第11章

Wecanunderstand，ontheseviews，theveryimportantdistinctionbetweenrealaffinitiesandanalogicaloradaptiveresemblances。Lamarckfirstcalledattentiontothisdistinction，andhehasbeenablyfollowedbyMacleayandothers。Theresemblance，intheshapeofthebodyandinthefin-likeanteriorlimbs，betweenthedugong，whichisapachydermatousanimal，andthewhale，andbetweenboththesemammalsandfishes，isanalogical。

Amongstinsectsthereareinnumerableinstances：thusLinnaeus，misledbyexternalappearances，actuallyclassedanhomopterousinsectasamoth。

Weseesomethingofthesamekindeveninourdomesticvarieties，asinthethickenedstemsofthecommonandswedishturnip。Theresemblanceofthegreyhoundandracehorseishardlymorefancifulthantheanalogieswhichhavebeendrawnbysomeauthorsbetweenverydistinctanimals。Onmyviewofcharactersbeingofrealimportanceforclassification，onlyinsofarastheyrevealdescent，wecanclearlyunderstandwhyanalogicaloradaptivecharacter，althoughoftheutmostimportancetothewelfareofthebeing，arealmostvaluelesstothesystematist。Foranimals，belongingtotwomostdistinctlinesofdescent，mayreadilybecomeadaptedtosimilarconditions，andthusassumeacloseexternalresemblance。butsuchresemblanceswillnotrevealwillrathertendtoconcealtheirblood-relationshiptotheirproperlinesofdescent。Wecanalsounderstandtheapparentparadox，thattheverysamecharactersareanalogicalwhenoneclassororderiscomparedwithanother，butgivetrueaffinitieswhenthemembersofthesameclassororderarecomparedonewithanother：thustheshapeofthebodyandfin-likelimbsareonlyanalogicalwhenwhalesarecomparedwithfishes，beingadaptationsinbothclassesforswimmingthroughthewater。

buttheshapeofthebodyandfin-likelimbsserveascharactersexhibitingtrueaffinitybetweentheseveralmembersofthewhalefamily。forthesecetaceansagreeinsomanycharacters，greatandsmall，thatwecannotdoubtthattheyhaveinheritedtheirgeneralshapeofbodyandstructureoflimbsfromacommonancestor。Soitiswithfishes。

Asmembersofdistinctclasseshaveoftenbeenadaptedbysuccessiveslightmodificationstoliveundernearlysimilarcircumstances，toinhabitforinstancethethreeelementsofland，air，andwater，wecanperhapsunderstandhowitisthatanumericalparallelismhassometimesbeenobservedbetweenthesub-groupsindistinctclasses。Anaturalist，struckbyaparallelismofthisnatureinanyoneclass，byarbitrarilyraisingorsinkingthevalueofthegroupsinotherclasses(andallourexperienceshowsthatthisvaluationhashithertobeenarbitrary)，couldeasilyextendtheparallelismoverawiderange。andthustheseptenary，quinary，quaternary，andternaryclassificationshaveprobablyarisen。

Asthemodifieddescendantsofdominantspecies，belongingtothelargergenera，tendtoinherittheadvantages，whichmadethegroupstowhichtheybelonglargeandtheirparentsdominant，theyarealmostsuretospreadwidely，andtoseizeonmoreandmoreplacesintheeconomyofnature。

Thelargerandmoredominantgroupsthustendtogoonincreasinginsize。

andtheyconsequentlysupplantmanysmallerandfeeblergroups。Thuswecanaccountforthefactthatallorganisms，recentandextinct，areincludedunderafewgreatorders，understillfewerclasses，andallinonegreatnaturalsystem。Asshowinghowfewthehighergroupsareinnumber，andhowwidelyspreadtheyarethroughouttheworld，thefactisstriking，thatthediscoveryofAustraliahasnotaddedasingleinsectbelongingtoaneworder。andthatinthevegetablekingdom，asIlearnfromDr。

Hooker，ithasaddedonlytwoorthreeordersofsmallsize。

InthechapterongeologicalsuccessionIattemptedtoshow，ontheprincipleofeachgrouphavinggenerallydivergedmuchincharacterduringthelong-continuedprocessofmodification，howitisthatthemoreancientformsoflifeoftenpresentcharactersinsomeslightdegreeintermediatebetweenexistinggroups。Afewoldandintermediateparent-formshavingoccasionallytransmittedtothepresentdaydescendantsbutlittlemodified，willgivetousourso-calledosculantoraberrantgroups。Themoreaberrantanyformis，thegreatermustbethenumberofconnectingformswhichonmytheoryhavebeenexterminatedandutterlylost。Andwehavesomeevidenceofaberrantformshavingsufferedseverelyfromextinction，fortheyaregenerallyrepresentedbyextremelyfewspecies。andsuchspeciesasdooccuraregenerallyverydistinctfromeachother，whichagainimpliesextinction。ThegeneraOrnithorhynchusandLepidosiren，forexample，wouldnothavebeenlessaberranthadeachbeenrepresentedbyadozenspeciesinsteadofbyasingleone。butsuchrichnessinspecies，asIfindaftersomeinvestigation，doesnotcommonlyfalltothelotofaberrantgenera。

Wecan，Ithink，accountforthisfactonlybylookingataberrantformsasfailinggroupsconqueredbymoresuccessfulcompetitors，withafewmemberspreservedbysomeunusualcoincidenceoffavourablecircumstances。

Mr。Waterhousehasremarkedthat，whenamemberbelongingtoonegroupofanimalsexhibitsanaffinitytoaquitedistinctgroup，thisaffinityinmostcasesisgeneralandnotspecial：thus，accordingtoMr。Waterhouse，ofallRodents，thebizcachaismostnearlyrelatedtoMarsupials。butinthepointsinwhichitapproachesthisorder，itsrelationsaregeneral，andnottoanyonemarsupialspeciesmorethantoanother。AsthepointsofaffinityofthebizcachatoMarsupialsarebelievedtoberealandnotmerelyadaptive，theyaredueonmytheorytoinheritanceincommon。ThereforewemustsupposeeitherthatallRodents，includingthebizcacha，branchedofffromsomeveryancientMarsupial，whichwillhavehadacharacterinsomedegreeintermediatewithrespecttoallexistingMarsupials。orthatbothRodentsandMarsupialsbranchedofffromacommonprogenitor，andthatbothgroupshavesinceundergonemuchmodificationindivergentdirections。

Oneitherviewwemaysupposethatthebizcachahasretained，byinheritance，moreofthecharacterofitsancientprogenitorthanhaveotherRodents。

andthereforeitwillnotbespeciallyrelatedtoanyoneexistingMarsupial，butindirectlytoallornearlyallMarsupials，fromhavingpartiallyretainedthecharacteroftheircommonprogenitor，orofanearlymemberofthegroup。Ontheotherhand，ofallMarsupials，asMr。Waterhousehasremarked，thephascolomysresemblesmostnearly，notanyonespecies，butthegeneralorderofRodents。Inthiscase，however，itmaybestronglysuspectedthattheresemblanceisonlyanalogical，owingtothephascolomyshavingbecomeadaptedtohabitslikethoseofaRodent。TheelderDeCandollehasmadenearlysimilarobservationsonthegeneralnatureoftheaffinitiesofdistinctordersofplants。

Ontheprincipleofthemultiplicationandgradualdivergenceincharacterofthespeciesdescendedfromacommonparent，togetherwiththeirretentionbyinheritanceofsomecharactersincommon，wecanunderstandtheexcessivelycomplexandradiatingaffinitiesbywhichallthemembersofthesamefamilyorhighergroupareconnectedtogether。Forthecommonparentofawholefamilyofspecies，nowbrokenupbyextinctionintodistinctgroupsandsub-groups，willhavetransmittedsomeofitscharacters，modifiedinvariouswaysanddegrees，toall。andtheseveralspecieswillconsequentlyberelatedtoeachotherbycircuitouslinesofaffinityofvariouslengths(asmaybeseeninthediagramsooftenreferredto)，mountingupthroughmanypredecessors。Asitisdifficulttoshowtheblood-relationshipbetweenthenumerouskindredofanyancientandnoblefamily，evenbytheaidofagenealogicaltree，andalmostimpossibletodothiswithoutthisaid，wecanunderstandtheextraordinarydifficultywhichnaturalistshaveexperiencedindescribing，withouttheaidofadiagram，thevariousaffinitieswhichtheyperceivebetweenthemanylivingandextinctmembersofthesamegreatnaturalclass。

Extinction，aswehaveseeninthefourthchapter，hasplayedanimportantpartindefiningandwideningtheintervalsbetweentheseveralgroupsineachclass。Wemaythusaccountevenforthedistinctnessofwholeclassesfromeachotherforinstance，ofbirdsfromallothervertebrateanimalsbythebeliefthatmanyancientformsoflifehavebeenutterlylost，throughwhichtheearlyprogenitorsofbirdswereformerlyconnectedwiththeearlyprogenitorsoftheothervertebrateclasses。Therehasbeenlessentireextinctionoftheformsoflifewhichonceconnectedfisheswithbatrachians。

Therehasbeenstilllessinsomeotherclasses，asinthatoftheCrustacea，forherethemostwonderfullydiverseformsarestilltiedtogetherbyalong，butbroken，chainofaffinities。Extinctionhasonlyseparatedgroups：ithasbynomeansmadethem。forifeveryformwhichhaseverlivedonthisearthweresuddenlytoreappear，thoughitwouldbequiteimpossibletogivedefinitionsbywhicheachgroupcouldbedistinguishedfromothergroups，asallwouldblendtogetherbystepsasfineasthosebetweenthefinestexistingvarieties，neverthelessanaturalclassification，oratleastanaturalarrangement，wouldbepossible。Weshallseethisbyturningtothediagram：theletters，AtoL，mayrepresentelevenSiluriangenera，someofwhichhaveproducedlargegroupsofmodifieddescendants。

