第5章

Finally，then，Iconcludethatthegreatervariabilityofspecificcharacters，orthosewhichdistinguishspeciesfromspecies，thanofgenericcharacters，orthosewhichthespeciespossessincommon。thatthefrequentextremevariabilityofanypartwhichisdevelopedinaspeciesinanextraordinarymannerincomparisonwiththesamepartinitscongeners。andthenotgreatdegreeofvariabilityinapart，howeverextraordinarilyitmaybedeveloped，ifitbecommontoawholegroupofspecies。thatthegreatvariabilityofsecondarysexualcharacters，andthegreatamountofdifferenceinthesesamecharactersbetweencloselyalliedspecies。thatsecondarysexualandordinaryspecificdifferencesaregenerallydisplayedinthesamepartsoftheorganisation，areallprinciplescloselyconnectedtogether。Allbeingmainlyduetothespeciesofthesamegrouphavingdescendedfromacommonprogenitor，fromwhomtheyhaveinheritedmuchincommon，topartswhichhaverecentlyandlargelyvariedbeingmorelikelystilltogoonvaryingthanpartswhichhavelongbeeninheritedandhavenotvaried，tonaturalselectionhavingmoreorlesscompletely，accordingtothelapseoftime，overmasteredthetendencytoreversionandtofurthervariability，tosexualselectionbeinglessrigidthanordinaryselection，andtovariationsinthesamepartshavingbeenaccumulatedbynaturalandsexualselection，andthusadaptedforsecondarysexual，andforordinaryspecificpurposes。Distinctspeciespresentanalogousvariations。andavarietyofonespeciesoftenassumessomeofthecharactersofanalliedspecies，orrevertstosomeofthecharactersofanearlyprogenitor。Thesepropositionswillbemostreadilyunderstoodbylookingtoourdomesticraces。Themostdistinctbreedsofpigeons，incountriesmostwidelyapart，presentsub-varietieswithreversedfeathersontheheadandfeathersonthefeet，charactersnotpossessedbytheaboriginalrock-pigeon。thesethenareanalogousvariationsintwoormoredistinctraces。Thefrequentpresenceoffourteenorevensixteentail-feathersinthepouter，maybeconsideredasavariationrepresentingthenormalstructureofanotherrace，thefantail。Ipresumethatnoonewilldoubtthatallsuchanalogousvariationsareduetotheseveralracesofthepigeonhavinginheritedfromacommonparentthesameconstitutionandtendencytovariation，whenactedonbysimilarunknowninfluences。Inthevegetablekingdomwehaveacaseofanalogousvariation，intheenlargedstems，orrootsascommonlycalled，oftheSwedishturnipandRutabaga，plantswhichseveralbotanistsrankasvarietiesproducedbycultivationfromacommonparent：ifthisbenotso，thecasewillthenbeoneofanalogousvariationintwoso-calleddistinctspecies。andtotheseathirdmaybeadded，namely，thecommonturnip。Accordingtotheordinaryviewofeachspecieshavingbeenindependentlycreated，weshouldhavetoattributethissimilarityintheenlargedstemsofthesethreeplants，nottotheveracausaofcommunityofdescent，andaconsequenttendencytovaryinalikemanner，buttothreeseparateyetcloselyrelatedactsofcreation。

Withpigeons，however，wehaveanothercase，namely，theoccasionalappearanceinallthebreeds，ofslaty-bluebirdswithtwoblackbarsonthewings，awhiterump，abarattheendofthetail，withtheouterfeathersexternallyedgedneartheirbaseswithwhite。Asallthesemarksarecharacteristicoftheparentrock-pigeon，Ipresumethatnoonewilldoubtthatthisisacaseofreversion，andnotofanewyetanalogousvariationappearingintheseveralbreeds。WemayIthinkconfidentlycometothisconclusion，because，aswehaveseen，thesecolouredmarksareeminentlyliabletoappearinthecrossedoffspringoftwodistinctanddifferentlycolouredbreeds。andinthiscasethereisnothingintheexternalconditionsoflifetocausethereappearanceoftheslaty-blue，withtheseveralmarks，beyondtheinfluenceofthemereactofcrossingonthelawsofinheritance。

Nodoubtitisaverysurprisingfactthatcharactersshouldreappearafterhavingbeenlostformany，perhapsforhundredsofgenerations。Butwhenabreedhasbeencrossedonlyoncebysomeotherbreed，theoffspringoccasionallyshowatendencytorevertincharactertotheforeignbreedformanygenerationssomesay，foradozenorevenascoreofgenerations。

Aftertwelvegenerations，theproportionofblood，touseacommonexpression，ofanyoneancestor，isonly1in2048。andyet，aswesee，itisgenerallybelievedthatatendencytoreversionisretainedbythisverysmallproportionofforeignblood。Inabreedwhichhasnotbeencrossed，butinwhichbothparentshavelostsomecharacterwhichtheirprogenitorpossessed，thetendency，whetherstrongorweak，toreproducethelostcharactermightbe，aswasformerlyremarked，forallthatwecanseetothecontrary，transmittedforalmostanynumberofgenerations。Whenacharacterwhichhasbeenlostinabreed，reappearsafteragreatnumberofgenerations，themostprobablehypothesisis，notthattheoffspringsuddenlytakesafteranancestorsomehundredgenerationsdistant，butthatineachsuccessivegenerationtherehasbeenatendencytoreproducethecharacterinquestion，whichatlast，underunknownfavourableconditions，gainsanascendancy。

Forinstance，itisprobablethatineachgenerationofthebarb-pigeon，whichproducesmostrarelyablueandblack-barredbird，therehasbeenatendencyineachgenerationintheplumagetoassumethiscolour。Thisviewishypothetical，butcouldbesupportedbysomefacts。andIcanseenomoreabstractimprobabilityinatendencytoproduceanycharacterbeinginheritedforanendlessnumberofgenerations，thaninquiteuselessorrudimentaryorgansbeing，asweallknowthemtobe，thusinherited。Indeed，wemaysometimesobserveameretendencytoproducearudimentinherited：

forinstance，inthecommonsnapdragon(Antirrhinum)arudimentofafifthstamensooftenappears，thatthisplantmusthaveaninheritedtendencytoproduceit。

Asallthespeciesofthesamegenusaresupposed，onmytheory，tohavedescendedfromacommonparent，itmightbeexpectedthattheywouldoccasionallyvaryinananalogousmanner。sothatavarietyofonespecieswouldresembleinsomeofitscharactersanotherspecies。thisotherspeciesbeingonmyviewonlyawell-markedandpermanentvariety。Butcharactersthusgainedwouldprobablybeofanunimportantnature，forthepresenceofallimportantcharacterswillbegovernedbynaturalselection，inaccordancewiththediversehabitsofthespecies，andwillnotbelefttothemutualactionoftheconditionsoflifeandofasimilarinheritedconstitution。

Itmightfurtherbeexpectedthatthespeciesofthesamegenuswouldoccasionallyexhibitreversionstolostancestralcharacters。As，however，weneverknowtheexactcharacterofthecommonancestorofagroup，wecouldnotdistinguishthesetwocases：if，forinstance，wedidnotknowthattherock-pigeonwasnotfeather-footedorturn-crowned，wecouldnothavetold，whetherthesecharactersinourdomesticbreedswerereversionsoronlyanalogousvariations。butwemighthaveinferredthatthebluenesswasacaseofreversion，fromthenumberofthemarkings，whicharecorrelatedwiththebluetint，andwhichitdoesnotappearprobablewouldallappeartogetherfromsimplevariation。Moreespeciallywemighthaveinferredthis，fromthebluecolourandmarkssooftenappearingwhendistinctbreedsofdiversecoloursarecrossed。Hence，thoughundernatureitmustgenerallybeleftdoubtful，whatcasesarereversionstoanancientlyexistingcharacter，andwhatarenewbutanalogousvariations，yetweought，onmytheory，sometimestofindthevaryingoffspringofaspeciesassumingcharacters(eitherfromreversionorfromanalogousvariation)whichalreadyoccurinsomemembersofthesamegroup。Andthisundoubtedlyisthecaseinnature。

Aconsiderablepartofthedifficultyinrecognisingavariablespeciesinoursystematicworks，isduetoitsvarietiesmocking，asitwere，comeoftheotherspeciesofthesamegenus。Aconsiderablecatalogue，also，couldbegivenofformsintermediatebetweentwootherforms，whichthemselvesmustbedoubtfullyrankedaseithervarietiesorspecies，thattheoneinvaryinghasassumedsomeofthecharactersoftheother，soastoproducetheintermediateform。Butthebestevidenceisaffordedbypartsororgansofanimportantanduniformnatureoccasionallyvaryingsoastoacquire，insomedegree，thecharacterofthesamepartororganinanalliedspecies。

Ihavecollectedalonglistofsuchcases。buthere，asbefore，Ilieunderagreatdisadvantageinnotbeingabletogivethem。Icanonlyrepeatthatsuchcasescertainlydooccur，andseemtomeveryremarkable。

Iwill，however，giveonecuriousandcomplexcase，notindeedasaffectinganyimportantcharacter，butfromoccurringinseveralspeciesofthesamegenus，partlyunderdomesticationandpartlyundernature。Itisacaseapparentlyofreversion。Theassnotrarelyhasverydistincttransversebarsonitslegs，likethoseofazebra：ithasbeenassertedthattheseareplainestinthefoal，andfrominquirieswhichIhavemade，Ibelievethistobetrue。Ithasalsobeenassertedthatthestripeoneachshoulderissometimesdouble。Theshoulder-stripeiscertainlyveryvariableinlengthandoutline。Awhiteass，butnotanalbino，hasbeendescribedwithouteitherspinalorshoulder-stripe。andthesestripesaresometimesveryobscure，oractuallyquitelost，indark-colouredasses。ThekoulanofPallasissaidtohavebeenseenwithadoubleshoulder-stripe。buttracesofit，asstatedbyMrBlythandothers，occasionallyappear：andIhavebeeninformedbyColonelPoolethatfoalsofthisspeciesaregenerallystripedonthelegs，andfaintlyontheshoulder。Thequagga，thoughsoplainlybarredlikeazebraoverthebody，iswithoutbarsonthelegs。

butDrGrayhasfiguredonespecimenwithverydistinctzebra-likebarsonthehocks。

Withrespecttothehorse，IhavecollectedcasesinEnglandofthespinalstripeinhorsesofthemostdistinctbreeds，andofallcolours。transversebarsonthelegsarenotrareinduns，mouse-duns，andinoneinstanceinachestnut：afaintshoulder-stripemaysometimesbeseeninduns，andIhaveseenatraceinabayhorse。MysonmadeacarefulexaminationandsketchformeofadunBelgiancart-horsewithadoublestripeoneachshoulderandwithleg-stripes。andaman，whomIcanimplicitlytrust，hasexaminedformeasmalldunWelchponywiththreeshortparallelstripesoneachshoulder。

