upon which a reef grew during the post- glacial period following, when the ocean became warmer and its level rose slowly. This interpretation is confirmed by the fact that in the de- tailed examination of the cores taken in the drill holes on Bikini islet, sections at depths corresponding to the level of the terrace contain shal- low-water organisms, some of them in position of growth and indicating lagoon edge or reef environments (Emery et al. 1954, pp. 215, 22h, 257.) Reef In a typical atoll the reef is the upper, flat part of the limestone structure capping the undersea volcanic mountain. It usually reaches the intertidal zone and is the seat of dynamic phenomena affecting the atoll: erosion and growth. In most atolls the reef, if not circular, is at least a closed geometric figure encircling the lagoon. The reef may be depressed in some places, forming relatively shallow passages or, espec- ially in some very large atolls, may be interrupted so that deep openings occur. In some atolls, the reef extends at about the same level all around the lagoon and, while there is some water transport between ocean and lagoon over the reef between islets, such inter-island channels are very shallow and cannot be used by boats. In the majority of atolls, dry land consists of islets, discontinuous little patches scattered along the reef surface. Such islets are of two main types (Fosberg and MacNeil 1956): some are only small mounds of unconsolidated debris piled up directly on the reef, others have a core of reef rock forming a sort of obstruction around which the loose material is accumulated (Tercinier BOSD De 98). Such cores are usually erosion remants of a higher reef surface the greatest part of which has been destroyed. In the case of Clipperton Island, a closed oval-shaped continuous reef is topped by a continuous land area. None of the closed atolls has been well studied, and the reasons for the occurrence of a few atolls with a continuous land strip have never been examined in detail. Of some of these atolls, various authors have said that they have been uplifted, and while this may be true in some cases, in others there seems little factual evidence demonstrating uplift. "Uplift" and "raised" are here used to mean only absolute uplift due to a change in the sea-floor in the vicinity of the undersea mountain, excluding the relative change in height in regard to ocean surface which is brought about by changes in ocean level. There is no evidence that Clipperton Island has been raised and no reason to think that its continuous land rim occurs as a result of uplift. I suggest that its closed form may be a reflection of its position in the ocean. The best known of atolls, the northern Marshall Islands, are very strongly asymmetric, a line separating two very different windward and leeward "sides." This line is perpendicular to the resultant of certain vectors, - 33 - the more or less constant directions of predominant winds, waves, and currents. The asymmetry does not affect so much the shape of the atolls as the characters of their reefs: the location of openings in the reefs, of islets on the reef, of areas of maximum growth and other such features. Clipperton Island may be looked wpon, on the contrary, as being to a certain extent radially symmetric, because the predominant winds, waves and currents vary throughout the year and fail to form the sort of oriented "field" which surrounds assymetric atolls. This explanation may fit other atolls with a continuous land rim and its applicability will be studied. If this theory is valid for Clipperton Island, it may explain, not only the presence of emerged land all around the atoll, but some ways in which the reef differs from the reefs of many well known atolls, par- ticularly the northern Marshalls. Around Clipperton Island, the reef flat lines the shore in a regular belt, but with some variation in width. The greatest width, about 130 m, is at the south corner of the atoil and the reef tends to be narrower on the east side of the atoll with a minimum width of about 50 m. During my stay on Clipperton the reef was almost always under some water at low tide. Living corals cannot survive long out of water but in some atolls coral heads or massive algae are sometimes left dry for very short periods by the receding tides. This was only once observed at Clipperton but probably occurs during some spring tides. Landward, the surface of the reef is often covered with sand, especially opposite the sand washes. Coral patches extend over some of the area. Seaward, coral colonies and various algae occupy the reef surface. The striking characteristic of the Clipperton reef is the practical absence of an algal (or "lithothamnion’) ridge. In atolls such as the Marshalls or Tuamotus, the seaward edge of the reef, at least on the windward side of the atoll, rises conspicuously in a rough ridge, of a bright pink color, principally made up of massive coralline algae. At low tide this ridge can be seen very