Everyintermediatelinkbetweentheseelevengeneraandtheirprimordialparent，andeveryintermediatelinkineachbranchandsub-branchoftheirdescendants，maybesupposedtobestillalive。andthelinkstobeasfineasthosebetweenthefinestvarieties。Inthiscaseitwouldbequiteimpossibletogiveanydefinitionbywhichtheseveralmembersoftheseveralgroupscouldbedistinguishedfromtheirmoreimmediateparents。ortheseparentsfromtheirancientandunknownprogenitor。Yetthenaturalarrangementinthediagramwouldstillholdgood。and，ontheprincipleofinheritance，alltheformsdescendedfromA，orfromI，wouldhavesomethingincommon。

Inatreewecanspecifythisorthatbranch，thoughattheactualforkthetwouniteandblendtogether。Wecouldnot，asIhavesaid，definetheseveralgroups。butwecouldpickouttypes，orforms，representingmostofthecharactersofeachgroup，whetherlargeorsmall，andthusgiveageneralideaofthevalueofthedifferencesbetweenthem。Thisiswhatweshouldbedrivento，ifwewereevertosucceedincollectingalltheformsinanyclasswhichhavelivedthroughoutalltimeandspace。

Weshallcertainlyneversucceedinmakingsoperfectacollection：nevertheless，incertainclasses，wearetendinginthisdirection。andMilneEdwardshaslatelyinsisted，inanablepaper，onthehighimportanceoflookingtotypes，whetherornotwecanseparateanddefinethegroupstowhichsuchtypesbelong。

Finally，wehaveseenthatnaturalselection，whichresultsfromthestruggleforexistence，andwhichalmostinevitablyinducesextinctionanddivergenceofcharacterinthemanydescendantsfromonedominantparent-species，explainsthatgreatanduniversalfeatureintheaffinitiesofallorganicbeings，namely，theirsubordinationingroupundergroup。Weusetheelementofdescentinclassingtheindividualsofbothsexesandofallages，althoughhavingfewcharactersincommon，underonespecies。weusedescentinclassingacknowledgedvarieties，howeverdifferenttheymaybefromtheirparent。

andIbelievethiselementofdescentisthehiddenbondofconnexionwhichnaturalistshavesoughtunderthetermoftheNaturalSystem。Onthisideaofthenaturalsystembeing，insofarasithasbeenperfected，genealogicalinitsarrangement，withthegradesofdifferencebetweenthedescendantsfromacommonparent，expressedbythetermsgenera，families，orders，&。c。，wecanunderstandtheruleswhichwearecompelledtofollowinourclassification。Wecanunderstandwhywevaluecertainresemblancesfarmorethanothers。whywearepermittedtouserudimentaryanduselessorgans，orothersoftriflingphysiologicalimportance。why，incomparingonegroupwithadistinctgroup，wesummarilyrejectanalogicaloradaptivecharacters，andyetusethesesamecharacterswithinthelimitsofthesamegroup。Wecanclearlyseehowitisthatalllivingandextinctformscanbegroupedtogetherinonegreatsystem。andhowtheseveralmembersofeachclassareconnectedtogetherbythemostcomplexandradiatinglinesofaffinities。Weshallnever，probably，disentangletheinextricablewebofaffinitiesbetweenthemembersofanyoneclass。butwhenwehaveadistinctobjectinview，anddonotlooktosomeunknownplanofcreation，wemayhopetomakesurebutslowprogress。MorphologyWehaveseenthatthemembersofthesameclass，independentlyoftheirhabitsoflife，resembleeachotherinthegeneralplanoftheirorganisation。

Thisresemblanceisoftenexpressedbytheterm`unityoftype。’orbysayingthattheseveralpartsandorgansinthedifferentspeciesoftheclassarehomologous。ThewholesubjectisincludedunderthegeneralnameofMorphology。Thisisthemostinterestingdepartmentofnaturalhistory，andmaybesaidtobeitsverysoul。Whatcanbemorecuriousthanthatthehandofaman，formedforgrasping，thatofamolefordigging，thelegofthehorse，thepaddleoftheporpoise，andthewingofthebat，shouldallbeconstructedonthesamepattern，andshouldincludethesamebones，inthesamerelativepositions？GeoffroyStHilairehasinsistedstronglyonthehighimportanceofrelativeconnexioninhomologousorgans：

thepartsmaychangetoalmostanyextentinformandsize，andyettheyalwaysremainconnectedtogetherinthesameorder。Weneverfind，forinstance，thebonesofthearmandforearm，orofthethighandleg，transposed。

Hencethesamenamescanbegiventothehomologousbonesinwidelydifferentanimals。Weseethesamegreatlawintheconstructionofthemouthsofinsects：whatcanbemoredifferentthantheimmenselylongspiralproboscisofasphinx-moth，thecuriousfoldedoneofabeeorbug，andthegreatjawsofabeetle？yetalltheseorgans，servingforsuchdifferentpurposes，areformedbyinfinitelynumerousmodificationsofanupperlip，mandibles，andtwopairsofmaxillae。Analogouslawsgoverntheconstructionofthemouthsandlimbsofcrustaceans。Soitiswiththeflowersofplants。

Nothingcanbemorehopelessthantoattempttoexplainthissimilarityofpatterninmembersofthesameclass，byutilityorbythedoctrineoffinalcauses。ThehopelessnessoftheattempthasbeenexpresslyadmittedbyOweninhismostinterestingworkonthe`NatureofLimbs。’Ontheordinaryviewoftheindependentcreationofeachbeing，wecanonlysaythatsoitis。thatithassopleasedtheCreatortoconstructeachanimalandplant。

Theexplanationismanifestonthetheoryofthenaturalselectionofsuccessiveslightmodifications，eachmodificationbeingprofitableinsomewaytothemodifiedform，butoftenaffectingbycorrelationofgrowthotherpartsoftheorganisation。Inchangesofthisnature，therewillbelittleornotendencytomodifytheoriginalpattern，ortotransposeparts。Thebonesofalimbmightbeshortenedandwidenedtoanyextent，andbecomegraduallyenvelopedinthickmembrane，soastoserveasafin。

orawebbedfootmighthaveallitsbones，orcertainbones，lengthenedtoanyextent，andthemembraneconnectingthemincreasedtoanyextent，soastoserveasawing：yetinallthisgreatamountofmodificationtherewillbenotendencytoaltertheframeworkofbonesortherelativeconnexionoftheseveralparts。Ifwesupposethattheancientprogenitor，thearchetypeasitmaybecalled，ofallmammals，haditslimbsconstructedontheexistinggeneralpattern，forwhateverpurposetheyserved，wecanatonceperceivetheplainsignificationofthehomologousconstructionofthelimbsthroughoutthewholeclass。Sowiththemouthsofinsects，wehaveonlytosupposethattheircommonprogenitorhadanupperlip，mandibles，andtwopairofmaxillae，thesepartsbeingperhapsverysimpleinform。andthennaturalselectionwillaccountfortheinfinitediversityinstructureandfunctionofthemouthsofinsects。Nevertheless，itisconceivablethatthegeneralpatternofanorganmightbecomesomuchobscuredastobefinallylost，bytheatrophyandultimatelybythecompleteabortionofcertainparts，bythesolderingtogetherofotherparts，andbythedoublingormultiplicationofothers，variationswhichweknowtobewithinthelimitsofpossibility。Inthepaddlesoftheextinctgiganticsea-lizards，andinthemouthsofcertainsuctorialcrustaceans，thegeneralpatternseemstohavebeenthustoacertainextentobscured。

Thereisanotherandequallycuriousbranchofthepresentsubject。

namely，thecomparisonnotofthesamepartindifferentmembersofaclass，butofthedifferentpartsororgansinthesameindividual。Mostphysiologistsbelievethatthebonesoftheskullarehomologouswiththatiscorrespondinnumberandinrelativeconnexionwiththeelementalpartsofacertainnumberofvertebrae。Theanteriorandposteriorlimbsineachmemberofthevertebrateandarticulateclassesareplainlyhomologous。Weseethesamelawincomparingthewonderfullycomplexjawsandlegsincrustaceans。

Itisfamiliartoalmosteveryone，thatinaflowertherelativepositionofthesepals，petals，stamens，andpistils，aswellastheirintimatestructure，areintelligibleintheviewthattheyconsistofmetamorphosedleaves，arrangedinaspire。Inmonstrousplants，weoftengetdirectevidenceofthepossibilityofoneorganbeingtransformedintoanother。andwecanactuallyseeinembryoniccrustaceansandinmanyotheranimals，andinflowers，thatorganswhichwhenmaturebecomeextremelydifferent，areatanearlystageofgrowthexactlyalike。

Howinexplicablearethesefactsontheordinaryviewofcreation！Whyshouldthebrainbeenclosedinaboxcomposedofsuchnumerousandsuchextraordinarilyshapedpiecesofbone？AsOwenhasremarked，thebenefitderivedfromtheyieldingoftheseparatepiecesintheactofparturitionofmammals，willbynomeansexplainthesameconstructionintheskullsofbirds。Whyshouldsimilarboneshavebeencreatedintheformationofthewingandlegofabat，usedastheyareforsuchtotallydifferentpurposes？Whyshouldonecrustacean，whichhasanextremelycomplexmouthformedofmanyparts，consequentlyalwayshavefewerlegs。orconversely，thosewithmanylegshavesimplermouths？Whyshouldthesepals，petals，stamens，andpistilsinanyindividualflower，thoughfittedforsuchwidelydifferentpurposes，beallconstructedonthesamepattern？

Onthetheoryofnaturalselection，wecansatisfactorilyanswerthesequestions。Inthevertebrata，weseeaseriesofinternalvertebraebearingcertainprocessesandappendages。inthearticulata，weseethebodydividedintoaseriesofsegments，bearingexternalappendages。andinfloweringplants，weseeaseriesofsuccessivespiralwhorlsofleaves。Anindefiniterepetitionofthesamepartororganisthecommoncharacteristic(asOwenhasobserved)ofallloworlittle-modifiedforms。thereforewemayreadilybelievethattheunknownprogenitorofthevertebratapossessedmanyvertebrae。

theunknownprogenitorofthearticulata，manysegments。andtheunknownprogenitoroffloweringplants，manyspiralwhorlsofleaves。Wehaveformerlyseenthatpartsmanytimesrepeatedareeminentlyliabletovaryinnumberandstructure。consequentlyitisquiteprobablethatnaturalselection，duringalong-continuedcourseofmodification，shouldhaveseizedonacertainnumberoftheprimordiallysimilarelements，manytimesrepeated，andhaveadaptedthemtothemostdiversepurposes。Andasthewholeamountofmodificationwillhavebeeneffectedbyslightsuccessivesteps，weneednotwonderatdiscoveringinsuchpartsororgans，acertaindegreeoffundamentalresemblance，retainedbythestrongprincipleofinheritance。

