’QuarterlyReview’1849page392。），insomeotherpoints:butthereisnoreasontobelievethatthissplit，theoriginofwhichisunknown，wasintentionallymade;itmaywithmuchmoreprobabilitybeattributedtodifferentbreedershavingattendedtodifferentpoints。Soagain，theBerkshirebreedofswineintheyear1810hadgreatlychangedfromwhatitwasin1780;andsince1810atleasttwodistinctsub—breedshavearisenbearingthesamename。（20/80。H。vonNathusius’Vorstudien……Schweineschadel’1864s140。）Keepinginmindhowrapidlyallanimalsincrease，andthatsomemustbeannuallyslaughteredandsomesavedforbreeding，then，ifthesamebreederduringalongcourseofyearsdeliberatelysettleswhichshallbesavedandwhichshallbekilled，itisalmostinevitablethathisindividualturnofmindwillinfluencethecharacterofhisstock，withouthishavinghadanyintentiontomodifythebreed。
Unconsciousselectioninthestrictestsenseoftheword，thatis，thesavingofthemoreusefulanimalsandtheneglectorslaughterofthelessuseful，withoutanythoughtofthefuture，musthavegoneonoccasionallyfromtheremotestperiodandamongstthemostbarbarousnations。Savagesoftensufferfromfamines，andaresometimesexpelledbywarfromtheirownhomes。Insuchcasesitcanhardlybedoubtedthattheywouldsavetheirmostusefulanimals。
WhentheFuegiansarehardpressedbywant，theykilltheiroldwomenforfoodratherthantheirdogs;for，aswewereassured，\"oldwomennouse——dogscatchotters。\"Thesamesoundsensewouldsurelyleadthemtopreservetheirmoreusefuldogswhenstillharderpressedbyfamine。Mr。Oldfield，whohasseensomuchoftheaboriginesofAustralia，informsmethat\"theyareallverygladtogetaEuropeankangaroodog，andseveralinstanceshavebeenknownofthefatherkillinghisowninfantthatthemothermightsucklethemuch—prizedpuppy。\"DifferentkindsofdogswouldbeusefultotheAustralianforhuntingopossumsandkangaroos，andtotheFuegianforcatchingfishandotters;andtheoccasionalpreservationinthetwocountriesofthemostusefulanimalswouldultimatelyleadtotheformationoftwowidelydistinctbreeds。
Withplants，fromtheearliestdawnofcivilisation，thebestvarietywhichwasknownwouldgenerallyhavebeencultivatedateachperiodanditsseedsoccasionallysown;sothattherewillhavebeensomeselectionfromanextremelyremoteperiod，butwithoutanyprefixedstandardofexcellenceorthoughtofthefuture。Weatthepresentdayprofitbyacourseofselectionoccasionallyandunconsciouslycarriedonduringthousandsofyears。ThisisprovedinaninterestingmannerbyOswaldHeer’sresearchesonthelake—
inhabitantsofSwitzerland，asgiveninaformerchapter;forheshowsthatthegrainandseedofourpresentvarietiesofwheat，barley，oats，peas，beans，lentils，andpoppy，exceedinsizethosewhichwerecultivatedinSwitzerlandduringtheNeolithicandBronzeperiods。Theseancientpeople，duringtheNeolithicperiod，possessedalsoacrabconsiderablylargerthanthatnowgrowingwildontheJura。（20/81。SeealsoDr。ChristinRutimeyer’s’Pfahlbauten’1861s。226。）ThepearsdescribedbyPlinywereevidentlyextremelyinferiorinqualitytoourpresentpears。Wecanrealisetheeffectsoflong—continuedselectionandcultivationinanotherway，forwouldanyoneinhissensesexpecttoraiseafirst—rateapplefromtheseedofatrulywildcrab，oralusciousmeltingpearfromthewildpear？AlphonsedeCandolleinformsmethathehaslatelyseenonanancientmosaicatRomearepresentationofthemelon;andastheRotnans，whoweresuchgourmands，aresilentonthisfruit，heinfersthatthemelonhasbeengreatlyamelioratedsincetheclassicalperiod。
Comingtolatertimes，Buffon（20/82。Thepassageisgiven’Bull。Soc。