well. iImmedi- ately on its landward side a moat of slightly deeper water separates it from the main part of the reef flat. At Clipperton, this ridge is very poorly developed and when observed from a distance the edge of the reef is marked only by the breakers. However, the marine biologists who were working on the reef-flat or diving outside the atoll noted that the outer edge of the reef was slightly raised, with a vigorous growth of corals and algae. The deep grooves and channels, observed by the divers on the reef front and mentioned earlier, rise to the edge of the reef, and form surge channels. On the vertical air photographs of the atoll the surge channels can be seen well, and appear to have approximately the same importance all around the island. In other atolls, surge channels cut through the algal ridge, and form passages, often roofed-over, between pillars and masses of fast-growing coralline algae, and through which the surf rushes toward the reef, and the backwash retreats. At Clipperton, the surge channels cut through the slightly raised reef edge, but they were not studied in detail. The greater amount of dissolved gasses in the breaking waters is no doubt responsible for the more active growth at the edge as is believed to be the case in other atolls. At low tide, the reef flat at Clipperton Island can be seen as strewn with a great many boulders which were probably torn off the slopes of the reef and thrown up by storm waves. These boulders are of various sizes, Bor _ mostly less than a meter in diameter, which is generally rather small compared to those on some other atolls. They occur all around the island but are especially abundant and concentrated in certain areas, such as near the north corner where they form a boulder field that lies opposite the gravel ridge on shore. The boulders are colored brown or partly bright green by various algae and many organisms grow on their lower parts, including sponges, encrusting Foraminifera, corals, many al- gae, molluscs, as well as associated free living animals. When an atoll reef dries or almost dries, at low tide, patterns of erosion, channels, cracks, pools or other forms normally can be observed. Such patterns were not recognized at Clipperton, partly because most cf the landward side of the reef flat is covered with fine sand. Much of this is only a thin film or is replaced by pebbles or cobbles, with interstices filled with sand. These areas can usually be identified by the abundance of algae growing on the pebbles. In certain places however, the reef is bordered by beach rock, which will be described below, and occasionally the beach rock is located in such a way that it is the seat of active erosion and forms typical erosion ramps. Such a ramp occurs along the northeast coast. A slightly sloping slab of beach rock several meters wide stretches between reef and beach and exhibits a polished surface unevenly pitted into shallow, rounded depressions, which are probably the site of mechanical erosion (abrasion). The walls between the depressions are not sharp and cutting as sometimes happens on other atolls, but form gently rounded, low shoulders between pits. ‘Some of the pits are joined together in shallow troughs, elongated in a beach- reef direction. The landward side of the ramp borders on the sandy beach and, at times, sand may be deposited over the ramp and hide it from view. The reef side of the ramp is occupied by algae, forming a brown felt or an orange crust. In some areas along the northeast coast of the atoll pieces of this ramp, undercut, broken and loosened by storm waves, have been displaced and rest on top of the sandy beach. In some parts of the atoll, small remmants of a ramp occur which may have been derived from an old reef surface rather than from old beach rock. Presumably the reef flat surface extends lagoonward under the island sediments, but nowhere inland did I see any indurated material that could be interpreted as an exposure of this surface. Obermllller (1959, p. 50) makes the same remark. At the lagoon edge, a gently sloping, irregular surface of consolidated rock sometimes forms the bottom for at least a little distance under water; what was seen of this seemed little different from the consolidated phosphatic rock above water, and none of it suggested either a reef surface or lagoon beach rock. Judging from the sample brought back (no. 35), broken from a rough place in this surface, it consists of coral debris indurated with a phosphatic cement. - 35 - SURFACE FEATURES OF LAND STRIP Outer shore Beaches Above the reef flat, often edged by beach rock, the island is surrounded by a beach of sand or gravel; usually this is a narrow stretch of fine coral sand, pinkish in color, sloping up from the beach rock ex- posures. Above it in most areas is a ridge of white coral fragments, in places branched even-sized pieces, elsewhere larger cobbles or boulders. In some areas, as on the southeast part of the island, south of the Hook, there is no sand