Inthegreatclassofmolluscs，thoughwecanhomologisethepartsofonespecieswiththoseofanotheranddistinctspecies，wecanindicatebutfewserialhomologies。thatis，weareseldomenabledtosaythatonepartororganishomologouswithanotherinthesameindividual。Andwecanunderstandthisfact。forinmolluscs，eveninthelowestmembersoftheclass，wedonotfindnearlysomuchindefiniterepetitionofanyonepart，aswefindintheothergreatclassesoftheanimalandvegetablekingdoms。

Naturalistsfrequentlyspeakoftheskullasformedofmetamorphosedvertebrae：thejawsofcrabsasmetamorphosedlegs。thestamensandpistilsofflowersasmetamorphosedleaves。butitwouldinthesecasesprobablybemorecorrect，asProfessorHuxleyhasremarked，tospeakofbothskullandvertebrae，bothjawsandlegs，&。c。，ashavingbeenmetamorphosed，notonefromtheother，butfromsomecommonelement。Naturalists，however，usesuchlanguageonlyinametaphoricalsense：theyarefarfrommeaningthatduringalongcourseofdescent，primordialorgansofanykindvertebraeintheonecaseandlegsintheotherhaveactuallybeenmodifiedintoskullsorjaws。Yetsostrongistheappearanceofamodificationofthisnaturehavingoccurred，thatnaturalistscanhardlyavoidemployinglanguagehavingthisplainsignification。Onmyviewthesetermsmaybeusedliterally。

andthewonderfulfactofthejaws，forinstance，ofacrabretainingnumerouscharacters，whichtheywouldprobablyhaveretainedthroughinheritance，iftheyhadreallybeenmetamorphosedduringalongcourseofdescentfromtruelegs，orfromsomesimpleappendage，isexplained。EmbryologyIthasalreadybeencasuallyremarkedthatcertainorgansintheindividual，whichwhenmaturebecomewidelydifferentandservefordifferentpurposes，areintheembryoexactlyalike。Theembryos，also，ofdistinctanimalswithinthesameclassareoftenstrikinglysimilar：abetterproofofthiscannotbegiven，thanacircumstancementionedbyAgassiz，namely，thathavingforgottentotickettheembryoofsomevertebrateanimal，hecannotnowtellwhetheritbethatofamammal，bird，orreptile。Thevermiformlarvaeofmoths，flies，beetles，&。c。，resembleeachothermuchmorecloselythandothematureinsects。butinthecaseoflarvae，theembryosareactive，andhavebeenadaptedforspeciallinesoflife。Atraceofthelawofembryonicresemblance，sometimeslaststillaratherlateage：

thusbirdsofthesamegenus，andofcloselyalliedgenera，oftenresembleeachotherintheirfirstandsecondplumage。asweseeinthespottedfeathersinthethrushgroup。Inthecattribe，mostofthespeciesarestripedorspottedinlines。andstripescanbeplainlydistinguishedinthewhelpofthelion。Weoccasionallythoughrarelyseesomethingofthiskindinplants：thustheembryonicleavesoftheulexorfurze，andthefirstleavesofthephyllodineousacaceas，arepinnateordividedliketheordinaryleavesoftheleguminosae。

Thepointsofstructure，inwhichtheembryosofwidelydifferentanimalsofthesameclassresembleeachother，oftenhavenodirectrelationtotheirconditionsofexistence。Wecannot，forinstance，supposethatintheembryosofthevertebratathepeculiarloop-likecourseofthearteriesnearthebranchialslitsarerelatedtosimilarconditions，intheyoungmammalwhichisnourishedinthewombofitsmother，intheeggofthebirdwhichishatchedinanest，andinthespawnofafrogunderwater。

Wehavenomorereasontobelieveinsucharelation，thanwehavetobelievethatthesamebonesinthehandofaman，wingofabat，andfinofaporpoise，arerelatedtosimilarconditionsoflife。Noonewillsupposethatthestripesonthewhelpofalion，orthespotsontheyoungblackbird，areofanyusetotheseanimals，orarerelatedtotheconditionstowhichtheyareexposed。

Thecase，however，isdifferentwhenananimalduringanypartofitsembryoniccareerisactive，andhastoprovideforitself。Theperiodofactivitymaycomeonearlierorlaterinlife。butwheneveritcomeson，theadaptationofthelarvatoitsconditionsoflifeisjustasperfectandasbeautifulasintheadultanimal。Fromsuchspecialadaptations，thesimilarityofthelarvaeoractiveembryosofalliedanimalsissometimesmuchobscured。andcasescouldbegivenofthelarvaeoftwospecies，oroftwogroupsofspecies，differingquiteasmuch，orevenmore，fromeachotherthandotheiradultparents。Inmostcases，however，thelarvae，thoughactive，stillobeymoreorlesscloselythelawofcommonembryonicresemblance。Cirripedesaffordagoodinstanceofthis：eventheillustriousCuvierdidnotperceivethatabarnaclewas，asitcertainlyis，acrustacean。

butaglanceatthelarvashowsthistobethecaseinanunmistakeablemanner。Soagainthetwomaindivisionsofcirripedes，thepedunculatedandsessile，whichdifferwidelyinexternalappearance，havelarvaeinalltheirseveralstagesbarelydistinguishable。

Theembryointhecourseofdevelopmentgenerallyrisesinorganisation：

Iusethisexpression，thoughIamawarethatitishardlypossibletodefineclearlywhatismeantbytheorganisationbeinghigherorlower。

Butnooneprobablywilldisputethatthebutterflyishigherthanthecaterpillar。Insomecases，however，thematureanimalisgenerallyconsideredaslowerinthescalethanthelarva，aswithcertainparasiticcrustaceans。

Toreferonceagaintocirripedes：thelarvaeinthefirststagehavethreepairsoflegs，averysimplesingleeye，andaprobosciformedmouth，withwhichtheyfeedlargely，fortheyincreasemuchinsize。Inthesecondstage，answeringtothechrysalisstageofbutterflies，theyhavesixpairsofbeautifullyconstructednatatorylegs，apairofmagnificentcompoundeyes，andextremelycomplexantennae。buttheyhaveaclosedandimperfectmouth，andcannotfeed：theirfunctionatthisstageis，tosearchbytheirwell-developedorgansofsense，andtoreachbytheiractivepowersofswimming，aproperplaceonwhichtobecomeattachedandtoundergotheirfinalmetamorphosis。Whenthisiscompletedtheyarefixedforlife：theirlegsarenowconvertedintoprehensileorgans。theyagainobtainawell-constructedmouth。buttheyhavenoantennae，andtheirtwoeyesarenowreconvertedintoaminute，single，andverysimpleeye-spot。Inthislastandcompletestate，cirripedesmaybeconsideredaseithermorehighlyormorelowlyorganisedthantheywereinthelarvalcondition。Butinsomegenerathelarvaebecomedevelopedeitherintohermaphroditeshavingtheordinarystructure，orintowhatIhavecalledcomplementalmales：andinthelatter，thedevelopmenthasassuredlybeenretrograde。forthemaleisameresack，whichlivesforashorttime，andisdestituteofmouth，stomach，orotherorganofimportance，exceptingforreproduction。

Wearesomuchaccustomedtoseedifferencesinstructurebetweentheembryoandtheadult，andlikewiseaclosesimilarityintheembryosofwidelydifferentanimalswithinthesameclass，thatwemightbeledtolookatthesefactsasnecessarilycontingentinsomemannerongrowth。

Butthereisnoobviousreasonwhy，forinstance，thewingofabat，orthefinofaporpoise，shouldnothavebeensketchedoutwithallthepartsinproperproportion，assoonasanystructurebecamevisibleintheembryo。

Andinsomewholegroupsofanimalsandincertainmembersofothergroups，theembryodoesnotatanyperioddifferwidelyfromtheadult：thusOwenhasremarkedinregardtocuttle-fish，`thereisnometamorphosis。thecephalopodiccharacterismanifestedlongbeforethepartsoftheembryoarecompleted。’andagaininspiders，`thereisnothingworthytobecalledametamorphosis。’Thelarvaeofinsects，whetheradaptedtothemostdiverseandactivehabits，orquiteinactive，beingfedbytheirparentsorplacedinthemidstofpropernutriment，yetnearlyallpassthroughasimilarworm-likestageofdevelopment。butinsomefewcases，asinthatofAphis，ifwelooktotheadmirabledrawingsbyProfessorHuxleyofthedevelopmentofthisinsect，weseenotraceofthevermiformstage。

How，then，canweexplaintheseseveralfactsinembryology，namelytheverygeneral，butnotuniversaldifferenceinstructurebetweentheembryoandtheadult。ofpartsinthesameindividualembryo，whichultimatelybecomeveryunlikeandservefordiversepurposes，beingatthisearlyperiodofgrowthalike。ofembryosofdifferentspecieswithinthesameclass，generally，butnotuniversally，resemblingeachother。ofthestructureoftheembryonotbeingcloselyrelatedtoitsconditionsofexistence，exceptwhentheembryobecomesatanyperiodoflifeactiveandhastoprovideforitself。oftheembryoapparentlyhavingsometimesahigherorganisationthanthematureanimal，intowhichitisdeveloped。Ibelievethatallthesefactscanbeexplained，asfollows，ontheviewofdescentwithmodification。