Inthenorth-westpartofIndiatheKattywarbreedofhorsesissogenerallystriped，that，asIhearfromColonelPoole，whoexaminedthebreedfortheIndianGovernment，ahorsewithoutstripesisnotconsideredaspurely-bred。

Thespineisalwaysstriped。thelegsaregenerallybarred。andtheshoulder-stripe，whichissometimesdoubleandsometimestreble，iscommon。thesideoftheface，moreover，issometimesstriped。Thestripesareplainestinthefoal。andsometimesquitedisappearinoldhorses。ColonelPoolehasseenbothgrayandbayKattywarhorsesstripedwhenfirstfoaled。Ihave，also，reasontosuspect，frominformationgivenmebyMr。W。W。Edwards，thatwiththeEnglishrace-horsethespinalstripeismuchcommonerinthefoalthaninthefull-grownanimal。Withouthereenteringonfurtherdetails，ImaystatethatIhavecollectedcasesoflegandshoulderstripesinhorsesofverydifferentbreeds，invariouscountriesfromBritaintoEasternChina。andfromNorwayinthenorthtotheMalayArchipelagointhesouth。

Inallpartsoftheworldthesestripesoccurfaroftenestindunsandmouse-duns。bythetermdunalargerangeofcolourisincluded，fromonebetweenbrownandblacktoacloseapproachtocream-colour。

IamawarethatColonelHamiltonSmith，whohaswrittenonthissubject，believesthattheseveralbreedsofthehorsehavedescendedfromseveralaboriginalspeciesoneofwhich，thedun，wasstriped。andthattheabove-describedappearancesareallduetoancientcrosseswiththedunstock。ButIamnotatallsatisfiedwiththistheory，andshouldbelothtoapplyittobreedssodistinctastheheavyBelgiancart-horse，Welchponies，cobs，thelankyKattywarrace，&。c。，inhabitingthemostdistantpartsoftheworld。

Nowletusturntotheeffectsofcrossingtheseveralspeciesofthehorse-genus。Rollinasserts，thatthecommonmulefromtheassandhorseisparticularlyapttohavebarsonitslegs。Ioncesawamulewithitslegssomuchstripedthatanyoneatfirstwouldhavethoughtthatitmusthavebeentheproductofazebra。andMr。W。C。Martin，inhisexcellenttreatiseonthehorse，hasgivenafigureofasimilarmule。Infourcoloureddrawings，whichIhaveseen，ofhybridsbetweentheassandzebra，thelegsweremuchmoreplainlybarredthantherestofthebody。andinoneofthemtherewasadoubleshoulder-stripe。InLordMoreton’sfamoushybridfromachestnutmareandmalequagga，thehybrid，andeventhepureoffspringsubsequentlyproducedfromthemarebyablackArabiansire，weremuchmoreplainlybarredacrossthelegsthaniseventhepurequagga。Lastly，andthisisanothermostremarkablecase，ahybridhasbeenfiguredbyDrGray(andheinformsmethatheknowsofasecondcase)fromtheassandthehemionus。andthishybrid，thoughtheassseldomhasstripesonitslegsandthehemionushasnoneandhasnotevenashoulder-stripe，neverthelesshadallfourlegsbarred，andhadthreeshortshoulder-stripes，likethoseonthedunWelchpony，andevenhadsomezebra-likestripesonthesidesofitsface。Withrespecttothislastfact，Iwassoconvincedthatnotevenastripeofcolourappearsfromwhatwouldcommonlybecalledanaccident，thatIwasledsolelyfromtheoccurrenceoftheface-stripesonthishybridfromtheassandhemionus，toaskColonelPoolewhethersuchface-stripeseveroccurintheeminentlystripedKattywarbreedofhorses，andwas，aswehaveseen，answeredintheaffirmative。

Whatnowarewetosaytotheseseveralfacts？Weseeseveralverydistinctspeciesofthehorse-genusbecoming，bysimplevariation，stripedonthelegslikeazebra，orstripedontheshoulderslikeanass。Inthehorseweseethistendencystrongwheneveraduntintappearsatintwhichapproachestothatofthegeneralcolouringoftheotherspeciesofthegenus。Theappearanceofthestripesisnotaccompaniedbyanychangeofformorbyanyothernewcharacter。Weseethistendencytobecomestripedmoststronglydisplayedinhybridsfrombetweenseveralofthemostdistinctspecies。

Nowobservethecaseoftheseveralbreedsofpigeons：theyaredescendedfromapigeon(includingtwoorthreesub-speciesorgeographicalraces)

ofabluishcolour，withcertainbarsandothermarks。andwhenanybreedassumesbysimplevariationabluishtint，thesebarsandothermarksinvariablyreappear。butwithoutanyotherchangeofformorcharacter。Whentheoldestandtruestbreedsofvariouscoloursarecrossed，weseeastrongtendencyforthebluetintandbarsandmarkstoreappearinthemongrels。Ihavestatedthatthemostprobablehypothesistoaccountforthereappearanceofveryancientcharacters，isthatthereisatendencyintheyoungofeachsuccessivegenerationtoproducethelong-lostcharacter，andthatthistendency，fromunknowncauses，sometimesprevails。Andwehavejustseenthatinseveralspeciesofthehorse-genusthestripesareeitherplainerorappearmorecommonlyintheyoungthanintheold。Callthebreedsofpigeons，someofwhichhavebredtrueforcenturies，species。

andhowexactlyparallelisthecasewiththatofthespeciesofthehorse-genus！

Formyself，Iventureconfidentlytolookbackthousandsonthousandsofgenerations，andIseeananimalstripedlikeazebra，butperhapsotherwiseverydifferentlyconstructed，thecommonparentofourdomestichorse，whetherornotitbedescendedfromoneormorewildstocks，oftheass，thehemionus，quagga，andzebra。

Hewhobelievesthateachequinespecieswasindependentlycreated，will，Ipresume，assertthateachspecieshasbeencreatedwithatendencytovary，bothundernatureandunderdomestication，inthisparticularmanner，soasoftentobecomestripedlikeotherspeciesofthegenus。

andthateachhasbeencreatedwithastrongtendency，whencrossedwithspeciesinhabitingdistantquartersoftheworld，toproducehybridsresemblingintheirstripes，nottheirownparents，butotherspeciesofthegenus。

Toadmitthisviewis，asitseemstome，torejectarealforanunreal，oratleastforanunknown，cause。ItmakestheworksofGodameremockeryanddeception。Iwouldalmostassoonbelievewiththeoldandignorantcosmogonists，thatfossilshellshadneverlived，buthadbeencreatedinstonesoastomocktheshellsnowlivingonthesea-shore。SummaryOurignoranceofthelawsofvariationisprofound。Notinonecaseoutofahundredcanwepretendtoassignanyreasonwhythisorthatpartdiffers，moreorless，fromthesamepartintheparents。Butwheneverwehavethemeansofinstitutingacomparison，thesamelawsappeartohaveactedinproducingthelesserdifferencesbetweenvarietiesofthesamespecies，andthegreaterdifferencesbetweenspeciesofthesamegenus。

Theexternalconditionsoflife，asclimateandfood，&。c。，seemtohaveinducedsomeslightmodifications。Habitinproducingconstitutionaldifferences，anduseinstrengthening，anddisuseinweakeninganddiminishingorgans，seemtohavebeenmorepotentintheireffects。Homologouspartstendtovaryinthesameway，andhomologouspartstendtocohere。Modificationsinhardpartsandinexternalpartssometimesaffectsofterandinternalparts。Whenonepartislargelydeveloped，perhapsittendstodrawnourishmentfromtheadjoiningparts。andeverypartofthestructurewhichcanbesavedwithoutdetrimenttotheindividual，willbesaved。Changesofstructureatanearlyagewillgenerallyaffectpartssubsequentlydeveloped。andthereareverymanyothercorrelationsofgrowth，thenatureofwhichweareutterlyunabletounderstand。Multiplepartsarevariableinnumberandinstructure，perhapsarisingfromsuchpartsnothavingbeencloselyspecializedtoanyparticularfunction，sothattheirmodificationshavenotbeencloselycheckedbynaturalselection。Itisprobablyfromthissamecausethatorganicbeingslowinthescaleofnaturearemorevariablethanthosewhichhavetheirwholeorganisationmorespecialized，andarehigherinthescale。Rudimentaryorgans，frombeinguseless，willbedisregardedbynaturalselection，andhenceprobablyarevariable。Specificcharactersthatis，thecharacterswhichhavecometodiffersincetheseveralspeciesofthesamegenusbranchedofffromacommonparentaremorevariablethangenericcharacters，orthosewhichhavelongbeeninherited，andhavenotdifferedwithinthissameperiod。Intheseremarkswehavereferredtospecialpartsororgansbeingstillvariable，becausetheyhaverecentlyvariedandthuscometodiffer。butwehavealsoseeninthesecondChapterthatthesameprincipleappliestothewholeindividual。forinadistrictwheremanyspeciesofanygenusarefoundthatis，wheretherehasbeenmuchformervariationanddifferentiation，orwherethemanufactoryofnewspecificformshasbeenactivelyatworkthere，onanaverage，wenowfindmostvarietiesorincipientspecies。Secondarysexualcharactersarehighlyvariable，andsuchcharactersdiffermuchinthespeciesofthesamegroup。Variabilityinthesamepartsoftheorganisationhasgenerallybeentakenadvantageofingivingsecondarysexualdifferencestothesexesofthesamespecies，andspecificdifferencestotheseveralspeciesofthesamegenus。Anypartororgandevelopedtoanextraordinarysizeorinanextraordinarymanner，incomparisonwiththesamepartororganinthealliedspecies，musthavegonethroughanextraordinaryamountofmodificationsincethegenusarose。andthuswecanunderstandwhyitshouldoftenstillbevariableinamuchhigherdegreethanotherparts。forvariationisalong-continuedandslowprocess，andnaturalselectionwillinsuchcasesnotasyethavehadtimetoovercomethetendencytofurthervariabilityandtoreversiontoalessmodifiedstate。Butwhenaspecieswithanyextraordinarily-developedorganhasbecometheparentofmanymodifieddescendantswhichonmyviewmustbeaveryslowprocess，requiringalonglapseoftimeinthiscase，naturalselectionmayreadilyhavesucceededingivingafixedcharactertotheorgan，inhoweverextraordinaryamanneritmaybedeveloped。Speciesinheritingnearlythesameconstitutionfromacommonparentandexposedtosimilarinfluenceswillnaturallytendtopresentanalogousvariations，andthesesamespeciesmayoccasionallyreverttosomeofthecharactersoftheirancientprogenitors。Althoughnewandimportantmodificationsmaynotarisefromreversionandanalogousvariation，suchmodificationswilladdtothebeautifulandharmoniousdiversityofnature。