beach but a steep ridge of coral fragments rises directly above the reef flat. In other places, the beack is wider and steeper and forms the whole oceanside slope, up to the crest of the land strip. In 1958, along a region 430 m long of the narrow northeast side of the atoll, sand had washed all the way from the ocean across the land strip to the lagoon. The beach here was lined oceanward by well-developed slabs of beach rock and on the landward side merged into a lagoonward sloping area, but the break in slope was faint, since the highest point was less than 1 m above estimated mean high tide level. This area will be referred to as the great sand wash. The sand was stained green at the surface and for the first 0.5 cm in depth, by blue-green algae. Below it was pale pink. Transverse lines of pebbles stretched across the wash, probably marking the direction of water flow. Any doubt that the ocean waves poured into the lagoon at this place when it was denuded of vegetation between Nov- ember 1957 and May 1958 would be dispelled by the fact that marine shells, drift seeds and drift pumice had been carried across to the lagoon edge. In the lagoonward part of the wash two low rocky ledges stretched parallel to the shore. Their upper surface was hidden but the sand had been removed from under their lagoonward free edges, so that each protruded slightly over a longitudinal depression several decimeters deep. That nearer the lagoon shore formed a moat full of water green with algae. There were several other, less spectacular, sand washes. Near the East corner of the island, the beach sand reached up to the crest of the island and spilled over the land in a very gentle (about 3 ) lagoonward slope. The beach here was especially mobile and, at the time of our visit, one day formed a slope of about 12° for most of its width, topped by a ~ little vertical cliff, about 30-50 cm high, cut in the sand at high water mark; on a later day the cliff had disappeared and the beach sand sloped regularly up from the ocean to the break in slope. On the southeast side, along Rock Bay, sand also spilled over from the beach onto the gentle lagoomward slope; this sand wash probably marked the location of one of the ancient passages into the lagoon, as the great northeast wash marked the other. Still another area occurred on the southwest side: The sandy beach occupied the whole width of the area between ocean and land crest and the dry land was again covered by a stretch of sand gently sloping to the lagoon. - 36 - In all these places, the sand was white or pale pink-orange where it was usually washed by seawater, and above the reach of ordinary waves the pink sand was hidden under a layer, 0.5 cm or more thick, of sand stained green by blue-green algae and sometimes slightly compacted into a friable crust. Except for this algal material, and for a very few seedlings of Ipomoea pes-caprae as well as occasional seedlings from drift seeds, there was no vegetation in 1958 on Clipperton Island beaches. However, on photographs taken in 1938, 1943 and 1957, Ipomoea vines can be seen to creep down the beaches from the land strip, at least on the east side of the atoll, and in 1958 masses of dead vines hung over the undercut edges of the land strip. The beaches are very mobile and even their upper regions, which could be called storm beaches, are constantly changing under the action of frequent storm waves. In contrast to this mobility, the relative permanence of the sand washes is worth investi- gating. Two of them mark the former shallow openings into the lagoon and most of them can already be seen on photographs taken in 1935 and recognized in older descriptions. They evidently correspond to particular configurations of the modern reef and underwater slopes and probably also of ancient, higher reef flats, that facilitate erosion and sand deposition but not gravel or boulder accumulation. Beach Rock Typical exposures of beach rock are well-marked along most of the seaward northeast coast of Clipperton. They can be clearly seen on the air photos of 1943, and were followed on the ground along most of their course. On the same photos, a dark line about the middle of the south- east coast may represent beach rock, but in 1958 this area was covered by a deposit of boulder or cobble gravel and no consolidated material was seen. Beach rock was also observed in 1958 by other members of the party (personal communication) on the southwest coast, after a storm washed off the coral sand and gravel which had accumulated over it. The several slabs had approximately the same slope as the present beach, and pieces of iron, including part of an old anchor, were embedded in it. Nothing that could be interpreted as beach rock was observed anywhere on the lagoon shore. Along the northeast side, some little distance northwest of the camp, the beach rock presents the following aspect: the beach consists of a thin layer of sand, occasionally removed and exposing a pavement-like layer of white