Itiscommonlyassumed，perhapsfrommonstrositiesoftenaffectingtheembryoataveryearlyperiod，thatslightvariationsnecessarilyappearatanequallyearlyperiod。Butwehavelittleevidenceonthisheadindeedtheevidenceratherpointstheotherway。foritisnotoriousthatbreedersofcattle，horses，andvariousfancyanimals，cannotpositivelytell，untilsometimeaftertheanimalhasbeenborn，whatitsmeritsorformwillultimatelyturnout。Weseethisplainlyinourownchildren。wecannotalwaystellwhetherthechildwillbetallorshort，orwhatitsprecisefeatureswillbe。Thequestionisnot，atwhatperiodoflifeanyvariationhasbeencaused，butatwhatperioditisfullydisplayed。Thecausemayhaveacted，andIbelievegenerallyhasacted，evenbeforetheembryoisformed。andthevariationmaybeduetothemaleandfemalesexualelementshavingbeenaffectedbytheconditionstowhicheitherparent，ortheirancestors，havebeenexposed。Neverthelessaneffectthuscausedataveryearlyperiod，evenbeforetheformationoftheembryo，mayappearlateinlife。aswhenanhereditarydisease，whichappearsinoldagealone，hasbeencommunicatedtotheoffspringfromthereproductiveelementofoneparent。Oragain，aswhenthehornsofcross-bredcattlehavebeenaffectedbytheshapeofthehornsofeitherparent。Forthewelfareofaveryyounganimal，aslongasitremainsinitsmother’swomb，orintheegg，oraslongasitisnourishedandprotectedbyitsparent，itmustbequiteunimportantwhethermostofitscharactersarefullyacquiredalittleearlierorlaterinlife。Itwouldnotsignify，forinstance，toabirdwhichobtaineditsfoodbestbyhavingalongbeak，whetherornotitassumedabeakofthisparticularlength，aslongasitwasfedbyitsparents。Hence，Iconclude，thatitisquitepossible，thateachofthemanysuccessivemodifications，bywhicheachspecieshasacquireditspresentstructure，mayhavesupervenedatanotveryearlyperiodoflife。andsomedirectevidencefromourdomesticanimalssupportsthisview。Butinothercasesitisquitepossiblethateachsuccessivemodification，ormostofthem，mayhaveappearedatanextremelyearlyperiod。

Ihavestatedinthefirstchapter，thatthereissomeevidencetorenderitprobable，thatatwhateverageanyvariationfirstappearsintheparent，ittendstoreappearatacorrespondingageintheoffspring。Certainvariationscanonlyappearatcorrespondingages，forinstance，peculiaritiesinthecaterpillar，cocoon，orimagostatesofthesilk-moth。or，again，inthehornsofalmostfull-growncattle。Butfurtherthanthis，variationswhich，forallthatwecansee，mighthaveappearedearlierorlaterinlife，tendtoappearatacorrespondingageintheoffspringandparent。Iamfarfrommeaningthatthisisinvariablythecase。andIcouldgiveagoodmanycasesofvariations(takingthewordinthelargestsense)whichhavesupervenedatanearlierageinthechildthanintheparent。

Thesetwoprinciples，iftheirtruthbeadmitted，will，Ibelieve，explainalltheabovespecifiedleadingfactsinembryology。Butfirstletuslookatafewanalogouscasesindomesticvarieties。SomeauthorswhohavewrittenonDogs，maintainthatthegreyhoundandbulldog，thoughappearingsodifferent，arereallyvarietiesmostcloselyallied，andhaveprobablydescendedfromthesamewildstock。henceIwascurioustoseehowfartheirpuppiesdifferedfromeachother：Iwastoldbybreedersthattheydifferedjustasmuchastheirparents，andthis，judgingbytheeye，seemedalmosttobethecase。butonactuallymeasuringtheolddogsandtheirsix-daysoldpuppies，Ifoundthatthepuppieshadnotnearlyacquiredtheirfullamountofproportionaldifference。So，again，Iwastoldthatthefoalsofcartandrace-horsesdifferedasmuchasthefull-grownanimals。andthissurprisedmegreatly，asIthinkitprobablethatthedifferencebetweenthesetwobreedshasbeenwhollycausedbyselectionunderdomestication。buthavinghadcarefulmeasurementsmadeofthedamandofathree-daysoldcoltofaraceandheavycart-horse，Ifindthatthecoltshavebynomeansacquiredtheirfullamountofproportionaldifference。

Astheevidenceappearstomeconclusive，thattheseveraldomesticbreedsofpigeonhavedescendedfromonewildspecies，Icomparedyoungpigeonsofvariousbreeds，withintwelvehoursafterbeinghatched。Icarefullymeasuredtheproportions(butwillnotheregivedetails)ofthebeak，widthofmouth，lengthofnostrilandofeyelid，sizeoffeetandlengthofleg，inthewildstock，inpouters，fantails，runts，barbs，dragons，carriers，andtumblers。Nowsomeofthesebirds，whenmature，differsoextraordinarilyinlengthandformofbeak，thattheywould，Icannotdoubt，berankedindistinctgenera，hadtheybeennaturalproductions。Butwhenthenestlingbirdsoftheseseveralbreedswereplacedinarow，thoughmostofthemcouldbedistinguishedfromeachother，yettheirproportionaldifferencesintheabovespecifiedseveralpointswereincomparablylessthaninthefull-grownbirds。Somecharacteristicpointsofdifferenceforinstance，thatofthewidthofmouth——couldhardlybedetectedintheyoung。Buttherewasoneremarkableexceptiontothisrule，fortheyoungoftheshort-facedtumblerdifferedfromtheyoungofthewildrock-pigeonandoftheotherbreeds，inallitsproportions，almostexactlyasmuchasintheadultstate。

Thetwoprinciplesabovegivenseemtometoexplainthesefactsinregardtothelaterembryonicstagesofourdomesticvarieties。Fanciersselecttheirhorses，dogs，andpigeons，forbreeding，whentheyarenearlygrownup：theyareindifferentwhetherthedesiredqualitiesandstructureshavebeenacquiredearlierorlaterinlife，ifthefull-grownanimalpossessesthem。Andthecasesjustgiven，moreespeciallythatofpigeons，seemtoshowthatthecharacteristicdifferenceswhichgivevaluetoeachbreed，andwhichhavebeenaccumulatedbyman’sselection，havenotgenerallyfirstappearedatanearlyperiodoflife，andhavebeeninheritedbytheoffspringatacorrespondingnotearlyperiod。Butthecaseoftheshort-facedtumbler，whichwhentwelvehoursoldhadacquireditsproperproportions，provesthatthisisnottheuniversalrule。forherethecharacteristicdifferencesmusteitherhaveappearedatanearlierperiodthanusual，or，ifnotso，thedifferencesmusthavebeeninherited，notatthecorresponding，butatanearlierage。

Nowletusapplythesefactsandtheabovetwoprincipleswhichlatter，thoughnotprovedtrue，canbeshowntobeinsomedegreeprobabletospeciesinastateofnature。Letustakeagenusofbirds，descendedonmytheoryfromsomeoneparent-species，andofwhichtheseveralnewspecieshavebecomemodifiedthroughnaturalselectioninaccordancewiththeirdiversehabits。Then，fromthemanyslightsuccessivestepsofvariationhavingsupervenedataratherlateage，andhavingbeeninheritedatacorrespondingage，theyoungofthenewspeciesofoursupposedgenuswillmanifestlytendtoresembleeachothermuchmorecloselythandotheadults，justaswehaveseeninthecaseofpigeons。Wemayextendthisviewtowholefamiliesorevenclasses。Thefore-limbs，forinstance，whichservedaslegsintheparent-species，maybecome，byalongcourseofmodification，adaptedinonedescendanttoactashands，inanotheraspaddles，inanotheraswings。andontheabovetwoprinciplesnamelyofeachsuccessivemodificationsuperveningataratherlateage，andbeinginheritedatacorrespondinglateagethefore-limbsintheembryosoftheseveraldescendantsoftheparent-specieswillstillresembleeachotherclosely，fortheywillnothavebeenmodified。Butineachindividualnewspecies，theembryonicfore-limbswilldiffergreatlyfromthefore-limbsinthematureanimal。thelimbsinthelatterhavingundergonemuchmodificationataratherlateperiodoflife，andhavingthusbeenconvertedintohands，orpaddles，orwings。

Whateverinfluencelong-continuedexerciseoruseontheonehand，anddisuseontheother，mayhaveinmodifyinganorgan，suchinfluencewillmainlyaffectthematureanimal，whichhascometoitsfullpowersofactivityandhastogainitsownliving。andtheeffectsthusproducedwillbeinheritedatacorrespondingmatureage。Whereastheyoungwillremainunmodified，orbemodifiedinalesserdegree，bytheeffectsofuseanddisuse。

Incertaincasesthesuccessivestepsofvariationmightsupervene，fromcausesofwhichwearewhollyignorant，ataveryearlyperiodoflife，oreachstepmightbeinheritedatanearlierperiodthanthatatwhichitfirstappeared。Ineithercase(aswiththeshort-facedtumbler)

theyoungorembryowouldcloselyresemblethematureparent-form。Wehaveseenthatthisistheruleofdevelopmentincertainwholegroupsofanimals，aswithcuttle-fishandspiders，andwithafewmembersofthegreatclassofinsects，aswithAphis。Withrespecttothefinalcauseoftheyounginthesecasesnotundergoinganymetamorphosis，orcloselyresemblingtheirparentsfromtheirearliestage，wecanseethatthiswouldresultfromthetwofollowingcontingencies。firstly，fromtheyoung，duringacourseofmodificationcarriedonformanygenerations，havingtoprovidefortheirownwantsataveryearlystageofdevelopment，andsecondly，fromtheirfollowingexactlythesamehabitsoflifewiththeirparents。

forinthiscase，itwouldbeindispensablefortheexistenceofthespecies，thatthechildshouldbemodifiedataveryearlyageinthesamemannerwithitsparents，inaccordancewiththeirsimilarhabits。Somefurtherexplanation，however，oftheembryonotundergoinganymetamorphosisisperhapsrequisite。If，ontheotherhand，itprofitedtheyoungtofollowhabitsoflifeinanydegreedifferentfromthoseoftheirparent，andconsequentlytobeconstructedinaslightlydifferentmanner，then，ontheprincipleofinheritanceatcorrespondingages，theactiveyoungorlarvaemighteasilyberenderedbynaturalselectiondifferenttoanyconceivableextentfromtheirparents。Suchdifferencesmight，also，becomecorrelatedwithsuccessivestagesofdevelopment。sothatthelarvae，inthefirststage，mightdiffergreatlyfromthelarvaeinthesecondstage，aswehaveseentobethecasewithcirripedes。Theadultmightbecomefittedforsitesorhabits，inwhichorgansoflocomotionorofthesenses，&。c。，wouldbeuseless。andinthiscasethefinalmetamorphosiswouldbesaidtoberetrograde。