Whateverthecausemaybeofeachslightdifferenceintheoffspringfromtheirparentsandacauseforeachmustexistitisthesteadyaccumulation，throughnaturalselection，ofsuchdifferences，whenbeneficialtotheindividual，thatgivesrisetoallthemoreimportantmodificationsofstructure，bywhichtheinnumerablebeingsonthefaceofthisearthareenabledtostrugglewitheachother，andthebestadaptedtosurvive。PreviousChapterNextChapterTheOriginofSpecies-Chapter6TheOriginofSpeciesbyCharlesDarwinPreviousChapterNextChapterChapter6-DifficultiesonTheoryLONGbeforehavingarrivedatthispartofmywork，acrowdofdifficultieswillhaveoccurredtothereader。SomeofthemaresogravethattothisdayIcanneverreflectonthemwithoutbeingstaggered。but，tothebestofmyjudgment，thegreaternumberareonlyapparent，andthosethatarerealarenot，Ithink，fataltomytheory。

Thesedifficultiesandobjectionsmaybeclassedunderthefollowingheads：-Firstly，why，ifspecieshavedescendedfromotherspeciesbyinsensiblyfinegradations，dowenoteverywhereseeinnumerabletransitionalforms？

Whyisnotallnatureinconfusioninsteadofthespeciesbeing，asweseethem，welldefined？

Secondly，isitpossiblethatananimalhaving，forinstance，thestructureandhabitsofabat，couldhavebeenformedbythemodificationofsomeanimalwithwhollydifferenthabits？Canwebelievethatnaturalselectioncouldproduce，ontheonehand，organsoftriflingimportance，suchasthetailofagiraffe，whichservesasafly-flapper，and，ontheotherhand，organsofsuchwonderfulstructure，astheeye，ofwhichwehardlyasyetfullyunderstandtheinimitableperfection？

Thirdly，caninstinctsbeacquiredandmodifiedthroughnaturalselection？

Whatshallwesaytosomarvellousaninstinctasthatwhichleadsthebeetomakecells，whichhavepracticallyanticipatedthediscoveriesofprofoundmathematicians？

Fourthly，howcanweaccountforspecies，whencrossed，beingsterileandproducingsterileoffspring，whereas，whenvarietiesarecrossed，theirfertilityisunimpaired？

ThetwofirstheadsshallbeherediscussedInstinctandHybridisminseparatechapters。

Ontheabsenceorrarityoftransitionalvarieties。Asnaturalselectionactssolelybythepreservationofprofitablemodifications，eachnewformwilltendinafully-stockedcountrytotaketheplaceof，andfinallytoexterminate，itsownlessimprovedparentorotherless-favouredformswithwhichitcomesintocompetition。Thusextinctionandnaturalselectionwill，aswehaveseen，gohandinhand。Hence，ifwelookateachspeciesasdescendedfromsomeotherunknownform，boththeparentandallthetransitionalvarietieswillgenerallyhavebeenexterminatedbytheveryprocessofformationandperfectionofthenewform。

But，asbythistheoryinnumerabletransitionalformsmusthaveexisted，whydowenotfindthemembeddedincountlessnumbersinthecrustoftheearth？ItwillbemuchmoreconvenienttodiscussthisquestioninthechapterontheImperfectionofthegeologicalrecord。andIwillhereonlystatethatIbelievetheanswermainlyliesintherecordbeingincomparablylessperfectthanisgenerallysupposed。theimperfectionoftherecordbeingchieflyduetoorganicbeingsnotinhabitingprofounddepthsofthesea，andtotheirremainsbeingembeddedandpreservedtoafutureageonlyinmassesofsedimentsufficientlythickandextensivetowithstandanenormousamountoffuturedegradation。andsuchfossiliferousmassescanbeaccumulatedonlywheremuchsedimentisdepositedontheshallowbedofthesea，whilstitslowlysubsides。Thesecontingencieswillconcuronlyrarely，andafterenormouslylongintervals。Whilstthebedoftheseaisstationaryorisrising，orwhenverylittlesedimentisbeingdeposited，therewillbeblanksinourgeologicalhistory。Thecrustoftheearthisavastmuseum。butthenaturalcollectionshavebeenmadeonlyatintervalsoftimeimmenselyremote。

Butitmaybeurgedthatwhenseveralclosely-alliedspeciesinhabitthesameterritorywesurelyoughttofindatthepresenttimemanytransitionalforms。Letustakeasimplecase：intravellingfromnorthtosouthoveracontinent，wegenerallymeetatsuccessiveintervalswithcloselyalliedorrepresentativespecies，evidentlyfillingnearlythesameplaceinthenaturaleconomyoftheland。Theserepresentativespeciesoftenmeetandinterlock。andastheonebecomesrarerandrarer，theotherbecomesmoreandmorefrequent，tilltheonereplacestheother。Butifwecomparethesespecieswheretheyintermingle，theyaregenerallyasabsolutelydistinctfromeachotherineverydetailofstructureasarespecimenstakenfromthemetropolisinhabitedbyeach。Bymytheorythesealliedspecieshavedescendedfromacommonparent。andduringtheprocessofmodification，eachhasbecomeadaptedtotheconditionsoflifeofitsownregion，andhassupplantedandexterminateditsoriginalparentandallthetransitionalvarietiesbetweenitspastandpresentstates。Henceweoughtnottoexpectatthepresenttimetomeetwithnumeroustransitionalvarietiesineachregion，thoughtheymusthaveexistedthere，andmaybeembeddedthereinafossilcondition。Butintheintermediateregion，havingintermediateconditionsoflife，whydowenotnowfindclosely-linkingintermediatevarieties？Thisdifficultyforalongtimequiteconfoundedme。ButIthinkitcanbeinlargepartexplained。

Inthefirstplaceweshouldbeextremelycautiousininferring，becauseanareaisnowcontinuous，thatithasbeencontinuousduringalongperiod。

Geologywouldleadustobelievethatalmosteverycontinenthasbeenbrokenupintoislandsevenduringthelatertertiaryperiods。andinsuchislandsdistinctspeciesmighthavebeenseparatelyformedwithoutthepossibilityofintermediatevarietiesexistingintheintermediatezones。Bychangesintheformofthelandandofclimate，marineareasnowcontinuousmustoftenhaveexistedwithinrecenttimesinafarlesscontinuousanduniformconditionthanatpresent。ButIwillpassoverthiswayofescapingfromthedifficulty。forIbelievethatmanyperfectlydefinedspecieshavebeenformedonstrictlycontinuousareas。thoughIdonotdoubtthattheformerlybrokenconditionofareasnowcontinuoushasplayedanimportantpartintheformationofnewspecies，moreespeciallywithfreely-crossingandwanderinganimals。

Inlookingatspeciesastheyarenowdistributedoverawidearea，wegenerallyfindthemtolerablynumerousoveralargeterritory，thenbecomingsomewhatabruptlyrarerandrarerontheconfines，andfinallydisappearing。Hencetheneutralterritorybetweentworepresentativespeciesisgenerallynarrowincomparisonwiththeterritorypropertoeach。Weseethesamefactinascendingmountains，andsometimesitisquiteremarkablehowabruptly，asAlph。DeCandollehasobserved，acommonalpinespeciesdisappears。ThesamefacthasbeennoticedbyForbesinsoundingthedepthsoftheseawiththedredge。Tothosewholookatclimateandthephysicalconditionsoflifeastheall-importantelementsofdistribution，thesefactsoughttocausesurprise，asclimateandheightordepthgraduateawayinsensibly。Butwhenwebearinmindthatalmosteveryspecies，eveninitsmetropolis，wouldincreaseimmenselyinnumbers，wereitnotforothercompetingspecies。thatnearlyalleitherpreyonorserveaspreyforothers。inshort，thateachorganicbeingiseitherdirectlyorindirectlyrelatedinthemostimportantmannertootherorganicbeings，wemustseethattherangeoftheinhabitantsofanycountrybynomeansexclusivelydependsoninsensiblychangingphysicalconditions，butinlargepartonthepresenceofotherspecies，onwhichitdepends，orbywhichitisdestroyed，orwithwhichitcomesintocompetition。andasthesespeciesarealreadydefinedobjects(howevertheymayhavebecomeso)，notblendingoneintoanotherbyinsensiblegradations，therangeofanyonespecies，dependingasitdoesontherangeofothers，willtendtobesharplydefined。Moreover，eachspeciesontheconfinesofitsrange，whereitexistsinlessenednumbers，will，duringfluctuationsinthenumberofitsenemiesorofitsprey，orintheseasons，beextremelyliabletoutterextermination。andthusitsgeographicalrangewillcometobestillmoresharplydefined。