rock. Over part of this pavement, lies another smooth slab also apparently in situ with a small overhang on its landward side. Further north, the upper surface of this slab becomes pitted with shallow pot holes and takes on the aspect of an erosion ramp. Over it lie other thin, broken and displaced slabs, very sonorous when hit with the hammer. A piece of one was collected (no. 17) which seemed to be softer on its undersurface than on top. Some small tufts of green algae were attached to its top surface. Other slabs were colored orange-brown by a felt of algae (no. 306). All the beach rock slopes oceanward, with a general dip of about 10°. Some slabs are cracked parallel to the coast line. In spite of the sandy nature of the beach, almost all of this beach rock is - 37 - not a sandstone but a conglomerate of small fragments of coral and material from a few other organisms, such as shells, with a sandy cement. Therefore beach conglomerate is the approprbte term for this rock. The thin, sonorous slabs are very like typical beach sandstone except for the difference in texture. Continuing along the ocean side, northwestward of the area just described, similar groups of slabs occur, with the lowest one again forming a white pavement about 4 m wide in places covered with beach sand. Above this and immediately landward of it lies a thicker strip over 6 meters wide of coarser conglomerate, also sloping oceanward, with a dip of 17 in its lower part. Its lower edge is undercut and dissected in deep scallops, perhaps somewhat overlapping the pavement. Numerous blue-black Littorina shells are attached to the crevices, as well as a very few specimens of Nerita. Because of its dip and because it is well indurated at its lower edge and over most of its surface, this thick bed shows at least a superficial resemblance to typical beach conglomer- ate. At its upper edge, in its texture and poor induration, it shows a strong resemblance to the flat beds of coarse, poorly indurated material lying to landward and extending across the island. Unfortunately no samples were collected. The material might be interpreted as resulting from the same induration process as formed the flat beds to landward, but applied to the seaward face of a beach ridge. If, on the other hand, it is interpreted as beach rock, the fact that it stretches well above present high tide levels would suggest that it was formed during a prev- ious higher stand of the sea. The thin slabs and the pavement below this layer are intertidal and probably represent modern beach rock. The area of the island here described is part of the land strip which had recently been affected by a violent storm. It may be as a result of this that a long depression, parallel to the shore, about one meter deep and lined with broken coral gravel, was formed between the thick sloping bed just described and the landward flat ledge of indurated gravel. The highest elevation in this general area is 2.25 m above esti- mated mean high tide level. The longitudinal depression flattens out as one continues to walk northwestward along the beach and, farther on, only a regular slope of loose gravel stretches from the flat landward ledge of consolidated gravel to the thick beach rock; still farther, the latter disappears under the gravel slope which extends all the way down to high water mark. The thick elevated "beach rock" was not seen in any other part of the island but stretches of it may be hidden under some of the beach ridges. Along much of the northeast coast where intertidal beach rock occurs, broken and displaced slabs, many too heavy to be moved by one person, are found at the top of the beach which is presumably reached only by storm waves, and are also strewn over the land area. These dry slabs are smooth and white, some with a marble-like glitter. Generally, like those in situ, they are made up of coral fragments of varying size, but none were fine ay in texture to be properly called beach sandstone (cf. sample no. 28). - 36 - Land Beach ridges Seaward of the Naturalists' Camp, there is no well-defined boulder ridge. Boulders and slabs of beach rock, some of them quite thick and heavy, are scattered at the top of the sandy beach, and above it rises the cliff-like edge of the consolidated rock layer which forms the land area: this is undercut and evidently has rather recently receded under the action of storm waves, as dead Ipomoea vines hang over the cliff; they must have formerly extended farther toward the ocean. Elsewhere around the atoll, especially on the south and west sides, outer beach ridges are well developed and consist of well-sorted coral fragments, most commonly about 7-10 cm long. The sorting is apparently facilitated by the fact that the coral fauna is rather poor, and in most places practically the whole mass of gravel comes from one species or a small number of related species. Around the north corner of the island, the ridge includes material of more varied size and also from a greater number of