Thecrabbeginsbytearingthehusk，fibrebyfibre，andalwaysfromthatendunderwhichthethreeeye-holesaresituated；whenthisiscompleted，thecrabcommenceshammeringwithitsheavyclawsononeoftheeye-holestillanopeningismade。Thenturningrounditsbody，bytheaidofitsposteriorandnarrowpairofpincers，itextractsthewhitealbuminoussubstance。IthinkthisisascuriousacaseofinstinctaseverIheardof，andlikewiseofadaptationinstructurebetweentwoobjectsapparentlysoremotefromeachotherintheschemeofnature，asacrabandacocoa-nuttree。TheBirgosisdiurnalinitshabits；buteverynightitissaidtopayavisittothesea，nodoubtforthepurposeofmoisteningitsbranchiae。
　　Theyoungarelikewisehatched，andliveforsometime，onthecoast。Thesecrabsinhabitdeepburrows，whichtheyhollowoutbeneaththerootsoftrees；andwheretheyaccumulatesurprisingquantitiesofthepickedfibresofthecocoa-nuthusk，onwhichtheyrestasonabed。TheMalayssometimestakeadvantageofthis，andcollectthefibrousmasstouseasjunk。Thesecrabsareverygoodtoeat；moreover，underthetailofthelargeronesthereisamassoffat，which，whenmelted，sometimesyieldsasmuchasaquartbottlefulloflimpidoil。IthasbeenstatedbysomeauthorsthattheBirgoscrawlsupthecocoa-nuttreesforthepurposeofstealingthenuts：Iverymuchdoubtthepossibilityofthis；butwiththePandanus[8]thetaskwouldbeverymucheasier。IwastoldbyMr。LieskthatontheseislandstheBirgoslivesonlyonthenutswhichhavefallentotheground。
　　CaptainMoresbyinformsmethatthiscrabinhabitstheChagosandSeychellegroups，butnottheneighbouringMaldivaarchipelago。ItformerlyaboundedatMauritius，butonlyafewsmallonesarenowfoundthere。InthePacific，thisspecies，oronewithcloselyalliedhabits，issaid[9]toinhabitasinglecoralisland，northoftheSocietygroup。Toshowthewonderfulstrengthofthefrontpairofpincers，Imaymention，thatCaptainMoresbyconfinedoneinastrongtin-box，whichhadheldbiscuits，thelidbeingsecuredwithwire；butthecrabturneddowntheedgesandescaped。
　　Inturningdowntheedges，itactuallypunchedmanysmallholesquitethroughthetin！
　　IwasagooddealsurprisedbyfindingtwospeciesofcoralofthegenusMilleporaM。complanataandalcicornis，possessedofthepowerofstinging。
　　Thestonybranchesorplates，whentakenfreshfromthewater，haveaharshfeelandarenotslimy，althoughpossessingastronganddisagreeablesmell。
　　Thestingingpropertyseemstovaryindifferentspecimens：whenapiecewaspressedorrubbedonthetenderskinofthefaceorarm，aprickingsensationwasusuallycaused，whichcameonaftertheintervalofasecond，andlastedonlyforafewminutes。Oneday，however，bymerelytouchingmyfacewithoneofthebranches，painwasinstantaneouslycaused；itincreasedasusualafterafewseconds，andremainingsharpforsomeminutes，wasperceptibleforhalfanhourafterwards。Thesensationwasasbadasthatfromanettle，butmorelikethatcausedbythePhysaliaorPortugueseman-of-war。Littleredspotswereproducedonthetenderskinofthearm，whichappearedasiftheywouldhaveformedwaterypustules，butdidnot。
　　M。QuoymentionsthiscaseoftheMillepora；andIhaveheardofstingingcoralsintheWestIndies。Manymarineanimalsseemtohavethispowerofstinging：besidesthePortugueseman-of-war，manyjelly-fish，andtheAplysiaorsea-slugoftheCapedeVerdIslands，itisstatedinthevoyageoftheAstrolabe，thatanActiniaorsea-anemone，aswellasaflexiblecorallinealliedtoSertularia，bothpossessthismeansofoffenceordefence。
　　IntheEastIndiansea，astingingsea-weedissaidtobefound。
　　Twospeciesoffish，ofthegenusScarus，whicharecommonhere，exclusivelyfeedoncoral：botharecolouredofasplendidbluish-green，onelivinginvariablyinthelagoon，andtheotheramongsttheouterbreakers。Mr。
　　Lieskassuredus，thathehadrepeatedlyseenwholeshoalsgrazingwiththeirstrongbonyjawsonthetopsofthecoralbranches：Iopenedtheintestinesofseveral，andfoundthemdistendedwithyellowishcalcareoussandymud。TheslimydisgustingHoluthuriaealliedtoourstar-fish，whichtheChinesegourmandsaresofondof，alsofeedlargely，asIaminformedbyDr。Allan，oncorals；andthebonyapparatuswithintheirbodiesseemswelladaptedforthisend。TheseHoluthuriae，thefish，thenumerousburrowingshells，andnereidousworms，whichperforateeveryblockofdeadcoral，mustbeveryefficientagentsinproducingthefinewhitemudwhichliesatthebottomandontheshoresofthelagoon。Aportion，however，ofthismud，whichwhenwetresembledpoundedchalk，wasfoundbyProfessorEhrenbergtobepartlycomposedofsiliceous-shieldedinfusoria。