IfIamrightinbelievingthatalliedorrepresentativespecies，wheninhabitingacontinuousarea，aregenerallysodistributedthateachhasawiderange，withacomparativelynarrowneutralterritorybetweenthem，inwhichtheybecomerathersuddenlyrarerandrarer。then，asvarietiesdonotessentiallydifferfromspecies，thesamerulewillprobablyapplytoboth。andifweinimaginationadaptavaryingspeciestoaverylargearea，weshallhavetoadapttwovarietiestotwolargeareas，andathirdvarietytoanarrowintermediatezone。Theintermediatevariety，consequently，willexistinlessernumbersfrominhabitinganarrowandlesserarea。

andpractically，asfarasIcanmakeout，thisruleholdsgoodwithvarietiesinastateofnature。Ihavemetwithstrikinginstancesoftheruleinthecaseofvarietiesintermediatebetweenwell-markedvarietiesinthegenusBalanus。AnditwouldappearfrominformationgivenmebyMrWatson，DrAsaGray，andMrWollaston，thatgenerallywhenvarietiesintermediatebetweentwootherformsoccur，theyaremuchrarernumericallythantheformswhichtheyconnect。Now，ifwemaytrustthesefactsandinferences，andthereforeconcludethatvarietieslinkingtwoothervarietiestogetherhavegenerallyexistedinlessernumbersthantheformswhichtheyconnect，then，Ithink，wecanunderstandwhyintermediatevarietiesshouldnotendureforverylongperiods。whyasageneralruletheyshouldbeexterminatedanddisappear，soonerthantheformswhichtheyoriginallylinkedtogether。

Foranyformexistinginlessernumberswould，asalreadyremarked，runagreaterchanceofbeingexterminatedthanoneexistinginlargenumbers。

andinthisparticularcasetheintermediateformwouldbeeminentlyliabletotheinroadsofcloselyalliedformsexistingonbothsidesofit。Butafarmoreimportantconsideration，asIbelieve，isthat，duringtheprocessoffurthermodification，bywhichtwovarietiesaresupposedonmytheorytobeconvertedandperfectedintotwodistinctspecies，thetwowhichexistinlargernumbersfrominhabitinglargerareas，willhaveagreatadvantageovertheintermediatevariety，whichexistsinsmallernumbersinanarrowandintermediatezone。Forformsexistinginlargernumberswillalwayshaveabetterchance，withinanygivenperiod，ofpresentingfurtherfavourablevariationsfornaturalselectiontoseizeon，thanwilltherarerformswhichexistinlessernumbers。Hence，themorecommonforms，intheraceforlife，willtendtobeatandsupplantthelesscommonforms，forthesewillbemoreslowlymodifiedandimproved。Itisthesameprinciplewhich，asIbelieve，accountsforthecommonspeciesineachcountry，asshowninthesecondchapter，presentingonanaverageagreaternumberofwell-markedvarietiesthandotherarerspecies。ImayillustratewhatImeanbysupposingthreevarietiesofsheeptobekept，oneadaptedtoanextensivemountainousregion。asecondtoacomparativelynarrow，hillytract。andathirdtowideplainsatthebase。andthattheinhabitantsarealltryingwithequalsteadinessandskilltoimprovetheirstocksbyselection。thechancesinthiscasewillbestronglyinfavourofthegreatholdersonthemountainsorontheplainsimprovingtheirbreedsmorequicklythanthesmallholdersontheintermediatenarrow，hillytract。

andconsequentlytheimprovedmountainorplainbreedwillsoontaketheplaceofthelessimprovedhillbreed。andthusthetwobreeds，whichoriginallyexistedingreaternumbers，willcomeintoclosecontactwitheachother，withouttheinterpositionofthesupplanted，intermediatehill-variety。

Tosumup，Ibelievethatspeciescometobetolerablywell-definedobjects，anddonotatanyoneperiodpresentaninextricablechaosofvaryingandintermediatelinks：firstly，becausenewvarietiesareveryslowlyformed，forvariationisaveryslowprocess，andnaturalselectioncandonothinguntilfavourablevariationschancetooccur，anduntilaplaceinthenaturalpolityofthecountrycanbebetterfilledbysomemodificationofsomeoneormoreofitsinhabitants。Andsuchnewplaceswilldependonslowchangesofclimate，orontheoccasionalimmigrationofnewinhabitants，and，probably，inastillmoreimportantdegree，onsomeoftheoldinhabitantsbecomingslowlymodified，withthenewformsthusproducedandtheoldonesactingandreactingoneachother。Sothat，inanyoneregionandatanyonetime，weoughtonlytoseeafewspeciespresentingslightmodificationsofstructureinsomedegreepermanent。

andthisassuredlywedosee。

Secondly，areasnowcontinuousmustoftenhaveexistedwithintherecentperiodinisolatedportions，inwhichmanyforms，moreespeciallyamongsttheclasseswhichuniteforeachbirthandwandermuch，mayhaveseparatelybeenrenderedsufficientlydistincttorankasrepresentativespecies。

Inthiscase，intermediatevarietiesbetweentheseveralrepresentativespeciesandtheircommonparent，mustformerlyhaveexistedineachbrokenportionoftheland，buttheselinkswillhavebeensupplantedandexterminatedduringtheprocessofnaturalselection，sothattheywillnolongerexistinalivingstate。

Thirdly，whentwoormorevarietieshavebeenformedindifferentportionsofastrictlycontinuousarea，intermediatevarietieswill，itisprobable，atfirsthavebeenformedintheintermediatezones，buttheywillgenerallyhavehadashortduration。Fortheseintermediatevarietieswill，fromreasonsalreadyassigned(namelyfromwhatweknowoftheactualdistributionofcloselyalliedorrepresentativespecies，andlikewiseofacknowledgedvarieties)，existintheintermediatezonesinlessernumbersthanthevarietieswhichtheytendtoconnect。Fromthiscausealonetheintermediatevarietieswillbeliabletoaccidentalextermination。andduringtheprocessoffurthermodificationthroughnaturalselection，theywillalmostcertainlybebeatenandsupplantedbytheformswhichtheyconnect。forthesefromexistingingreaternumberswill，intheaggregate，presentmorevariation，andthusbefurtherimprovedthroughnaturalselectionandgainfurtheradvantages。

Lastly，lookingnottoanyonetime，buttoalltime，ifmytheorybetrue，numberlessintermediatevarieties，linkingmostcloselyallthespeciesofthesamegrouptogether，mustassuredlyhaveexisted。buttheveryprocessofnaturalselectionconstantlytends，ashasbeensooftenremarked，toexterminatetheparentformsandtheintermediatelinks。Consequentlyevidenceoftheirformerexistencecouldbefoundonlyamongstfossilremains，whicharepreserved，asweshallinafuturechapterattempttoshow，inanextremelyimperfectandintermittentrecord。

Ontheoriginandtransitionsoforganicbeingswithpeculiarhabitsandstructure。IthasbeenaskedbytheopponentsofsuchviewsasIhold，how，forinstance，alandcarnivorousanimalcouldhavebeenconvertedintoonewithaquatichabits。forhowcouldtheanimalinitstransitionalstatehavesubsisted？Itwouldbeeasytoshowthatwithinthesamegroupcarnivorousanimalsexisthavingeveryintermediategradebetweentrulyaquaticandstrictlyterrestrialhabits。andaseachexistsbyastruggleforlife，itisclearthateachiswelladaptedinitshabitstoitsplaceinnature。LookattheMustelavisonofNorthAmerica，whichhaswebbedfeetandwhichresemblesanotterinitsfur，shortlegs，andformoftail。

duringsummerthisanimaldivesforandpreysonfish，butduringthelongwinteritleavesthefrozenwaters，andpreyslikeotherpolecatsonmiceandlandanimals。Ifadifferentcasehadbeentaken，andithadbeenaskedhowaninsectivorousquadrupedcouldpossiblyhavebeenconvertedintoaflyingbat，thequestionwouldhavebeenfarmoredifficult，andIcouldhavegivennoanswer。YetIthinksuchdifficultieshaveverylittleweight。

Here，asonotheroccasions，Ilieunderaheavydisadvantage，foroutofthemanystrikingcaseswhichIhavecollected，Icangiveonlyoneortwoinstancesoftransitionalhabitsandstructuresincloselyalliedspeciesofthesamegenus。andofdiversifiedhabits，eitherconstantoroccasional，inthesamespecies。Anditseemstomethatnothinglessthanalonglistofsuchcasesissufficienttolessenthedifficultyinanyparticularcaselikethatofthebat。

Lookatthefamilyofsquirrels。herewehavethefinestgradationfromanimalswiththeirtailsonlyslightlyflattened，andfromothers，asSirJ。Richardsonhasremarked，withtheposteriorpartoftheirbodiesratherwideandwiththeskinontheirflanksratherfull，totheso-calledflyingsquirrels。andflyingsquirrelshavetheirlimbsandeventhebaseofthetailunitedbyabroadexpanseofskin，whichservesasaparachuteandallowsthemtoglidethroughtheairtoanastonishingdistancefromtreetotree。Wecannotdoubtthateachstructureisofusetoeachkindofsquirrelinitsowncountry，byenablingittoescapebirdsorbeastsofprey，ortocollectfoodmorequickly，or，asthereisreasontobelieve，bylesseningthedangerfromoccasionalfalls。Butitdoesnotfollowfromthisfactthatthestructureofeachsquirrelisthebestthatitispossibletoconceiveunderallnaturalconditions。Lettheclimateandvegetationchange，letothercompetingrodentsornewbeastsofpreyimmigrate，oroldonesbecomemodified，andallanalogywouldleadustobelievethatsomeatleastofthesquirrelswoulddecreaseinnumbersorbecomeexterminated，unlesstheyalsobecamemodifiedandimprovedinstructureinacorrespondingmanner。Therefore，Icanseenodifficulty，moreespeciallyunderchangingconditionsoflife，inthecontinuedpreservationofindividualswithfullerandfullerflank-membranes，eachmodificationbeinguseful，eachbeingpropagated，untilbytheaccumulatedeffectsofthisprocessofnaturalselection，aperfectso-calledflyingsquirrelwasproduced。