species. The range is from pebbles to cobbles and a few boulders. In that area, the beach ridge is an enormous one and extends inland into a narrow pebble and cobble field, which may have been reworked somewhat at a recent date, as some of the blackened cobbles have been displaced and some of their lighter undersurfaces are exposed. In the same area, a second~ ary ridge or gravel bar has been formed on the seaward side of the reef flat, and cuts off part of the flat as a reef pool. This is reminiscent of a feature observed on Jaluit atoll after the typhoon of January 1958 (Fosberg, personal communication and Blumenstock, ed., 1961). ‘This secondary ridge may be a fugitive feature and will perhaps move land- ward until its material is added to the main ridge. This is what is happening on Jaluit (Blumenstock et al. 1961). ‘There are areas on the southwest coast where the beach ridge of even-sized gravel has a double erest, with a slight longitudinal depression, indicating that the ridge may have been added to secondarily. However it is not absolutely sure that this double-crested ridge is natural, as further north, along the main coGonut grove and the abandoned quonset village, the ridge has obviously been artificially added to, in order to provide greater pro- tection for the camp. Double crested ridges, however with very different Dae are well known from other atolls (Fosberg, personal communi- cation). All the beach ridges are white on their ocean facing slopes. Their flat tops, sometimes depressed in the center in the case of the double- creSted ridges, and their shorter landward slopes are blackened. The line of demarcation between blackened and white coral is sometimes very sharp. The discoloration is due to microscopic algae. The under surfaces of the coral pieces are usually paler and range from white to dark gray. In. some areas and under certain lighting conditions, the dark coral looks very much like a desert of black volcanic lava. - 39 - On the isthmus leading to Clipperton Rock, and more particularly on the land strip parallel to it on the other side of Rock Bay, the pigs had worn conspicuous pale trails in the dark gray coral gravel. These trails can be seen very well in the same area on the air photos taken in 1943. It will be interesting to see how long they will per- sist now that the pigs are gone. Land surface As mentioned earlier, the land slopes from the top of the beach ridges lagoonward all around the island. Generally the slope is steep at first, along the landward side of the beach ridge, then becomes more gentle. Obermilller (1959, p. 50) estimates the latter as 5-10%. In many areas the grade is slight but a series of step-like ledges bring the level down. The surface of the land strip may consist of loose or consolidated material. Unconsolidated material: The sand washes and the inland slopes of the gravel and boulder ridges, as well as the boulder fields have already been mentioned as landward extensions of shore features. Other areas of the land, gener- ally rather flat, are sandy or covered with a very weakly developed soil and often sprinkled with a thin layer of coral fragments. The predomin- ant soil is a mixture of small coral gravel and pale brown silt (cf. samples 5, 34, pp57%50, shown by analyses to be highly phosphatic. In places this may be covered by gravel sheets of varying thiclmess, by linear ridges or stripes of loose blackened coral fragments, or by a layer of sand. Where the consolidated material is exposed, a thin irregular deposit of sand or scattered gravel usually lies in depres- sions or on the surfaces. Of course, due to the extensive disturbance by guano digging, little can be said about the natural characteristics of the soil profiles or the original disposition of the older loose sediments. The more characteristic physiographic features of the land are formed by consolidated sediments. These occur in three main aspects: ledges or steps, pavements, and cliffs at the lagoon edge. The ledges: There are actually two kinds of ledges, both probably exposures of the same flat or almost flat consolidated beds. The first type is exposed only on the ocean side of the northeast land strip: Stretching between Naturalists' Camp and the beginning of the sand wash, the seaward margin of the flat bed just mentioned forms a wide ledge with an abrupt vertical or overhanging escarpment, 0.5 to 1m high, but with the lower part covered by gravel deposits. This ledge occurs opposite and just above the thick coarse oceanward-sloping conglomerate described on Pp. 37. The upper surface of the ledge has a scarcely perceptible slope toward the lagoon. This exposed surface varies in width from almost nothing to perhaps as much as 1/3 or 1/4 of the width of the land strip. In most Swe O aes areas, the consolidated layer is covered landward by a storm-deposited sheet of white gravel composed of somewhat rounded coral branches (see De Tle) Very possibly, the consolidated bed, and its oceanside edge forming the