　　April12th——InthemorningwestoodoutofthelagoononourpassagetotheIsleofFrance。Iamgladwehavevisitedtheseislands：suchformationssurelyrankhighamongstthewonderfulobjectsofthisworld。CaptainFitzRoyfoundnobottomwithaline7200feetinlength，atthedistanceofonly2200yardsfromtheshore；hencethisislandformsaloftysubmarinemountain，withsidessteepereventhanthoseofthemostabruptvolcaniccone。Thesaucer-shapedsummitisnearlytenmilesacross；andeverysingleatom，[10]fromtheleastparticletothelargestfragmentofrock，inthisgreatpile，whichhoweverissmallcomparedwithverymanyotherlagoon-islands，bearsthestampofhavingbeensubjectedtoorganicarrangement。WefeelsurprisewhentravellerstellusofthevastdimensionsofthePyramidsandothergreatruins，buthowutterlyinsignificantarethegreatestofthese，whencomparedtothesemountainsofstoneaccumulatedbytheagencyofvariousminuteandtenderanimals！
　　Thisisawonderwhichdoesnotatfirststriketheeyeofthebody，but，afterreflection，theeyeofreason。
　　Iwillnowgiveaverybriefaccountofthethreegreatclassesofcoral-reefs；
　　namely，Atolls，Barrier，andFringing-reefs，andwillexplainmyviews[11]ontheirformation。AlmosteveryvoyagerwhohascrossedthePacifichasexpressedhisunboundedastonishmentatthelagoon-islands，orasIshallforthefuturecallthembytheirIndiannameofatolls，andhasattemptedsomeexplanation。Evenaslongagoastheyear1605，PyrarddeLavalwellexclaimed，“C’est[picture]
　　unemerveilledevoirchacundecesatollons，environned’ungrandbancdepierretoutautour，n’yayantpointd’artificehumain。”TheaccompanyingsketchofWhitsundayIslandinthePacific，copiedfrom，Capt。Beechey’sadmirableVoyage，givesbutafaintideaofthesingularaspectofanatoll：
　　itisoneofthesmallestsize，andhasitsnarrowisletsunitedtogetherinaring。Theimmensityoftheocean，thefuryofthebreakers，contrastedwiththelownessofthelandandthesmoothnessofthebrightgreenwaterwithinthelagoon，canhardlybeimaginedwithouthavingbeenseen。
　　Theearliervoyagersfanciedthatthecoral-buildinganimalsinstinctivelybuiltuptheirgreatcirclestoaffordthemselvesprotectionintheinnerparts；butsofaristhisfromthetruth，thatthosemassivekinds，towhosegrowthontheexposedoutershorestheveryexistenceofthereefdepends，cannotlivewithinthelagoon，whereotherdelicately-branchingkindsflourish。Moreover，onthisview，manyspeciesofdistinctgeneraandfamiliesaresupposedtocombineforoneend；andofsuchacombination，notasingleinstancecanbefoundinthewholeofnature。Thetheorythathasbeenmostgenerallyreceivedis，thatatollsarebasedonsubmarinecraters；butwhenweconsidertheformandsizeofsome，thenumber，proximity，andrelativepositionsofothers，thisidealosesitsplausiblecharacter：
　　thusSuadivaatollis44geographicalmilesindiameterinoneline，by34milesinanotherline；Rimskyis54by20milesacross，andithasastrangelysinuousmargin；Bowatollis30mileslong，andonanaverageonly6inwidth；Menchicoffatollconsistsofthreeatollsunitedortiedtogether。Thistheory，moreover，istotallyinapplicabletothenorthernMaldivaatollsintheIndianOceanoneofwhichis88milesinlength，andbetween10and20inbreadth，fortheyarenotboundedlikeordinaryatollsbynarrowreefs，butbyavastnumberofseparatelittleatolls；
　　otherlittleatollsrisingoutofthegreatcentrallagoon-likespaces。
　　AthirdandbettertheorywasadvancedbyChamisso，whothoughtthatfromthecoralsgrowingmorevigorouslywhereexposedtotheopensea，asundoubtedlyisthecase，theouteredgeswouldgrowupfromthegeneralfoundationbeforeanyotherpart，andthatthiswouldaccountfortheringorcup-shapedstructure。Butweshallimmediatelysee，thatinthis，aswellasinthecrater-theory，amostimportantconsiderationhasbeenoverlooked，namely，onwhathavethereef-buildingcorals，whichcannotliveatagreatdepth，basedtheirmassivestructures？