NowlookattheGaleopithecusorflyinglemur，whichformerlywasfalselyrankedamongstbats。Ithasanextremelywideflank-membrane，stretchingfromthecornersofthejawtothetail，andincludingthelimbsandtheelongatedfingers：theflankmembraneis，also，furnishedwithanextensormuscle。Althoughnograduatedlinksofstructure，fittedforglidingthroughtheair，nowconnecttheGaleopithecuswiththeotherLemuridae，yetI

canseenodifficultyinsupposingthatsuchlinksformerlyexisted，andthateachhadbeenformedbythesamestepsasinthecaseofthelessperfectlyglidingsquirrels。andthateachgradeofstructurehadbeenusefultoitspossessor。NorcanIseeanyinsuperabledifficultyinfurtherbelievingitpossiblethatthemembrane-connectedfingersandfore-armoftheGaleopithecusmightbegreatlylengthenedbynaturalselection。

andthis，asfarastheorgansofflightareconcerned，wouldconvertitintoabat。Inbatswhichhavethewing-membraneextendedfromthetopoftheshouldertothetail，includingthehind-legs，weperhapsseetracesofanapparatusoriginallyconstructedforglidingthroughtheairratherthanforflight。

Ifaboutadozengeneraofbirdshadbecomeextinctorwereunknown，whowouldhaveventuredtohavesurmisedthatbirdsmighthaveexistedwhichusedtheirwingssolelyasflappers，likethelogger-headedduck(MicropterusofEyton)。asfinsinthewaterandfrontlegsontheland，likethepenguin。assails，liketheostrich。andfunctionallyfornopurpose，liketheApteryx。Yetthestructureofeachofthesebirdsisgoodforit，undertheconditionsoflifetowhichitisexposed，foreachhastolivebyastruggle。butitisnotnecessarilythebestpossibleunderallpossibleconditions。Itmustnotbeinferredfromtheseremarksthatanyofthegradesofwing-structureherealludedto，whichperhapsmayallhaveresultedfromdisuse，indicatethenaturalstepsbywhichbirdshaveacquiredtheirperfectpowerofflight。buttheyserve，atleast，toshowwhatdiversifiedmeansoftransitionarepossible。

Seeingthatafewmembersofsuchwater-breathingclassesastheCrustaceaandMolluscaareadaptedtoliveontheland，andseeingthatwehaveflyingbirdsandmammals，flyinginsectsofthemostdiversifiedtypes，andformerlyhadflyingreptiles，itisconceivablethatflying-fish，whichnowglidefarthroughtheair，slightlyrisingandturningbytheaidoftheirflutteringfins，mighthavebeenmodifiedintoperfectlywingedanimals。Ifearlytransitionalstatetheyhadbeeninhabitantsoftheopenocean，andhadusedtheirincipientorgansofflightexclusively，asfarasweknow，toescapebeingdevouredbyotherfish？

Whenweseeanystructurehighlyperfectedforanyparticularhabit，asthewingsofabirdforflight，weshouldbearinmindthatanimalsdisplayingearlytransitionalgradesofthestructurewillseldomcontinuetoexisttothepresentday，fortheywillhavebeensupplantedbytheveryprocessofperfectionthroughnaturalselection。Furthermore，wemayconcludethattransitionalgradesbetweenstructuresfittedforverydifferenthabitsoflifewillrarelyhavebeendevelopedatanearlyperiodingreatnumbersandundermanysubordinateforms。Thus，toreturntoourimaginaryillustrationoftheflying-fish，itdoesnotseemprobablethatfishescapableoftrueflightwouldhavebeendevelopedundermanysubordinateforms，fortakingpreyofmanykindsinmanyways，onthelandandinthewater，untiltheirorgansofflighthadcometoahighstageofperfection，soastohavegiventhemadecidedadvantageoverotheranimalsinthebattleforlife。Hencethechanceofdiscoveringspecieswithtransitionalgradesofstructureinafossilconditionwillalwaysbeless，fromtheirhavingexistedinlessernumbers，thaninthecaseofspecieswithfullydevelopedstructures。

Iwillnowgivetwoorthreeinstancesofdiversifiedandofchangedhabitsintheindividualsofthesamespecies。Wheneithercaseoccurs，itwouldbeeasyfornaturalselectiontofittheanimal，bysomemodificationofitsstructure，foritschangedhabits，orexclusivelyforoneofitsseveraldifferenthabits。Butitisdifficulttotell，andimmaterialforus，whetherhabitsgenerallychangefirstandstructureafterwards。orwhetherslightmodificationsofstructureleadtochangedhabits。bothprobablyoftenchangealmostsimultaneously。OfcasesofchangedhabitsitwillsufficemerelytoalludetothatofthemanyBritishinsectswhichnowfeedonexoticplants，orexclusivelyonartificialsubstances。Ofdiversifiedhabitsinnumerableinstancescouldbegiven：Ihaveoftenwatchedatyrantflycatcher(Saurophagussulphuratus)inSouthAmerica，hoveringoveronespotandthenproceedingtoanother，likeakestrel，andatothertimesstandingstationaryonthemarginofwater，andthendashinglikeakingfisheratafish。Inourowncountrythelargertitmouse(Parusmajor)

maybeseenclimbingbranches，almostlikeacreeper。itoften，likeashrike，killssmallbirdsbyblowsonthehead。andIhavemanytimesseenandheardithammeringtheseedsoftheyewonabranch，andthusbreakingthemlikeanuthatch。InNorthAmericatheblackbearwasseenbyHearneswimmingforhourswithwidelyopenmouth，thuscatching，likeawhale，insectsinthewater。Eveninsoextremeacaseasthis，ifthesupplyofinsectswereconstant，andifbetteradaptedcompetitorsdidnotalreadyexistinthecountry，Icanseenodifficultyinaraceofbearsbeingrendered，bynaturalselection，moreandmoreaquaticintheirstructureandhabits，withlargerandlargermouths，tillacreaturewasproducedasmonstrousasawhale。

Aswesometimesseeindividualsofaspeciesfollowinghabitswidelydifferentfromthosebothoftheirownspeciesandoftheotherspeciesofthesamegenus，wemightexpect，onmytheory，thatsuchindividualswouldoccasionallyhavegivenrisetonewspecies，havinganomaloushabits，andwiththeirstructureeitherslightlyorconsiderablymodifiedfromthatoftheirpropertype。Andsuchinstancesdooccurinnature。Canamorestrikinginstanceofadaptationbegiventhanthatofawoodpeckerforclimbingtreesandforseizinginsectsinthechinksofthebark？YetinNorthAmericatherearewoodpeckerswhichfeedlargelyonfruit，andotherswithelongatedwingswhichchaseinsectsonthewing。andontheplainsofLaPlata，wherenotatreegrows，thereisawoodpecker，whichineveryessentialpartofitsorganisation，eveninitscolouring，intheharshtoneofitsvoice，andundulatoryflight，toldmeplainlyofitscloseblood-relationshiptoourcommonspecies。yetitisawoodpeckerwhichneverclimbsatree！

Petrelsarethemostaë。rialandoceanicofbirds，yetinthequietSoundsofTierradelFuego，thePuffinuriaberardi，initsgeneralhabits，initsastonishingpowerofdiving，itsmannerofswimming，andofflyingwhenunwillinglyittakesflight，wouldbemistakenbyanyoneforanaukorgrebe。nevertheless，itisessentiallyapetrel，butwithmanypartsofitsorganisationprofoundlymodified。Ontheotherhand，theacutestobserverbyexaminingthedeadbodyofthewater-ouzelwouldneverhavesuspecteditssub-aquatichabits。yetthisanomalousmemberofthestrictlyterrestrialthrushfamilywhollysubsistsbydiving，graspingthestoneswithitsfeetandusingitswingsunderwater。

Hewhobelievesthateachbeinghasbeencreatedaswenowseeit，mustoccasionallyhavefeltsurprisewhenhehasmetwithananimalhavinghabitsandstructurenotatallinagreement。Whatcanbeplainerthanthatthewebbedfeetofducksandgeeseareformedforswimming。yetthereareuplandgeesewithwebbedfeetwhichrarelyornevergonearthewater。andnooneexceptAudubonhasseenthefrigate-bird，whichhasallitsfourtoeswebbed，alightonthesurfaceofthesea。Ontheotherhand，grebesandcootsareeminentlyaquatic，althoughtheirtoesareonlyborderedbymembrane。

Whatseemsplainerthanthatthelongtoesofgrallatoresareformedforwalkingoverswampsandfloatingplants，yetthewater-henisnearlyasaquaticasthecoot。andthelandrailnearlyasterrestrialasthequailorpartridge。Insuchcases，andmanyotherscouldbegiven，habitshavechangedwithoutacorrespondingchangeofstructure。Thewebbedfeetoftheuplandgoosemaybesaidtohavebecomerudimentaryinfunction，thoughnotinstructure。Inthefrigate-bird，thedeeply-scoopedmembranebetweenthetoesshowsthatstructurehasbeguntochange。

Hewhobelievesinseparateandinnumerableactsofcreationwillsay，thatinthesecasesithaspleasedtheCreatortocauseabeingofonetypetotaketheplaceofoneofanothertype。butthisseemstomeonlyrestatingthefactindignifiedlanguage。Hewhobelievesinthestruggleforexistenceandintheprincipleofnaturalselection，willacknowledgethateveryorganicbeingisconstantlyendeavouringtoincreaseinnumbers。

andthatifanyonebeingvaryeversolittle，eitherinhabitsorstructure，andthusgainanadvantageoversomeotherinhabitantofthecountry，itwillseizeontheplaceofthatinhabitant，howeverdifferentitmaybefromitsownplace。Henceitwillcausehimnosurprisethatthereshouldbegeeseandfrigate-birdswithwebbedfeet，eitherlivingonthedrylandormostrarelyalightingonthewater。thatthereshouldbelong-toedcorncrakeslivinginmeadowsinsteadofinswamps。thatthereshouldbewoodpeckerswherenotatreegrows。thatthereshouldbedivingthrushes，andpetrelswiththehabitsofauks。