first type of ledge, were exposed at the time of the same storm by the removal of loose material and elsewhere are hidden under the beach ridges. Around the far greater part of the periphery of the atoll, except for the sand washes, runs an interrupted and irregular series of a second type of very low concentric ledges of what appears to be the same material as the first type. Their scarps face the lagoon and are never more than 2 or 3 dm high. Where they are well developed, as on the northeast side just west of the wreck and in some areas of the south- west side, the ledges, from the lagoon shore, look like the bleachers of an amphitheater and one can walk up them as a series of irregular steps. They are often arranged in groups, several steps very close together, then a wide pavement and another group of steps above it. The free edges of the steps often form overhangs sheltering cavernous spaces inhabited by many land crabs. Seen from below, the free edges of a group of steps look somewhat like those of bedded beach rock slabs, but as far as can be told from their irregular surface the ledges do not dip as beach rock would. This agrees with Obermllller's observations (personal communi- cation and 1959, p. 50). Seen from above, the ledges show a consolidated surface often covered in places by loose sand or fine gravel, especially near the base of the next upper ledge. Plants sometimes grow on the sur- face of the ledges, particularly in this loose material. In some areas, for instance near the Hast corner, the lines of ledges are rather regularly parallel to the shore. Elsewhere, they may follow the line of the shore for a while then swing inland away from it. Near the south corner at the base of the Hook there is a vast area where vegetation is rather sparse, and where low ledges are very conspicuous. They form a complicated pattern of arcs which have little relation to the present lagoon coast line. On the other side of the Rock, some series of low ledges swing away from the Isthmus to form the low rocky Thumb Point. Some of the ledges can be fOllowed for considerable distances and it is not impossible that by digging some could be shown to extend more or less continuously all around the island. Measuring their height in rdation to a fixed point (for instance the marker on top of Astro 1957) and mapping their courses might give useful information and help understand their origin. Where they were shown to be absent it might be assumed that a passage once existed, now filled with loose sediments. Exposed pavements: These are especially well developed in the southern part of the atoll, i.e. on part of the Isthmus and on Thumb Point, and on the other side of the Rock, at the base of the Hook. In most areas, they appear as the top surface of ledges, in others they are less obviously connected with them, and disappear under loose material or denser vegetation. Their surfaces often seem horizontal. In places they are rather smooth, more often they appear dissected by erosion, particularly by rain water, into a miniature karst surface a few centimeters high. According to Obermuller, they represent a rich phosphate deposit. my ty es Lagoon shores Lagoon cliffs: When they reach the lagoon, groups of ledges and pavements may end abruptly above the level of the water forming "cliffs."* Both on the northeast lagoon coast, a little north of Naturalists' Camp, and almost diagonally opposite, on the southwest side of the atoll, north and south of the coconut groves, lagoon cliffs oceur which are up to 1.50 m high above lagoon level. The rock forming these cliffs is uneven in structure and hardness so that the cliffs may present overhangs or, on the contrary, be cut back at the top. Commonly they have been worn by erosion and blocks have fallen off from the overhangs, or harder parts have occasion- ally been left in place in front of the new cliff face as "mushrooms." These small remnants, surrounded by water or by loose gravel, are much used by sea-birds as nesting sites as they afford protection from the pigs. On the northeast side, such cliffs extend for a long distance, arising somewhat back from the water's edge and separated from it by a Slope or beach of gravel. Some of this probably originated from the erosion of the cliff and some may have been deposited by the 1957-1958 storm. The cliff was apparently much damaged by this storm and large over- hangs of its upper, more firmly consolidated, part lay broken off and tilted near the lagoon shore. Other lagoon shore types: Along most of the lagoon shore, the land is not so high above water level and rocky overhangs a few decimeters above water level replace the cliffs. They are often hidden under a thick blanket of Ipomoea pes- caprae. They appear to consist of the same type of consolidated rock as the upper part of the taller cliffs. Elsewhere, as in the vicinity of the Rock, a low pavement may extend to the water's edge, forming a very low rocky shore. Consolidated rock or pavement may continue