Asalltheorganicbeings，extinctandrecent，whichhaveeverlivedonthisearthhavetobeclassedtogether，andasallhavebeenconnectedbythefinestgradations，thebest，orindeed，ifourcollectionswerenearlyperfect，theonlypossiblearrangement，wouldbegenealogical。Descentbeingonmyviewthehiddenbondofconnexionwhichnaturalistshavebeenseekingunderthetermofthenaturalsystem。Onthisviewwecanunderstandhowitisthat，intheeyesofmostnaturalists，thestructureoftheembryoisevenmoreimportantforclassificationthanthatoftheadult。Fortheembryoistheanimalinitslessmodifiedstate。andinsofaritrevealsthestructureofitsprogenitor。Intwogroupsofanimal，howevermuchtheymayatpresentdifferfromeachotherinstructureandhabits，iftheypassthroughthesameorsimilarembryonicstages，wemayfeelassuredthattheyhavebothdescendedfromthesameornearlysimilarparents，andarethereforeinthatdegreecloselyrelated。Thus，communityinembryonicstructurerevealscommunityofdescent。Itwillrevealthiscommunityofdescent，howevermuchthestructureoftheadultmayhavebeenmodifiedandobscured。wehaveseen，forinstance，thatcirripedescanatonceberecognisedbytheirlarvaeasbelongingtothegreatclassofcrustaceans。

Astheembryonicstateofeachspeciesandgroupofspeciespartiallyshowsusthestructureoftheirlessmodifiedancientprogenitors，wecanclearlyseewhyancientandextinctformsoflifeshouldresembletheembryosoftheirdescendants，ourexistingspecies。Agassizbelievesthistobealawofnature。butIamboundtoconfessthatIonlyhopetoseethelawhereafterprovedtrue。Itcanbeprovedtrueinthosecasesaloneinwhichtheancientstate，nowsupposedtoberepresentedinmanyembryos，hasnotbeenobliterated，eitherbythesuccessivevariationsinalongcourseofmodificationhavingsupervenedataveryearlyage，orbythevariationshavingbeeninheritedatanearlierperiodthanthatatwhichtheyfirstappeared。Itshouldalsobeborneinmind，thatthesupposedlawofresemblanceofancientformsoflifetotheembryonicstagesofrecentforms，maybetrue，butyet，owingtothegeologicalrecordnotextendingfarenoughbackintime，mayremainforalongperiod，orforever，incapableofdemonstration。

Thus，asitseemstome，theleadingfactsinembryology，whicharesecondinimportancetononeinnaturalhistory，areexplainedontheprincipleofslightmodificationsnotappearing，inthemanydescendantsfromsomeoneancientprogenitor，ataveryearlyperiodinthelifeofeach，thoughperhapscausedattheearliest，andbeinginheritedatacorrespondingnotearlyperiod。Embryologyrisesgreatlyininterest，whenwethuslookattheembryoasapicture，moreorlessobscured，ofthecommonparent-formofeachgreatclassofanimals。Rudimentary，atrophied，orabortedorgansOrgansorpartsinthisstrangecondition，bearingthestampofinutility，areextremelycommonthroughoutnature。Forinstance，rudimentarymammaeareverygeneralinthemalesofmammals：Ipresumethatthe`bastard-wing’

inbirdsmaybesafelyconsideredasadigitinarudimentarystate：inverymanysnakesonelobeofthelungsisrudimentary。inothersnakestherearerudimentsofthepelvisandhindlimbs。Someofthecasesofrudimentaryorgansareextremelycurious。forinstance，thepresenceofteethinfoetalwhales，whichwhengrownuphavenotatoothintheirheads。

andthepresenceofteeth，whichnevercutthroughthegums，intheupperjawsofourunborncalves。Ithasevenbeenstatedongoodauthoritythatrudimentsofteethcanbedetectedinthebeaksofcertainembryonicbirds。

Nothingcanbeplainerthanthatwingsareformedforflight，yetinhowmanyinsectsdoweseewingssoreducedinsizeastobeutterlyincapableofflight，andnotrarelylyingunderwing-cases，firmlysolderedtogether！

Themeaningofrudimentaryorgansisoftenquiteunmistakeable：forinstancetherearebeetlesofthesamegenus(andevenofthesamespecies)

resemblingeachothermostcloselyinallrespects，oneofwhichwillhavefull-sizedwings，andanothermererudimentsofmembrane。andhereitisimpossibletodoubt，thattherudimentsrepresentwings。Rudimentaryorganssometimesretaintheirpotentiality，andaremerelynotdeveloped：thisseemstobethecasewiththemammaeofmalemammals，formanyinstancesareonrecordoftheseorganshavingbecomewelldevelopedinfull-grownmales，andhavingsecretedmilk。SoagaintherearenormallyfourdevelopedandtworudimentaryteatsintheuddersofthegenusBos，butinourdomesticcowsthetwosometimesbecomedevelopedandgivemilk。Inindividualplantsofthesamespeciesthepetalssometimesoccurasmererudiments，andsometimesinawell-developedstate。Inplantswithseparatedsexes，themaleflowersoftenhavearudimentofapistil。andKö。lreuterfoundthatbycrossingsuchmaleplantswithanhermaphroditespecies，therudimentofthepistilinthehybridoffspringwasmuchincreasedinsize。andthisshowsthattherudimentandtheperfectpistilareessentiallyalikeinnature。

Anorganservingfortwopurposes，maybecomerudimentaryorutterlyabortedforone，eventhemoreimportantpurpose。，andremainperfectlyefficientfortheother。Thusinplants，theofficeofthepistilistoallowthepollen-tubestoreachtheovulesprotectedintheovariumatitsbase。Thepistilconsistsofastigmasupportedonthestyle。butinsomeCompositae，themaleflorets，whichofcoursecannotbefecundated，haveapistil，whichisinarudimentarystate，foritisnotcrownedwithastigma。butthestyleremainswelldeveloped，andisclothedwithhairsasinothercompositae，forthepurposeofbrushingthepollenoutofthesurroundinganthers。Again，anorganmaybecomerudimentaryforitsproperpurpose，andbeusedforadistinctobject：incertainfishtheswim-bladderseemstoberudimentaryforitsproperfunctionofgivingbuoyancy，buthasbecomeconvertedintoanascentbreathingorganorlung。Othersimilarinstancescouldbegiven。

Rudimentaryorgansintheindividualsofthesamespeciesareveryliabletovaryindegreeofdevelopmentandinotherrespects。Moreover，incloselyalliedspecies，thedegreetowhichthesameorganhasbeenrenderedrudimentaryoccasionallydiffersmuch。Thislatterfactiswellexemplifiedinthestateofthewingsofthefemalemothsincertaingroups。Rudimentaryorgansmaybeutterlyaborted。andthisimplies，thatwefindinananimalorplantnotraceofanorgan，whichanalogywouldleadustoexpecttofind，andwhichisoccasionallyfoundinmonstrousindividualsofthespecies。

Thusinthesnapdragon(antirrhinum)wegenerallydonotfindarudimentofafifthstamen。butthismaysometimesbeseen。Intracingthehomologiesofthesamepartindifferentmembersofaclass，nothingismorecommon，ormorenecessary，thantheuseanddiscoveryofrudiments。ThisiswellshowninthedrawingsgivenbyOwenofthebonesofthelegofthehorse，ox，andrhinoceros。

Itisanimportantfactthatrudimentaryorgans，suchasteethintheupperjawsofwhalesandruminants，canoftenbedetectedintheembryo，butafterwardswhollydisappear。Itisalso，Ibelieve，auniversalrule，thatarudimentarypartororganisofgreatersizerelativelytotheadjoiningpartsintheembryo，thanintheadult。sothattheorganatthisearlyageislessrudimentary，orevencannotbesaidtobeinanydegreerudimentary。

Hence，also，arudimentaryorganintheadult，isoftensaidtohaveretaineditsembryoniccondition。

Ihavenowgiventheleadingfactswithrespecttorudimentaryorgans。

Inreflectingonthem，everyonemustbestruckwithastonishment：forthesamereasoningpowerwhichtellsusplainlythatmostpartsandorgansareexquisitelyadaptedforcertainpurposes，tellsuswithequalplainnessthattheserudimentaryoratrophiedorgans，areimperfectanduseless。

Inworksonnaturalhistoryrudimentaryorgansaregenerallysaidtohavebeencreated`forthesakeofsymmetry，’orinorder`tocompletetheschemeofnature。’butthisseemstomenoexplanation，merelyarestatementofthefact。Woulditbethoughtsufficienttosaythatbecauseplanetsrevolveinellipticcoursesroundthesun，satellitesfollowthesamecourseroundtheplanets，forthesakeofsymmetry，andtocompletetheschemeofnature？

Aneminentphysiologistaccountsforthepresenceofrudimentaryorgans，bysupposingthattheyservetoexcretematterinexcess，orinjurioustothesystem。butcanwesupposethattheminutepapilla，whichoftenrepresentsthepistilinmaleflowers，andwhichisformedmerelyofcellulartissue，canthusact？Canwesupposethattheformationofrudimentaryteethwhicharesubsequentlyabsorbed，canbeofanyservicetotherapidlygrowingembryoniccalfbytheexcretionofpreciousphosphateoflime？

Whenaman’sfingershavebeenamputated，imperfectnailssometimesappearonthestumps：Icouldassoonbelievethatthesevestigesofnailshaveappeared，notfromunknownlawsofgrowth，butinordertoexcretehornymatter，asthattherudimentarynailsonthefinofthemanateewereformedforthispurpose。