Organsofextremeperfectionandcomplication。Tosupposethattheeye，withallitsinimitablecontrivancesforadjustingthefocustodifferentdistances，foradmittingdifferentamountsoflight，andforthecorrectionofsphericalandchromaticaberration，couldhavebeenformedbynaturalselection，seems，Ifreelyconfess，absurdinthehighestpossibledegree。Yetreasontellsme，thatifnumerousgradationsfromaperfectandcomplexeyetooneveryimperfectandsimple，eachgradebeingusefultoitspossessor，canbeshowntoexist。iffurther，theeyedoesvaryeversoslightly，andthevariationsbeinherited，whichiscertainlythecase。andifanyvariationormodificationintheorganbeeverusefultoananimalunderchangingconditionsoflife，thenthedifficultyofbelievingthataperfectandcomplexeyecouldbeformedbynaturalselection，thoughinsuperablebyourimagination，canhardlybeconsideredreal。Howanervecomestobesensitivetolight，hardlyconcernsusmorethanhowlifeitselffirstoriginated。butImayremarkthatseveralfactsmakemesuspectthatanysensitivenervemayberenderedsensitivetolight，andlikewisetothosecoarservibrationsoftheairwhichproducesound。

Inlookingforthegradationsbywhichanorganinanyspecieshasbeenperfected，weoughttolookexclusivelytoitslinealancestors。butthisisscarcelyeverpossible，andweareforcedineachcasetolooktospeciesofthesamegroup，thatistothecollateraldescendantsfromthesameoriginalparent-form，inordertoseewhatgradationsarepossible，andforthechanceofsomegradationshavingbeentransmittedfromtheearlierstagesofdescent，inanunalteredorlittlealteredcondition。AmongstexistingVertebrata，wefindbutasmallamountofgradationinthestructureoftheeye，andfromfossilspecieswecanlearnnothingonthishead。

Inthisgreatclassweshouldprobablyhavetodescendfarbeneaththelowestknownfossiliferousstratumtodiscovertheearlierstages，bywhichtheeyehasbeenperfected。

IntheArticulatawecancommenceaserieswithanopticnervemerelycoatedwithpigment，andwithoutanyothermechanism。andfromthislowstage，numerousgradationsofstructure，branchingoffintwofundamentallydifferentlines，canbeshowntoexist，untilwereachamoderatelyhighstageofperfection。Incertaincrustaceans，forinstance，thereisadoublecornea，theinneronedividedintofacets，withineachofwhichthereisalensshapedswelling。Inothercrustaceansthetransparentconeswhicharecoatedbypigment，andwhichproperlyactonlybyexcludinglateralpencilsoflight，areconvexattheirupperendsandmustactbyconvergence。

andattheirlowerendsthereseemstobeanimperfectvitreoussubstance。

Withthesefacts，herefartoobrieflyandimperfectlygiven，whichshowthatthereismuchgraduateddiversityintheeyesoflivingcrustaceans，andbearinginmindhowsmallthenumberoflivinganimalsisinproportiontothosewhichhavebecomeextinct，Icanseenoverygreatdifficulty(notmorethaninthecaseofmanyotherstructures)inbelievingthatnaturalselectionhasconvertedthesimpleapparatusofanopticnervemerelycoatedwithpigmentandinvestedbytransparentmembrane，intoanopticalinstrumentasperfectasispossessedbyanymemberofthegreatArticulateclass。

Hewhowillgothusfar，ifhefindonfinishingthistreatisethatlargebodiesoffacts，otherwiseinexplicable，canbeexplainedbythetheoryofdescent，oughtnottohesitatetogofurther，andtoadmitthatastructureevenasperfectastheeyeofaneaglemightbeformedbynaturalselection，althoughinthiscasehedoesnotknowanyofthetransitionalgrades。Hisreasonoughttoconquerhisimagination。thoughIhavefeltthedifficultyfartookeenlytobesurprisedatanydegreeofhesitationinextendingtheprincipleofnaturalselectiontosuchstartlinglengths。

Itisscarcelypossibletoavoidcomparingtheeyetoatelescope。Weknowthatthisinstrumenthasbeenperfectedbythelong-continuedeffortsofthehighesthumanintellects。andwenaturallyinferthattheeyehasbeenformedbyasomewhatanalogousprocess。Butmaynotthisinferencebepresumptuous？HaveweanyrighttoassumethattheCreatorworksbyintellectualpowerslikethoseofman？Ifwemustcomparetheeyetoanopticalinstrument，weoughtinimaginationtotakeathicklayeroftransparenttissue，withanervesensitivetolightbeneath，andthensupposeeverypartofthislayertobecontinuallychangingslowlyindensity，soastoseparateintolayersofdifferentdensitiesandthicknesses，placedatdifferentdistancesfromeachother，andwiththesurfacesofeachlayerslowlychanginginform。Furtherwemustsupposethatthereisapoweralwaysintentlywatchingeachslightaccidentalalterationinthetransparentlayers。andcarefullyselectingeachalterationwhich，undervariedcircumstances，mayinanyway，orinanydegree，tendtoproduceadistincterimage。Wemustsupposeeachnewstateoftheinstrumenttobemultipliedbythemillion。

andeachtobepreservedtillabetterbeproduced，andthentheoldonestobedestroyed。Inlivingbodies，variationwillcausetheslightalterations，generationwillmultiplythemalmostinfinitely，andnaturalselectionwillpickoutwithunerringskilleachimprovement。Letthisprocessgoonformillionsonmillionsofyears。andduringeachyearonmillionsofindividualsofmanykinds。andmaywenotbelievethatalivingopticalinstrumentmightthusbeformedassuperiortooneofglass，astheworksoftheCreatoraretothoseofman？

Ifitcouldbedemonstratedthatanycomplexorganexisted，whichcouldnotpossiblyhavebeenformedbynumerous，successive，slightmodifications，mytheorywouldabsolutelybreakdown。ButIcanfindoutnosuchcase。

Nodoubtmanyorgansexistofwhichwedonotknowthetransitionalgrades，moreespeciallyifwelooktomuch-isolatedspecies，roundwhich，accordingtomytheory，therehasbeenmuchextinction。Oragain，ifwelooktoanorgancommontoallthemembersofalargeclass，forinthislattercasetheorganmusthavebeenfirstformedatanextremelyremoteperiod，sincewhichallthemanymembersoftheclasshavebeendeveloped。andinordertodiscovertheearlytransitionalgradesthroughwhichtheorganhaspassed，weshouldhavetolooktoveryancientancestralforms，longsincebecomeextinct。

Weshouldbeextremelycautiousinconcludingthatanorgancouldnothavebeenformedbytransitionalgradationsofsomekind。Numerouscasescouldbegivenamongsttheloweranimalsofthesameorganperformingatthesametimewhollydistinctfunctions。thusthealimentarycanalrespires，digests，andexcretesinthelarvaofthedragon-flyandinthefishCobites。

IntheHydra，theanimalmaybeturnedinsideout，andtheexteriorsurfacewillthendigestandthestomachrespire。Insuchcasesnaturalselectionmighteasilyspecialise，ifanyadvantagewerethusgained，apartororgan，whichhadperformedtwofunctions，foronefunctionalone，andthuswhollychangeitsnaturebyinsensiblesteps。Twodistinctorganssometimesperformsimultaneouslythesamefunctioninthesameindividual。togiveoneinstance，therearefishwithgillsorbranchiaethatbreathetheairdissolvedinthewater，atthesametimethattheybreathefreeairintheirswimbladders，thislatterorganhavingaductuspneumaticusforitssupply，andbeingdividedbyhighlyvascularpartitions。Inthesecases，oneofthetwoorgansmightwitheasebemodifiedandperfectedsoastoperformalltheworkbyitself，beingaidedduringtheprocessofmodificationbytheotherorgan。andthenthisotherorganmightbemodifiedforsomeotherandquitedistinctpurpose，orbequiteobliterated。

Theillustrationoftheswimbladderinfishesisagoodone，becauseitshowsusclearlythehighlyimportantfactthatanorganoriginallyconstructedforonepurpose，namelyflotation，maybeconvertedintooneforawhollydifferentpurpose，namelyrespiration。Theswimbladderhas，also，beenworkedinasanaccessorytotheauditoryorgansofcertainfish，or，forIdonotknowwhichviewisnowgenerallyheld，apartoftheauditoryapparatushasbeenworkedinasacomplementtotheswimbladder。

Allphysiologistsadmitthattheswimbladderishomologous，or’ideallysimilar，’inpositionandstructurewiththelungsofthehighervertebrateanimals：hencethereseemstometobenogreatdifficultyinbelievingthatnaturalselectionhasactuallyconvertedaswimbladderintoalung，ororganusedexclusivelyforrespiration。

Ican，indeed，hardlydoubtthatallvertebrateanimalshavingtruelungshavedescendedbyordinarygenerationfromanancientprototype，ofwhichweknownothing，furnishedwithafloatingapparatusorswimbladder。

Wecanthus，asIinferfromProfessorOwen’sinterestingdescriptionoftheseparts，understandthestrangefactthateveryparticleoffoodanddrinkwhichweswallowhastopassovertheorificeofthetrachea，withsomeriskoffallingintothelungs，notwithstandingthebeautifulcontrivancebywhichtheglottisisclosed。InthehigherVertebratathebranchiaehavewhollydisappearedtheslitsonthesidesoftheneckandtheloop-likecourseofthearteriesstillmarkingintheembryotheirformerposition。

Butitisconceivablethatthenowutterlylostbranchiaemighthavebeengraduallyworkedinbynaturalselectionforsomequitedistinctpurpose：

inthesamemanneras，ontheviewentertainedbysomenaturaliststhatthebranchiaeanddorsalscalesofAnnelidsarehomologouswiththewingsandwing-coversofinsects，itisprobablethatorganswhichataveryancientperiodservedforrespirationhavebeenactuallyconvertedintoorgansofflight。

Inconsideringtransitionsoforgans，itissoimportanttobearinmindtheprobabilityofconversionfromonefunctiontoanother，thatI

willgiveonemoreinstance。Pedunculatedcirripedeshavetwominutefoldsofskin，calledbymetheovigerousfrena，whichserve，throughthemeansofastickysecretion，toretaintheeggsuntiltheyarehatchedwithinthesack。Thesecirripedeshavenobranchiae，thewholesurfaceofthebodyandsack，includingthesmallfrena，servingforrespiration。TheBalanidaeorsessilecirripedes，ontheotherhand，havenoovigerousfrena，theeggslyinglooseatthebottomofthesack，inthewell-enclosedshell。

buttheyhavelargefoldedbranchiae。NowIthinknoonewilldisputethattheovigerousfrenaintheonefamilyarestrictlyhomologouswiththebranchiaeoftheotherfamily。indeed，theygraduateintoeachother。ThereforeIdonotdoubtthatlittlefoldsofskin，whichoriginallyservedasovigerousfrena，butwhich，likewise，veryslightlyaidedtheactofrespiration，havebeengraduallyconvertedbynaturalselectionintobranchiae，simplythroughanincreaseintheirsizeandtheobliterationoftheiradhesiveglands。Ifallpedunculatedcirripedeshadbecomeextinct，andtheyhavealreadysufferedfarmoreextinctionthanhavesessilecirripedes，whowouldeverhaveimaginedthatthebranchiaeinthislatterfamilyhadoriginallyexistedasorgansforpreventingtheovafrombeingwashedoutofthesack？