under water for some little distance into the lagoon. In other areas, fine sediments or plant debris accumulate on the bottom at the edge of the lagoon. Mud flats may aiso form the lagoon shore. Because of the con- stantly fluctuating level of the lagoon (tides and perhaps rainfall and evaporation) their width varies and their material may be reworked with the movement of the water. Often, however, they are held by vegetation, i.e. beds of sedges. Near the east corner of the island, and for a long distance south- ward, the consolidated rock layers and ledges swing inland and a low lying area of generally fine sediments spreads between them and the water. The lagoon shore here is of fine white sand or small gravel forming a narrow beach or more often arranged in a low (a few decimeters) and narrow beach ridge. Lower land often lies behind this ridge and may be covered by stagnant pools of water and occupied by sedges. Strong * "Cliff" perhaps gives an exaggerated idea of such low features, but it describes well their abrupt faces, and avoids confusion with some other miniature atoll landforms. ie lagoon waves during storms keep reworking the ridge and at times cut channels across it and flush the pools. The sand seems to extend from the beach for some distance under water into the lagoon. Because of the very gentle slope, the lagoon water covers and uncovers sand flats with every small change in water level. During our visit, there were many wind and rain storms from the southwest which drove great quantities of water plants, principally Najas, toward the eastern half of the atoll. They formed thick masses in the shallow edge of the lagoon and were often deposited in windrows along the beach and beach ridge. Some dried up lines of this material on the top of the beach ridge, or curling crusts of lagoon algae some distance behind it in the mud flats, indicated that at times the lagoon must reach higher and spread farther inland than we ever saw it. The sand beach ridge had probably been relatively stable for a while before our visit, as Ipomoea pes-caprae vines were creeping across it from a large patch inland. The sand was alive with tiger beetles. Mobile sandy beaches and low beach ridges occur also on the north part of the northeast shore, at the lagoon edge of the great sand wash. Here sand or gravel bars often form scallops sometimes closing off tempor- ary ponds, especially at low lagoon levels. Between the camp and the great sand wash, gravel bars often occur at the foot of the cliffs or, at Green Point, along low lying ground,and may cut off from the lagoon some stagnant ponds or moats usually full of algae. These bars occasion- ally are cut by channels and the ponds or moats flushed by lagoon water. The gravel is similar in size and appearance with that of the recent gravel sheet and probably was deposited at the same time. Clipperton Rock In the southeast part of the atoll, a narrow tongue of land called the Isthmus juts into the lagoon in an east west direction and at its free western end rises the irregular volcanic mass called Clipperton Rock. Its height has been variously indicated as from 19 m to 29 m or even 80 m. The latter was estimated from the distance at which the Rock disappeared from view, and is certainly wrong. None of the recent scientific and other surveys included measurements of the Rock, as far as could be ascer- tained. The most recent figure that is based on well-described surveys is that of 29 m above ocean level obtained in 1935 by the officers of the Jeanne-d'Are (Lacroix 1939, Gauthier 1949) but it may be a bit high. Many visitors have been struck by the resemblance of the Rock to a distant sail, or from closer up, to a ruined castle. Indeed these are apt compari- sons. The base of the Rock is roughly lozenge-shaped, with its long axis in a general east west direction. The highest pinnacle is near the center of the mass. Seen from the Isthmus, the east face rises as an almost vertical wall, as does the north face when seen from the lagoon. As one walks around the Rock, several passages, mostly roofless, roughly parallel and separated by thick vertical walls, are seen to lead into it. The floors of these passages slope upward and are covered by fine material - 3 - mixed with coral pebbles, guano, feathers and other debris (cf. sample 15 p- 56). Some of the passages can be followed completely through the Rock. The parallel walls separated by open passages are well shown on some of the photos taken from helicopters. From a superficial examination, the most obvious possible origin for the Rock is that it is part of the crater of the undersea volcano, and this was suggested by Geikie (in Teall, 1898, p. 233). The detailed study of the lithology of the volcanic material, and of descriptions and photographs of the structure of the Rock, led A. Lacroix (1939) to describe it as an extrusive dome, comparable to that of Bogosloff, and formed in the same manner as that of Montagne Pelée. This would make it a cumub-