Onmyviewofdescentwithmodification，theoriginofrudimentaryorgansissimple。Wehaveplentyofcasesofrudimentaryorgansinourdomesticproductions，asthestumpofatailintaillessbreeds，thevestigeofanearinearlessbreeds，——thereappearanceofminutedanglinghornsinhornlessbreedsofcattle，moreespecially，accordingtoYouatt，inyounganimals，andthestateofthewholeflowerinthecauliflower。Weoftenseerudimentsofvariouspartsinmonsters。ButIdoubtwhetheranyofthesecasesthrowlightontheoriginofrudimentaryorgansinastateofnature，furtherthanbyshowingthatrudimentscanbeproduced。forIdoubtwhetherspeciesundernatureeverundergoabruptchanges。Ibelievethatdisusehasbeenthemainagency。thatithasledinsuccessivegenerationstothegradualreductionofvariousorgans，untiltheyhavebecomerudimentary，asinthecaseoftheeyesofanimalsinhabitingdarkcaverns，andofthewingsofbirdsinhabitingoceanicislands，whichhaveseldombeenforcedtotakeflight，andhaveultimatelylostthepowerofflying。Again，anorganusefulundercertainconditions，mightbecomeinjuriousunderothers，aswiththewingsofbeetleslivingonsmallandexposedislands。andinthiscasenaturalselectionwouldcontinueslowlytoreducetheorgan，untilitwasrenderedharmlessandrudimentary。

Anychangeinfunction，whichcanbeeffectedbyinsensiblysmallsteps，iswithinthepowerofnaturalselection。sothatanorganrendered，duringchangedhabitsoflife，uselessorinjuriousforonepurpose，mighteasilybemodifiedandusedforanotherpurpose。Oranorganmightberetainedforonealoneofitsformerfunctions。Anorgan，whenrendereduseless，maywellbevariable，foritsvariationscannotbecheckedbynaturalselection。

Atwhateverperiodoflifedisuseorselectionreducesanorgan，andthiswillgenerallybewhenthebeinghascometomaturityandtoitsfullpowersofaction，theprincipleofinheritanceatcorrespondingageswillreproducetheorganinitsreducedstateatthesameage，andconsequentlywillseldomaffectorreduceitintheembryo。Thuswecanunderstandthegreaterrelativesizeofrudimentaryorgansintheembryo，andtheirlesserrelativesizeintheadult。Butifeachstepoftheprocessofreductionweretobeinherited，notatthecorrespondingage，butatanextremelyearlyperiodoflife(aswehavegoodreasontobelievetobepossible)therudimentarypartwouldtendtobewhollylost，andweshouldhaveacaseofcompleteabortion。

Theprinciple，also，ofeconomy，explainedinaformerchapter，bywhichthematerialsforminganypartorstructure，ifnotusefultothepossessor，willbesavedasfarasispossible，willprobablyoftencomeintoplay。

andthiswilltendtocausetheentireobliterationofarudimentaryorgan。

Asthepresenceofrudimentaryorgansisthusduetothetendencyineverypartoftheorganisation，whichhaslongexisted，tobeinheritedwecanunderstand，onthegenealogicalviewofclassification，howitisthatsystematistshavefoundrudimentarypartsasusefulas，orevensometimesmoreusefulthan，partsofhighphysiologicalimportance。Rudimentaryorgansmaybecomparedwiththelettersinaword，stillretainedinthespelling，butbecomeuselessinthepronunciation，butwhichserveasaclueinseekingforitsderivation。Ontheviewofdescentwithmodification，wemayconcludethattheexistenceoforgansinarudimentary，imperfect，anduselesscondition，orquiteaborted，farfrompresentingastrangedifficulty，astheyassuredlydoontheordinarydoctrineofcreation，mightevenhavebeenanticipated，andcanbeaccountedforbythelawsofinheritance。SummaryInthischapterIhaveattemptedtoshow，thatthesubordinationofgrouptogroupinallorganismsthroughoutalltime。thatthenatureoftherelationship，bywhichalllivingandextinctbeingsareunitedbycomplex，radiating，andcircuitouslinesofaffinitiesintoonegrandsystem。therulesfollowedandthedifficultiesencounteredbynaturalistsintheirclassifications。

thevaluesetuponcharacters，ifconstantandprevalent，whetherofhighvitalimportance，orofthemosttriflingimportance，or，asinrudimentaryorgans，ofnoimportance。thewideoppositioninvaluebetweenanalogicaloradaptivecharacters，andcharactersoftrueaffinity。andothersuchrules。allnaturallyfollowontheviewofthecommonparentageofthoseformswhichareconsideredbynaturalistsasallied，togetherwiththeirmodificationthroughnaturalselection，withitscontingenciesofextinctionanddivergenceofcharacter。Inconsideringthisviewofclassification，itshouldbeborneinmindthattheelementofdescenthasbeenuniversallyusedinrankingtogetherthesexes，ages，andacknowledgedvarietiesofthesamespecies，howeverdifferenttheymaybeinstructure。Ifweextendtheuseofthiselementofdescent，theonlycertainlyknowncauseofsimilarityinorganicbeings，weshallunderstandwhatismeantbythenaturalsystem：

itisgenealogicalinitsattemptedarrangement，withthegradesofacquireddifferencemarkedbythetermsvarieties，species，genera，families，orders，andclasses。

Onthissameviewofdescentwithmodification，allthegreatfactsinMorphologybecomeintelligible，whetherwelooktothesamepatterndisplayedinthehomologousorgans，towhateverpurposeapplied，ofthedifferentspeciesofaclass。ortothehomologouspartsconstructedonthesamepatternineachindividualanimalandplant。

Ontheprincipleofsuccessiveslightvariations，notnecessarilyorgenerallysuperveningataveryearlyperiodoflife，andbeinginheritedatacorrespondingperiod，wecanunderstandthegreatleadingfactsinEmbryology。namely，theresemblanceinanindividualembryoofthehomologousparts，whichwhenmaturedwillbecomewidelydifferentfromeachotherinstructureandfunction。andtheresemblanceindifferentspeciesofaclassofthehomologouspartsororgans，thoughfittedintheadultmembersforpurposesasdifferentaspossible。Larvaeareactiveembryos，whichhavebecomespeciallymodifiedinrelationtotheirhabitsoflife，throughtheprincipleofmodificationsbeinginheritedatcorrespondingages。Onthissameprincipleandbearinginmind，thatwhenorgansarereducedinsize，eitherfromdisuseorselection，itwillgenerallybeatthatperiodoflifewhenthebeinghastoprovideforitsownwants，andbearinginmindhowstrongistheprincipleofinheritancetheoccurrenceofrudimentaryorgansandtheirfinalabortion，presenttousnoinexplicabledifficulties。

onthecontrary，theirpresencemighthavebeenevenanticipated。Theimportanceofembryologicalcharactersandofrudimentaryorgansinclassificationisintelligible，ontheviewthatanarrangementisonlysofarnaturalasitisgenealogical。

Finally，theseveralclassesoffactswhichhavebeenconsideredinthischapter，seemtometoproclaimsoplainly，thattheinnumerablespecies，genera，andfamiliesoforganicbeings，withwhichthisworldispeopled，havealldescended，eachwithinitsownclassorgroup，fromcommonparents，andhaveallbeenmodifiedinthecourseofdescent，thatIshouldwithouthesitationadoptthisview，evenifitwereunsupportedbyotherfactsorarguments。PreviousChapterNextChapterTheOriginofSpecies-Chapter14TheOriginofSpeciesbyCharlesDarwinPreviousChapterChapter14-RecapitulationandConclusionAsthiswholevolumeisonelongargument，itmaybeconvenienttothereadertohavetheleadingfactsandinferencesbrieflyrecapitulated。

Thatmanyandgraveobjectionsmaybeadvancedagainstthetheoryofdescentwithmodificationthroughnaturalselection，Idonotdeny。Ihaveendeavouredtogivetothemtheirfullforce。Nothingatfirstcanappearmoredifficulttobelievethanthatthemorecomplexorgansandinstinctsshouldhavebeenperfectednotbymeanssuperiorto，thoughanalogouswith，humanreason，butbytheaccumulationofinnumerableslightvariations，eachgoodfortheindividualpossessor。Nevertheless，thisdifficulty，thoughappearingtoourimaginationinsuperablygreat，cannotbeconsideredrealifweadmitthefollowingpropositions，namely，——thatgradationsintheperfectionofanyorganorinstinct，whichwemayconsider，eitherdonowexistorcouldhaveexisted，eachgoodofitskind，——thatallorgansandinstinctsare，ineversoslightadegree，variable，——and，lastly，thatthereisastruggleforexistenceleadingtothepreservationofeachprofitabledeviationofstructureorinstinct。Thetruthofthesepropositionscannot，Ithink，bedisputed。

Itis，nodoubt，extremelydifficulteventoconjecturebywhatgradationsmanystructureshavebeenperfected，moreespeciallyamongstbrokenandfailinggroupsoforganicbeings。butweseesomanystrangegradationsinnature，asisproclaimedbythecanon，`Naturanonfacitsaltum，’thatweoughttobeextremelycautiousinsayingthatanyorganorinstinct，oranywholebeing，couldnothavearrivedatitspresentstatebymanygraduatedsteps。Thereare，itmustbeadmitted，casesofspecialdifficultyonthetheoryofnaturalselection。andoneofthemostcuriousoftheseistheexistenceoftwoorthreedefinedcastesofworkersorsterilefemalesinthesamecommunityofantsbutIhaveattemptedtoshowhowthisdifficultycanbemastered。Withrespecttothealmostuniversalsterilityofspecieswhenfirstcrossed，whichformssoremarkableacontrastwiththealmostuniversalfertilityofvarietieswhencrossed，Imustreferthereadertotherecapitulationofthefactsgivenattheendoftheeighthchapter，whichseemtomeconclusivelytoshowthatthissterilityisnomoreaspecialendowmentthanistheincapacityoftwotreestobegraftedtogether，butthatitisincidentalonconstitutionaldifferencesinthereproductivesystemsoftheintercrossedspecies。Weseethetruthofthisconclusioninthevastdifferenceintheresult，whenthesametwospeciesarecrossedreciprocally。thatis，whenonespeciesisfirstusedasthefatherandthenasthemother。