Althoughwemustbeextremelycautiousinconcludingthatanyorgancouldnotpossiblyhavebeenproducedbysuccessivetransitionalgradations，yet，undoubtedly，gravecasesofdifficultyoccur，someofwhichwillbediscussedinmyfuturework。

Oneofthegravestisthatofneuterinsects，whichareoftenverydifferentlyconstructedfromeitherthemalesorfertilefemales。butthiscasewillbetreatedofinthenextchapter。Theelectricorgansoffishesofferanothercaseofspecialdifficulty。itisimpossibletoconceivebywhatstepsthesewondrousorganshavebeenproduced。but，asOwenandothershaveremarked，theirintimatestructurecloselyresemblesthatofcommonmuscle。andasithaslatelybeenshownthatRayshaveanorgancloselyanalogoustotheelectricapparatus，andyetdonot，asMatteuchiasserts，dischargeanyelectricity，wemustownthatwearefartooignoranttoarguethatnotransitionofanykindispossible。

Theelectricorgansofferanotherandevenmoreseriousdifficulty。

fortheyoccurinonlyaboutadozenfishes，ofwhichseveralarewidelyremoteintheiraffinities。Generallywhenthesameorganappearsinseveralmembersofthesameclass，especiallyifinmembershavingverydifferenthabitsoflife，wemayattributeitspresencetoinheritancefromacommonancestor。anditsabsenceinsomeofthememberstoitslossthroughdisuseornaturalselection。Butiftheelectricorganshadbeeninheritedfromoneancientprogenitorthusprovided，wemighthaveexpectedthatallelectricfisheswouldhavebeenspeciallyrelatedtoeachother。Nordoesgeologyatallleadtothebeliefthatformerlymostfisheshadelectricorgans，whichmostoftheirmodifieddescendantshavelost。Thepresenceofluminousorgansinafewinsects，belongingtodifferentfamiliesandorders，offersaparallelcaseofdifficulty。Othercasescouldbegiven。forinstanceinplants，theverycuriouscontrivanceofamassofpollen-grains，borneonafoot-stalkwithastickyglandattheend，isthesameinOrchisandAsclepias，generaalmostasremoteaspossibleamongstfloweringplants。

Inallthesecasesoftwoverydistinctspeciesfurnishedwithapparentlythesameanomalousorgan，itshouldbeobservedthat，althoughthegeneralappearanceandfunctionoftheorganmaybethesame，yetsomefundamentaldifferencecangenerallybedetected。Iaminclinedtobelievethatinnearlythesamewayastwomenhavesometimesindependentlyhitontheverysameinvention，sonaturalselection，workingforthegoodofeachbeingandtakingadvantageofanalogousvariations，hassometimesmodifiedinverynearlythesamemannertwopartsintwoorganicbeings，whichowebutlittleoftheirstructureincommontoinheritancefromthesameancestor。

Althoughinmanycasesitismostdifficulttoconjecturebywhattransitionsanorgancouldhavearrivedatitspresentstate。yet，consideringthattheproportionoflivingandknownformstotheextinctandunknownisverysmall，Ihavebeenastonishedhowrarelyanorgancanbenamed，towardswhichnotransitionalgradeisknowntolead。Thetruthofthisremarkisindeedshownbythatoldcanoninnaturalhistoryof’Naturanonfacitsaltum。’Wemeetwiththisadmissioninthewritingsofalmosteveryexperiencednaturalist。or，asMilneEdwardshaswellexpressedit，natureisprodigalinvariety，butniggardininnovation。Why，onthetheoryofCreation，shouldthisbeso？Whyshouldallthepartsandorgansofmanyindependentbeings，eachsupposedtohavebeenseparatelycreatedforitsproperplaceinnature，besoinvariablylinkedtogetherbygraduatedsteps？WhyshouldnotNaturehavetakenaleapfromstructuretostructure？Onthetheoryofnaturalselection，wecanclearlyunderstandwhysheshouldnot。fornaturalselectioncanactonlybytakingadvantageofslightsuccessivevariations。shecannevertakealeap，butmustadvancebytheshortestandsloweststeps。

Organsoflittleapparentimportance。Asnaturalselectionactsbylifeanddeath，bythepreservationofindividualswithanyfavourablevariation，andbythedestructionofthosewithanyunfavourabledeviationofstructure，Ihavesometimesfeltmuchdifficultyinunderstandingtheoriginofsimpleparts，ofwhichtheimportancedoesnotseemsufficienttocausethepreservationofsuccessivelyvaryingindividuals。Ihavesometimesfeltasmuchdifficulty，thoughofaverydifferentkind，onthishead，asinthecaseofanorganasperfectandcomplexastheeye。

Inthefirstplace，wearemuchtooignorantinregardtothewholeeconomyofanyoneorganicbeing，tosaywhatslightmodificationswouldbeofimportanceornot。InaformerchapterIhavegiveninstancesofmosttriflingcharacters，suchasthedownonfruitandthecolouroftheflesh，which，fromdeterminingtheattacksofinsectsorfrombeingcorrelatedwithconstitutionaldifferences，mightassuredlybeactedonbynaturalselection。Thetailofthegiraffelookslikeanartificiallyconstructedfly-flapper。anditseemsatfirstincrediblethatthiscouldhavebeenadaptedforitspresentpurposebysuccessiveslightmodifications，eachbetterandbetter，forsotriflinganobjectasdrivingawayflies。yetweshouldpausebeforebeingtoopositiveeveninthiscase，forweknowthatthedistributionandexistenceofcattleandotheranimalsinSouthAmericaabsolutelydependsontheirpowerofresistingtheattacksofinsects：

sothatindividualswhichcouldbyanymeansdefendthemselvesfromthesesmallenemies，wouldbeabletorangeintonewpasturesandthusgainagreatadvantage。Itisnotthatthelargerquadrupedsareactuallydestroyed(exceptinsomerarecases)bytheflies，buttheyareincessantlyharassedandtheirstrengthreduced，sothattheyaremoresubjecttodisease，ornotsowellenabledinacomingdearthtosearchforfood，ortoescapefrombeastsofprey。

Organsnowoftriflingimportancehaveprobablyinsomecasesbeenofhighimportancetoanearlyprogenitor，and，afterhavingbeenslowlyperfectedataformerperiod，havebeentransmittedinnearlythesamestate，althoughnowbecomeofveryslightuse。andanyactuallyinjuriousdeviationsintheirstructurewillalwayshavebeencheckedbynaturalselection。Seeinghowimportantanorganoflocomotionthetailisinmostaquaticanimals，itsgeneralpresenceanduseformanypurposesinsomanylandanimals，whichintheirlungsormodifiedswim-bladdersbetraytheiraquaticorigin，mayperhapsbethusaccountedfor。Awell-developedtailhavingbeenformedinanaquaticanimal，itmightsubsequentlycometobeworkedinforallsortsofpurposes，asafly-flapper，anorganofprehension，orasanaidinturning，aswiththedog，thoughtheaidmustbeslight，forthehare，withhardlyanytail，candoublequicklyenough。

Inthesecondplace，wemaysometimesattributeimportancetocharacterswhicharereallyofverylittleimportance，andwhichhaveoriginatedfromquitesecondarycauses，independentlyofnaturalselection。Weshouldrememberthatclimate，food，&。c。，probablyhavesomelittledirectinfluenceontheorganisation。thatcharactersreappearfromthelawofreversion。，thatcorrelationofgrowthwillhavehadamostimportantinfluenceinmodifyingvariousstructures。andfinally，thatsexualselectionwilloftenhavelargelymodifiedtheexternalcharactersofanimalshavingawill，togiveonemaleanadvantageinfightingwithanotherorincharmingthefemales。Moreoverwhenamodificationofstructurehasprimarilyarisenfromtheaboveorotherunknowncauses，itmayatfirsthavebeenofnoadvantagetothespecies，butmaysubsequentlyhavebeentakenadvantageofbythedescendantsofthespeciesundernewconditionsoflifeandwithnewlyacquiredhabits。

Togiveafewinstancestoillustratetheselatterremarks。Ifgreenwoodpeckersalonehadexisted，andwedidnotknowthatthereweremanyblackandpiedkinds，Idaresaythatweshouldhavethoughtthatthegreencolourwasabeautifuladaptationtohidethistree-frequentingbirdfromitsenemies。andconsequentlythatitwasacharacterofimportanceandmighthavebeenacquiredthroughnaturalselection。asitis，Ihavenodoubtthatthecolourisduetosomequitedistinctcause，probablytosexualselection。AtrailingbamboointheMalayArchipelagoclimbstheloftiesttreesbytheaidofexquisitelyconstructedhooksclusteredaroundtheendsofthebranches，andthiscontrivance，nodoubt，isofthehighestservicetotheplant。butasweseenearlysimilarhooksonmanytreeswhicharenotclimbersthehooksonthebamboomayhavearisenfromunknownlawsofgrowth，andhavebeensubsequentlytakenadvantageofbytheplantundergoingfurthermodificationandbecomingaclimber。Thenakedskinontheheadofavultureisgenerallylookedatasadirectadaptationforwallowinginputridity。andsoitmaybe，oritmaypossiblybeduetothedirectactionofputridmatter。butweshouldbeverycautiousindrawinganysuchinference，whenweseethattheskinontheheadoftheclean-feedingmaleturkeyislikewisenaked。Thesuturesintheskullsofyoungmammalshavebeenadvancedasabeautifuladaptationforaidingparturition，andnodoubttheyfacilitate，ormaybeindispensableforthisact。butassuturesoccurintheskullsofyoungbirdsandreptiles，whichhaveonlytoescapefromabrokenegg，wemayinferthatthisstructurehasarisenfromthelawsofgrowth，andhasbeentakenadvantageofintheparturitionofthehigheranimals。