　　NumeroussoundingswerecarefullytakenbyCaptainFitzRoyonthesteepoutsideofKeelingatoll，anditwasfoundthatwithintenfathoms，thepreparedtallowatthebottomofthelead，invariablycameupmarkedwiththeimpressionoflivingcorals，butasperfectlycleanasifithadbeendroppedonacarpetofturf；asthedepthincreased，theimpressionsbecamelessnumerous，buttheadheringparticlesofsandmoreandmorenumerous，untilatlastitwasevidentthatthebottomconsistedofasmoothsandylayer：tocarryontheanalogyoftheturf，thebladesofgrassgrewthinnerandthinner，tillatlastthesoilwassosterile，thatnothingsprangfromit。Fromtheseobservations，confirmedbymanyothers，itmaybesafelyinferredthattheutmostdepthatwhichcoralscanconstructreefsisbetween20and30fathoms。NowthereareenormousareasinthePacificandIndianOcean，inwhicheverysingleislandisofcoralformation，andisraisedonlytothatheighttowhichthewavescanthrowupfragments，andthewindspileupsand。ThusRadackgroupofatollsisanirregularsquare，520mileslongand240broad；theLowArchipelagoiselliptic-formed，840
　　milesinitslonger，and420initsshorteraxis：thereareothersmallgroupsandsinglelowislandsbetweenthesetwoarchipelagoes，makingalinearspaceofoceanactuallymorethan4000milesinlength，inwhichnotonesingleislandrisesabovethespecifiedheight。Again，intheIndianOceanthereisaspaceofocean1500milesinlength，includingthreearchipelagoes，inwhicheveryislandislowandofcoralformation。Fromthefactofthereef-buildingcoralsnotlivingatgreatdepths，itisabsolutelycertainthatthroughoutthesevastareas，whereverthereisnowanatoll，afoundationmusthaveoriginallyexistedwithinadepthoffrom20to30fathomsfromthesurface。Itisimprobableinthehighestdegreethatbroad，lofty，isolated，steep-sidedbanksofsediment，arrangedingroupsandlineshundredsofleaguesinlength，couldhavebeendepositedinthecentralandprofoundestpartsofthePacificandIndianOceans，atanimmensedistancefromanycontinent，andwherethewaterisperfectlylimpid。Itisequallyimprobablethattheelevatoryforcesshouldhaveupliftedthroughouttheabovevastareas，innumerablegreatrockybankswithin20to30fathoms，or120to180feet，ofthesurfaceofthesea，andnotonesinglepointabovethatlevel；forwhereonthewholesurfaceoftheglobecanwefindasinglechainofmountains，evenafewhundredmilesinlength，withtheirmanysummitsrisingwithinafewfeetofagivenlevel，andnotonepinnacleaboveit？Ifthenthefoundations，whencetheatoll-buildingcoralssprang，werenotformedofsediment，andiftheywerenotlifteduptotherequiredlevel，theymustofnecessityhavesubsidedintoit；andthisatoncesolvesthedifficulty。Forasmountainaftermountain，andislandafterisland，slowlysankbeneaththewater，freshbaseswouldbesuccessivelyaffordedforthegrowthofthecorals。Itisimpossibleheretoenterintoallthenecessarydetails，butIventuretodefy[12]anyonetoexplaininanyothermannerhowitispossiblethatnumerousislandsshouldbedistributedthroughoutvastareas——alltheislandsbeinglow——allbeingbuiltofcorals，absolutelyrequiringafoundationwithinalimiteddepthfromthesurface。
　　Beforeexplaininghowatoll-formedreefsacquiretheirpeculiarstructure，wemustturntothesecondgreatclass，namely，Barrier-reefs。Theseeitherextendinstraightlinesinfrontoftheshoresofacontinentorofalargeisland，ortheyencirclesmallerislands；inbothcases，beingseparatedfromthelandbyabroadandratherdeepchannelofwater，analogoustothelagoonwithinanatoll。Itisremarkablehowlittleattentionhasbeenpaidtoencirclingbarrier-reefs；yettheyaretrulywonderfulstructures。
　　ThefollowingsketchrepresentspartofthebarrierencirclingtheislandofBolabolainthePacific，asseenfromoneofthecentralpeaks。Inthisinstancethewholelineofreefhasbeenconvertedintoland；butusuallyasnow-whitelineofgreatbreakers，withonlyhereandthereasinglelowisletcrownedwithcocoa-nuttrees，dividesthedarkheavingwatersoftheoceanfromthelight-greenexpanseofthelagoon-channel。Andthequietwatersofthischannelgenerallybatheafringeoflowalluvialsoil，loadedwiththemostbeautifulproductionsofthetropics，andlyingatthefootofthewild，abrupt，centralmountains。