Thefertilityofvarietieswhenintercrossedandoftheirmongreloffspringcannotbeconsideredasuniversal。noristheirverygeneralfertilitysurprisingwhenwerememberthatitisnotlikelythateithertheirconstitutionsortheirreproductivesystemsshouldhavebeenprofoundlymodified。Moreover，mostofthevarietieswhichhavebeenexperimentisedonhavebeenproducedunderdomestication。andasdomesticationapparentlytendstoeliminatesterility，weoughtnottoexpectitalsotoproducesterility。

Thesterilityofhybridsisaverydifferentcasefromthatoffirstcrosses，fortheirreproductiveorgansaremoreorlessfunctionallyimpotent。

whereasinfirstcrossestheorgansonbothsidesareinaperfectcondition。

Aswecontinuallyseethatorganismsofallkindsarerenderedinsomedegreesterilefromtheirconstitutionshavingbeendisturbedbyslightlydifferentandnewconditionsoflife，weneednotfeelsurpriseathybridsbeinginsomedegreesterile，fortheirconstitutionscanhardlyfailtohavebeendisturbedfrombeingcompoundedoftwodistinctorganisations。

Thisparallelismissupportedbyanotherparallel，butdirectlyopposite，classoffacts。namely，thatthevigourandfertilityofallorganicbeingsareincreasedbyslightchangesintheirconditionsoflife，andthattheoffspringofslightlymodifiedformsorvarietiesacquirefrombeingcrossedincreasedvigourandfertility。Sothat，ontheonehand，considerablechangesintheconditionsoflifeandcrossesbetweengreatlymodifiedforms，lessenfertility。andontheotherhand，lesserchangesintheconditionsoflifeandcrossesbetweenlessmodifiedforms，increasefertility。

Turningtogeographicaldistribution，thedifficultiesencounteredonthetheoryofdescentwithmodificationaregraveenough。Alltheindividualsofthesamespecies，andallthespeciesofthesamegenus，orevenhighergroup，musthavedescendedfromcommonparents。andtherefore，inhoweverdistantandisolatedpartsoftheworldtheyarenowfound，theymustinthecourseofsuccessivegenerationshavepassedfromsomeoneparttotheothers。Weareoftenwhollyunableeventoconjecturehowthiscouldhavebeeneffected。Yet，aswehavereasontobelievethatsomespecieshaveretainedthesamespecificformforverylongperiods，enormouslylongasmeasuredbyyears，toomuchstressoughtnottobelaidontheoccasionalwidediffusionofthesamespecies。forduringverylongperiodsoftimetherewillalwaysbeagoodchanceforwidemigrationbymanymeans。

Abrokenorinterruptedrangemayoftenbeaccountedforbytheextinctionofthespeciesintheintermediateregions。Itcannotbedeniedthatweareasyetveryignorantofthefullextentofthevariousclimatalandgeographicalchangeswhichhaveaffectedtheearthduringmodernperiods。

andsuchchangeswillobviouslyhavegreatlyfacilitatedmigration。Asanexample，IhaveattemptedtoshowhowpotenthasbeentheinfluenceoftheGlacialperiodonthedistributionbothofthesameandofrepresentativespeciesthroughouttheworld。Weareasyetprofoundlyignorantofthemanyoccasionalmeansoftransport。Withrespecttodistinctspeciesofthesamegenusinhabitingverydistantandisolatedregions，astheprocessofmodificationhasnecessarilybeenslow，allthemeansofmigrationwillhavebeenpossibleduringaverylongperiod。andconsequentlythedifficultyofthewidediffusionofspeciesofthesamegenusisinsomedegreelessened。

Asonthetheoryofnaturalselectionaninterminablenumberofintermediateformsmusthaveexisted，linkingtogetherallthespeciesineachgroupbygradationsasfineasourpresentvarieties，itmaybeasked，Whydowenotseetheselinkingformsallaroundus？Whyarenotallorganicbeingsblendedtogetherinaninextricablechaos？Withrespecttoexistingforms，weshouldrememberthatwehavenorighttoexpect(exceptinginrarecases)

todiscoverdirectlyconnectinglinksbetweenthem，butonlybetweeneachandsomeextinctandsupplantedform。Evenonawidearea，whichhasduringalongperiodremainedcontinuous，andofwhichtheclimateandotherconditionsoflifechangeinsensiblyingoingfromadistrictoccupiedbyonespeciesintoanotherdistrictoccupiedbyacloselyalliedspecies，wehavenojustrighttoexpectoftentofindintermediatevarietiesintheintermediatezone。Forwehavereasontobelievethatonlyafewspeciesareundergoingchangeatanyoneperiod。andallchangesareslowlyeffected。

Ihavealsoshownthattheintermediatevarietieswhichwillatfirstprobablyexistintheintermediatezones，willbeliabletobesupplantedbythealliedformsoneitherhand。andthelatter，fromexistingingreaternumbers，willgenerallybemodifiedandimprovedataquickerratethantheintermediatevarieties，whichexistinlessernumbers。sothattheintermediatevarietieswill，inthelongrun，besupplantedandexterminated。

Onthisdoctrineoftheexterminationofaninfinitudeofconnectinglinks，betweenthelivingandextinctinhabitantsoftheworld，andateachsuccessiveperiodbetweentheextinctandstillolderspecies，whyisnoteverygeologicalformationchargedwithsuchlinks？Whydoesnoteverycollectionoffossilremainsaffordplainevidenceofthegradationandmutationoftheformsoflife？Wemeetwithnosuchevidence，andthisisthemostobviousandforcibleofthemanyobjectionswhichmaybeurgedagainstmytheory。Why，again，dowholegroupsofalliedspeciesappear，thoughcertainlytheyoftenfalselyappear，tohavecomeinsuddenlyontheseveralgeologicalstages？WhydowenotfindgreatpilesofstratabeneaththeSiluriansystem，storedwiththeremainsoftheprogenitorsoftheSiluriangroupsoffossils？Forcertainlyonmytheorysuchstratamustsomewherehavebeendepositedattheseancientandutterlyunknownepochsintheworld’shistory。

Icananswerthesequestionsandgraveobjectionsonlyonthesuppositionthatthegeologicalrecordisfarmoreimperfectthanmostgeologistsbelieve。

Itcannotbeobjectedthattherehasnotbeentimesufficientforanyamountoforganicchange。forthelapseoftimehasbeensogreatastobeutterlyinappreciablebythehumanintellect。Thenumberofspecimensinallourmuseumsisabsolutelyasnothingcomparedwiththecountlessgenerationsofcountlessspecieswhichcertainlyhaveexisted。Weshouldnotbeabletorecogniseaspeciesastheparentofanyoneormorespeciesifweweretoexaminethemeversoclosely，unlesswelikewisepossessedmanyoftheintermediatelinksbetweentheirpastorparentandpresentstates。andthesemanylinkswecouldhardlyeverexpecttodiscover，owingtotheimperfectionofthegeologicalrecord。Numerousexistingdoubtfulformscouldbenamedwhichareprobablyvarieties。butwhowillpretendthatinfutureagessomanyfossillinkswillbediscovered，thatnaturalistswillbeabletodecide，onthecommonview，whetherornotthesedoubtfulformsarevarieties？Aslongasmostofthelinksbetweenanytwospeciesareunknown，ifanyonelinkorintermediatevarietybediscovered，itwillsimplybeclassedasanotheranddistinctspecies。Onlyasmallportionoftheworldhasbeengeologicallyexplored。Onlyorganicbeingsofcertainclassescanbepreservedinafossilcondition，atleastinanygreatnumber。

Widelyrangingspeciesvarymost，andvarietiesareoftenatfirstlocal，——bothcausesrenderingthediscoveryofintermediatelinkslesslikely。

Localvarietieswillnotspreadintootheranddistantregionsuntiltheyareconsiderablymodifiedandimproved。andwhentheydospread，ifdiscoveredinageologicalformation，theywillappearasifsuddenlycreatedthere，andwillbesimplyclassedasnewspecies。Mostformationshavebeenintermittentintheiraccumulation。andtheirduration，Iaminclinedtobelieve，hasbeenshorterthantheaveragedurationofspecificforms。Successiveformationsareseparatedfromeachotherbyenormousblankintervalsoftime。forfossiliferousformations，thickenoughtoresistfuturedegradation，canbeaccumulatedonlywheremuchsedimentisdepositedonthesubsidingbedofthesea。Duringthealternateperiodsofelevationandofstationaryleveltherecordwillbeblank。Duringtheselatterperiodstherewillprobablybemorevariabilityintheformsoflife。duringperiodsofsubsidence，moreextinction。

WithrespecttotheabsenceoffossiliferousformationsbeneaththelowestSilurianstrata，Icanonlyrecurtothehypothesisgivenintheninthchapter。Thatthegeologicalrecordisimperfectallwilladmit。

butthatitisimperfecttothedegreewhichIrequire，fewwillbeinclinedtoadmit。Ifwelooktolongenoughintervalsoftime，geologyplainlydeclaresthatallspecieshavechanged。andtheyhavechangedinthemannerwhichmytheoryrequires，fortheyhavechangedslowlyandinagraduatedmanner。Weclearlyseethisinthefossilremainsfromconsecutiveformationsinvariablybeingmuchmorecloselyrelatedtoeachother，thanarethefossilsfromformationsdistantfromeachotherintime。

Suchisthesumoftheseveralchiefobjectionsanddifficultieswhichmayjustlybeurgedagainstmytheory。andIhavenowbrieflyrecapitulatedtheanswersandexplanationswhichcanbegiventothem。Ihavefeltthesedifficultiesfartooheavilyduringmanyyearstodoubttheirweight。Butitdeservesespecialnoticethatthemoreimportantobjectionsrelatetoquestionsonwhichweareconfessedlyignorant。nordoweknowhowignorantweare。Wedonotknowallthepossibletransitionalgradationsbetweenthesimplestandthemostperfectorgans。itcannotbepretendedthatweknowallthevariedmeansofDistributionduringthelonglapseofyears，orthatweknowhowimperfecttheGeologicalRecordis。Graveastheseseveraldifficultiesare，inmyjudgementtheydonotoverthrowthetheoryofdescentwithmodification。