Weareprofoundlyignorantofthecausesproducingslightandunimportantvariations。andweareimmediatelymadeconsciousofthisbyreflectingonthedifferencesinthebreedsofourdomesticatedanimalsindifferentcountries，moreespeciallyinthelesscivilizedcountrieswheretherehasbeenbutlittleartificialselection。Carefulobserversareconvincedthatadampclimateaffectsthegrowthofthehair，andthatwiththehairthehornsarecorrelated。Mountainbreedsalwaysdifferfromlowlandbreeds。

andamountainouscountrywouldprobablyaffectthehindlimbsfromexercisingthemmore，andpossiblyeventheformofthepelvis。andthenbythelawofhomologousvariation，thefrontlimbsandeventheheadwouldprobablybeaffected。Theshape，also，ofthepelvismightaffectbypressuretheshapeoftheheadoftheyounginthewomb。Thelaboriousbreathingnecessaryinhighregionswould，wehavesomereasontobelieve，increasethesizeofthechest。andagaincorrelationwouldcomeintoplay。Animalskeptbysavagesindifferentcountriesoftenhavetostrugglefortheirownsubsistence，andwouldbeexposedtoacertainextenttonaturalselection，andindividualswithslightlydifferentconstitutionswouldsucceedbestunderdifferentclimates。andthereisreasontobelievethatconstitutionandcolourarecorrelated。Agoodobserver，also，statesthatincattlesusceptibilitytotheattacksoffliesiscorrelatedwithcolour，asistheliabilitytobepoisonedbycertainplants。sothatcolourwouldbethussubjectedtotheactionofnaturalselection。Butwearefartooignoranttospeculateontherelativeimportanceoftheseveralknownandunknownlawsofvariation。andIhaveherealludedtothemonlytoshowthat，ifweareunabletoaccountforthecharacteristicdifferencesofourdomesticbreeds，whichneverthelesswegenerallyadmittohavearisenthroughordinarygeneration，weoughtnottolaytoomuchstressonourignoranceoftheprecisecauseoftheslightanalogousdifferencesbetweenspecies。Imighthaveadducedforthissamepurposethedifferencesbetweentheracesofman，whicharesostronglymarked。Imayaddthatsomelittlelightcanapparentlybethrownontheoriginofthesedifferences，chieflythroughsexualselectionofaparticularkind，butwithouthereenteringoncopiousdetailsmyreasoningwouldappearfrivolous。

Theforegoingremarksleadmetosayafewwordsontheprotestlatelymadebysomenaturalists，againsttheutilitariandoctrinethateverydetailofstructurehasbeenproducedforthegoodofitspossessor。Theybelievethatverymanystructureshavebeencreatedforbeautyintheeyesofman，orformerevariety。Thisdoctrine，iftrue，wouldbeabsolutelyfataltomytheory。YetIfullyadmitthatmanystructuresareofnodirectusetotheirpossessors。Physicalconditionsprobablyhavehadsomelittleeffectonstructure，quiteindependentlyofanygoodthusgained。Correlationofgrowthhasnodoubtplayedamostimportantpart，andausefulmodificationofonepartwilloftenhaveentailedonotherpartsdiversifiedchangesofnodirectuse。Soagaincharacterswhichformerlywereuseful，orwhichformerlyhadarisenfromcorrelationofgrowth，orfromotherunknowncause，mayreappearfromthelawofreversion，thoughnowofnodirectuse。Theeffectsofsexualselection，whendisplayedinbeautytocharmthefemales，canbecalledusefulonlyinratheraforcedsense。Butbyfarthemostimportantconsiderationisthatthechiefpartoftheorganisationofeverybeingissimplyduetoinheritance。andconsequently，thougheachbeingassuredlyiswellfittedforitsplaceinnature，manystructuresnowhavenodirectrelationtothehabitsoflifeofeachspecies。Thus，wecanhardlybelievethatthewebbedfeetoftheuplandgooseorofthefrigate-birdareofspecialusetothesebirds。wecannotbelievethatthesamebonesinthearmofthemonkey，intheforelegofthehorse，inthewingofthebat，andintheflipperoftheseal，areofspecialusetotheseanimals。

Wemaysafelyattributethesestructurestoinheritance。Buttotheprogenitoroftheuplandgooseandofthefrigate-bird，webbedfeetnodoubtwereasusefulastheynowaretothemostaquaticofexistingbirds。Sowemaybelievethattheprogenitorofthesealhadnotaflipper，butafootwithfivetoesfittedforwalkingorgrasping。andwemayfurtherventuretobelievethattheseveralbonesinthelimbsofthemonkey，horse，andbat，whichhavebeeninheritedfromacommonprogenitor，wereformerlyofmorespecialusetothatprogenitor，oritsprogenitors，thantheynowaretotheseanimalshavingsuchwidelydiversifiedhabits。Thereforewemayinferthattheseseveralbonesmighthavebeenacquiredthroughnaturalselection，subjectedformerly，asnow，totheseverallawsofinheritance，reversion，correlationofgrowth，&。c。Henceeverydetailofstructureineverylivingcreature(makingsomelittleallowanceforthedirectactionofphysicalconditions)maybeviewed，eitherashavingbeenofspecialusetosomeancestralform，orasbeingnowofspecialusetothedescendantsofthisformeitherdirectly，orindirectlythroughthecomplexlawsofgrowth。

Naturalselectioncannotpossiblyproduceanymodificationinanyonespeciesexclusivelyforthegoodofanotherspecies。thoughthroughoutnatureonespeciesincessantlytakesadvantageof，andprofitsby，thestructureofanother。Butnaturalselectioncananddoesoftenproducestructuresforthedirectinjuryofotherspecies，asweseeinthefangoftheadder，andintheovipositoroftheichneumon，bywhichitseggsaredepositedinthelivingbodiesofotherinsects。Ifitcouldbeprovedthatanypartofthestructureofanyonespecieshadbeenformedfortheexclusivegoodofanotherspecies，itwouldannihilatemytheory，forsuchcouldnothavebeenproducedthroughnaturalselection。Althoughmanystatementsmaybefoundinworksonnaturalhistorytothiseffect，Icannotfindevenonewhichseemstomeofanyweight。Itisadmittedthattherattlesnakehasapoison-fangforitsowndefenceandforthedestructionofitsprey。

butsomeauthorssupposethatatthesametimethissnakeisfurnishedwitharattleforitsowninjury，namely，towarnitspreytoescape。I

wouldalmostassoonbelievethatthecatcurlstheendofitstailwhenpreparingtospring，inordertowarnthedoomedmouse。ButIhavenotspaceheretoenteronthisandothersuchcases。

Naturalselectionwillneverproduceinabeinganythinginjurioustoitself，fornaturalselectionactssolelybyandforthegoodofeach。

Noorganwillbeformed，asPaleyhasremarked，forthepurposeofcausingpainorfordoinganinjurytoitspossessor。Ifafairbalancebestruckbetweenthegoodandevilcausedbyeachpart，eachwillbefoundonthewholeadvantageous。Afterthelapseoftime，underchangingconditionsoflife，ifanypartcomestobeinjurious，itwillbemodified。orifitbenotso，thebeingwillbecomeextinct，asmyriadshavebecomeextinct。

Naturalselectiontendsonlytomakeeachorganicbeingasperfectas，orslightlymoreperfectthan，theotherinhabitantsofthesamecountrywithwhichithastostruggleforexistence。Andweseethatthisisthedegreeofperfectionattainedundernature。TheendemicproductionsofNewZealand，forinstance，areperfectonecomparedwithanother。buttheyarenowrapidlyyieldingbeforetheadvancinglegionsofplantsandanimalsintroducedfromEurope。Naturalselectionwillnotproduceabsoluteperfection，nordowealwaysmeet，asfaraswecanjudge，withthishighstandardundernature。Thecorrectionfortheaberrationoflightissaid，onhighauthority，nottobeperfecteveninthatmostperfectorgan，theeye。

Ifourreasonleadsustoadmirewithenthusiasmamultitudeofinimitablecontrivancesinnature，thissamereasontellsus，thoughwemayeasilyerronbothsides，thatsomeothercontrivancesarelessperfect。Canweconsiderthestingofthewasporofthebeeasperfect，which，whenusedagainstmanyattackinganimals，cannotbewithdrawn，owingtothebackwardserratures，andsoinevitablycausesthedeathoftheinsectbytearingoutitsviscera？

Ifwelookatthestingofthebee，ashavingoriginallyexistedinaremoteprogenitorasaboringandserratedinstrument，likethatinsomanymembersofthesamegreatorder，andwhichhasbeenmodifiedbutnotperfectedforitspresentpurpose，withthepoisonoriginallyadaptedtocausegallssubsequentlyintensified，wecanperhapsunderstandhowitisthattheuseofthestingshouldsooftencausetheinsect’sowndeath：

forifonthewholethepowerofstingingbeusefultothecommunity，itwillfulfilalltherequirementsofnaturalselection，thoughitmaycausethedeathofsomefewmembers。Ifweadmirethetrulywonderfulpowerofscentbywhichthemalesofmanyinsectsfindtheirfemales，canweadmiretheproductionforthissinglepurposeofthousandsofdrones，whichareutterlyuselesstothecommunityforanyotherend，andwhichareultimatelyslaughteredbytheirindustriousandsterilesisters？Itmaybedifficult，butweoughttoadmirethesavageinstinctivehatredofthequeen-bee，whichurgesherinstantlytodestroytheyoungqueensherdaughtersassoonasborn，ortoperishherselfinthecombat。forundoubtedlythisisforthegoodofthecommunity。andmaternalloveormaternalhatred，thoughthelatterfortunatelyismostrare，isallthesametotheinexorableprincipleofnaturalselection。Ifweadmiretheseveralingeniouscontrivances，bywhichtheflowersoftheorchisandofmanyotherplantsarefertilisedthroughinsectagency，canweconsiderasequallyperfecttheelaborationbyourfir-treesofdensecloudsofpollen，inorderthatafewgranulesmaybewaftedbyachancebreezeontotheovules？