Nowletusturntotheothersideoftheargument。Underdomesticationweseemuchvariability。Thisseemstobemainlyduetothereproductivesystembeingeminentlysusceptibletochangesintheconditionsoflifesothatthissystem，whennotrenderedimpotent，failstoreproduceoffspringexactlyliketheparent-form。Variabilityisgovernedbymanycomplexlaws，——bycorrelationofgrowth，byuseanddisuse，andbythedirectactionofthephysicalconditionsoflife。Thereismuchdifficultyinascertaininghowmuchmodificationourdomesticproductionshaveundergone。butwemaysafelyinferthattheamounthasbeenlarge，andthatmodificationscanbeinheritedforlongperiods。Aslongastheconditionsofliferemainthesame，wehavereasontobelievethatamodification，whichhasalreadybeeninheritedformanygenerations，maycontinuetobeinheritedforanalmostinfinitenumberofgenerations。Ontheotherhandwehaveevidencethatvariability，whenithasoncecomeintoplay，doesnotwhollycease。

fornewvarietiesarestilloccasionallyproducedbyourmostancientlydomesticatedproductions。

Mandoesnotactuallyproducevariability。heonlyunintentionallyexposesorganicbeingstonewconditionsoflife，andthennatureactsontheorganisation，andcausesvariability。Butmancananddoesselectthevariationsgiventohimbynature，andthusaccumulatetheminanydesiredmanner。Hethusadaptsanimalsandplantsforhisownbenefitorpleasure。Hemaydothismethodically，orhemaydoitunconsciouslybypreservingtheindividualsmostusefultohimatthetime，withoutanythoughtofalteringthebreed。

Itiscertainthathecanlargelyinfluencethecharacterofabreedbyselecting，ineachsuccessivegeneration，individualdifferencessoslightastobequiteinappreciablebyanuneducatedeye。Thisprocessofselectionhasbeenthegreatagencyintheproductionofthemostdistinctandusefuldomesticbreeds。Thatmanyofthebreedsproducedbymanhavetoalargeextentthecharacterofnaturalspecies，isshownbytheinextricabledoubtswhetherverymanyofthemarevarietiesoraboriginalspecies。

Thereisnoobviousreasonwhytheprincipleswhichhaveactedsoefficientlyunderdomesticationshouldnothaveactedundernature。Inthepreservationoffavouredindividualsandraces，duringtheconstantly-recurrentStruggleforExistence，weseethemostpowerfulandever-actingmeansofselection。

Thestruggleforexistenceinevitablyfollowsfromthehighgeometricalratioofincreasewhichiscommontoallorganicbeings。Thishighrateofincreaseisprovedbycalculation，bytheeffectsofasuccessionofpeculiarseasons，andbytheresultsofnaturalisation，asexplainedinthethirdchapter。Moreindividualsarebornthancanpossiblysurvive。

Agraininthebalancewilldeterminewhichindividualshallliveandwhichshalldie，——whichvarietyorspeciesshallincreaseinnumber，andwhichshalldecrease，orfinallybecomeextinct。Astheindividualsofthesamespeciescomeinallrespectsintotheclosestcompetitionwitheachother，thestrugglewillgenerallybemostseverebetweenthem。itwillbealmostequallyseverebetweenthevarietiesofthesamespecies，andnextinseveritybetweenthespeciesofthesamegenus。Butthestrugglewilloftenbeveryseverebetweenbeingsmostremoteinthescaleofnature。Theslightestadvantageinonebeing，atanyageorduringanyseason，overthosewithwhichitcomesintocompetition，orbetteradaptationinhoweverslightadegreetothesurroundingphysicalconditions，willturnthebalance。

Withanimalshavingseparatedsexestherewillinmostcasesbeastrugglebetweenthemalesforpossessionofthefemales。Themostvigorousindividuals，orthosewhichhavemostsuccessfullystruggledwiththeirconditionsoflife，willgenerallyleavemostprogeny。Butsuccesswilloftendependonhavingspecialweaponsormeansofdefence，oronthecharmsofthemales。andtheslightestadvantagewillleadtovictory。

Asgeologyplainlyproclaimsthateachlandhasundergonegreatphysicalchanges，wemighthaveexpectedthatorganicbeingswouldhavevariedundernature，inthesamewayastheygenerallyhavevariedunderthechangedconditionsofdomestication。Andiftherebeanyvariabilityundernature，itwouldbeanunaccountablefactifnaturalselectionhadnotcomeintoplay。Ithasoftenbeenasserted，buttheassertionisquiteincapableofproof，thattheamountofvariationundernatureisastrictlylimitedquantity。Man，thoughactingonexternalcharactersaloneandoftencapriciously，canproducewithinashortperiodagreatresultbyaddingupmereindividualdifferencesinhisdomesticproductions。andeveryoneadmitsthatthereareatleastindividualdifferencesinspeciesundernature。But，besidessuchdifferences，allnaturalistshaveadmittedtheexistenceofvarieties，whichtheythinksufficientlydistincttobeworthyofrecordinsystematicworks。Noonecandrawanycleardistinctionbetweenindividualdifferencesandslightvarieties。orbetweenmoreplainlymarkedvarietiesandsubspecies，andspecies。LetitbeobservedhownaturalistsdifferintherankwhichtheyassigntothemanyrepresentativeformsinEuropeandNorthAmerica。

Ifthenwehaveundernaturevariabilityandapowerfulagentalwaysreadytoactandselect，whyshouldwedoubtthatvariationsinanywayusefultobeings，undertheirexcessivelycomplexrelationsoflife，wouldbepreserved，accumulated，andinherited？Why，ifmancanbypatienceselectvariationsmostusefultohimself，shouldnaturefailinselectingvariationsuseful，underchangingconditionsoflife，toherlivingproducts？Whatlimitcanbeputtothispower，actingduringlongagesandrigidlyscrutinisingthewholeconstitution，structure，andhabitsofeachcreature，&。mdash。

favouringthegoodandrejectingthebad？Icanseenolimittothispower，inslowlyandbeautifullyadaptingeachformtothemostcomplexrelationsoflife。Thetheoryofnaturalselection，evenifwelookednofurtherthanthis，seemstometobeinitselfprobable。Ihavealreadyrecapitulated，asfairlyasIcould，theopposeddifficultiesandobjections：nowletusturntothespecialfactsandargumentsinfavourofthetheory。

Ontheviewthatspeciesareonlystronglymarkedandpermanentvarieties，andthateachspeciesfirstexistedasavariety，wecanseewhyitisthatnolineofdemarcationcanbedrawnbetweenspecies，commonlysupposedtohavebeenproducedbyspecialactsofcreation，andvarietieswhichareacknowledgedtohavebeenproducedbysecondarylaws。Onthissameviewwecanunderstandhowitisthatineachregionwheremanyspeciesofagenushavebeenproduced，andwheretheynowflourish，thesesamespeciesshouldpresentmanyvarieties。forwherethemanufactoryofspecieshasbeenactive，wemightexpect，asageneralrule，tofinditstillinaction。andthisisthecaseifvarietiesbeincipientspecies。Moreover，thespeciesofthelargegenera，whichaffordthegreaternumberofvarietiesorincipientspecies，retaintoacertaindegreethecharacterofvarieties。

fortheydifferfromeachotherbyalessamountofdifferencethandothespeciesofsmallergenera。Thecloselyalliedspeciesalsoofthelargergeneraapparentlyhaverestrictedranges，andtheyareclusteredinlittlegroupsroundotherspecies——inwhichrespectstheyresemblevarieties。

Thesearestrangerelationsontheviewofeachspecieshavingbeenindependentlycreated，butareintelligibleifallspeciesfirstexistedasvarieties。

Aseachspeciestendsbyitsgeometricalratioofreproductiontoincreaseinordinatelyinnumber。andasthemodifieddescendantsofeachspecieswillbeenabledtoincreasebysomuchthemoreastheybecomemorediversifiedinhabitsandstructure，soastobeenabledtoseizeonmanyandwidelydifferentplacesintheeconomyofnature，therewillbeaconstanttendencyinnaturalselectiontopreservethemostdivergentoffspringofanyonespecies。Henceduringalong-continuedcourseofmodification，theslightdifferences，characteristicofvarietiesofthesamespecies，tendtobeaugmentedintothegreaterdifferencescharacteristicofspeciesofthesamegenus。Newandimprovedvarietieswillinevitablysupplantandexterminatetheolder，lessimprovedandintermediatevarieties。andthusspeciesarerenderedtoalargeextentdefinedanddistinctobjects。Dominantspeciesbelongingtothelargergroupstendtogivebirthtonewanddominantforms。

sothateachlargegrouptendstobecomestilllarger，andatthesametimemoredivergentincharacter。Butasallgroupscannotthussucceedinincreasinginsize，fortheworldwouldnotholdthem，themoredominantgroupsbeatthelessdominant。Thistendencyinthelargegroupstogoonincreasinginsizeanddivergingincharacter，togetherwiththealmostinevitablecontingencyofmuchextinction，explainsthearrangementofalltheformsoflife，ingroupssubordinatetogroups，allwithinafewgreatclasses，whichwenowseeeverywherearoundus，andwhichhasprevailedthroughoutalltime。Thisgrandfactofthegroupingofallorganicbeingsseemstomeutterlyinexplicableonthetheoryofcreation。

Asnaturalselectionactssolelybyaccumulatingslight，successive，favourablevariations，itcanproducenogreatorsuddenmodification。

itcanactonlybyveryshortandslowsteps。Hencethecanonof`Naturanonfacitsaltum，’whicheveryfreshadditiontoourknowledgetendstomakemorestrictlycorrect，isonthistheorysimplyintelligible。Wecanplainlyseewhynatureisprodigalinvariety，thoughniggardininnovation。