d’Acclimat。’1858page11。）oncomparingtheflowers，fruit，andvegetableswhichwerethencultivatedwithsomeexcellentdrawingsmadeahundredandfiftyyearspreviously，wasstruckwithsurpriseatthegreatimprovementwhichhadbeeneffected;andremarksthattheseancientflowersandvegetableswouldnowberejected，notonlybyafloristbutbyavillagegardener。SincethetimeofBuffontheworkofimprovementhassteadilyandrapidlygoneon。
Everyfloristwhocomparesourpresentflowerswiththosefiguredinbookspublishednotlongsince，isastonishedatthechange。Awell—knownamateur（20/83。’JournalofHorticulture’1862page394。），inspeakingofthevarietiesofPelargoniumraisedbyMr。Garthonlytwenty—twoyearsbefore，remarks，\"Whataragetheyexcited:surelywehadattainedperfection，itwassaid;andnownotoneoftheflowersofthosedayswillbelookedat。Butnonethelessisthedebtofgratitudewhichweowetothosewhosawwhatwastobedone，anddidit。\"Mr。Paul，thewell—knownhorticulturist，inwritingofthesameflower（20/84。’Gardener’sChronicle’1857page85。），saysherememberswhenyoungbeingdelightedwiththeportraitsinSweet’swork;\"butwhataretheyinpointofbeautycomparedwiththePelargoniumsofthisday？Hereagainnaturedidnotadvancebyleaps;theimprovementwasgradual，andifwehadneglectedthoseverygradualadvances，wemusthaveforegonethepresentgrandresults。\"Howwellthispracticalhorticulturistappreciatesandillustratesthegradualandaccumulativeforceofselection！TheDahliahasadvancedinbeautyinalikemanner;thelineofimprovementbeingguidedbyfashion，andbythesuccessivemodificationswhichtheflowerslowlyunderwent。（20/85。SeeMr。Wildman’saddresstotheFloricult。Soc。in’Gardener’sChronicle’1843
page86。）Asteadyandgradualchangehasbeennoticedinmanyotherflowers:
thusanoldflorist（20/86。’JournalofHorticulture’October24，1865page239。），afterdescribingtheleadingvarietiesofthePinkwhichweregrownin1813adds，\"thepinksofthosedayswouldnowbescarcelygrownasborder—
flowers。\"Theimprovementofsomanyflowersandthenumberofthevarietieswhichhavebeenraisedisallthemorestrikingwhenwehearthattheearliestknownflower—gardeninEurope，namelyatPadua，datesonlyfromtheyear1545。
（20/87。Prescott’Hist。ofMexico’volume2page61。）
EFFECTSOFSELECTION，ASSHOWNBYTHEPARTSMOSTVALUEDBYMANPRESENTINGTHE
GREATESTAMOUNTOFDIFFERENCE。
Thepoweroflong—continuedselection，whethermethodicalorunconscious，orbothcombined，iswellshowninageneralway，namely，bythecomparisonofthedifferencesbetweenthevarietiesofdistinctspecies，whicharevaluedfordifferentparts，suchasfortheleaves，orstems，ortubers，theseed，orfruit，orflowers。Whateverpartmanvaluesmost，thatpartwillbefoundtopresentthegreatestamountofdifference。Withtreescultivatedfortheirfruit，Sageretremarksthatthefruitislargerthanintheparent—species，whilstwiththosecultivatedfortheseed，aswithnuts，walnuts，almonds，chestnuts，etc。，itistheseeditselfwhichislarger;andheaccountsforthisfactbythefruitintheonecase，andbytheseedintheother，havingbeencarefullyattendedtoandselectedduringmanyages。Gallesiohasmadethesameobservation。Godroninsistsonthediversityofthetuberinthepotato，ofthebulbintheonion，andofthefruitinthemelon;andontheclosesimilarityoftheotherpartsinthesesameplants。（20/88。Sagaret’PomologiePhysiologique’1830page47;Gallesio’TeoriadellaRiproduzione’
1816page88;Godron’Del’Espece’1859tome2pages63，67，70。InmytenthandeleventhchaptersIhavegivendetailsonthepotato;andIcanconfirmsimilarremarkswithrespecttotheonion。IhavealsoshownhowfarNaudinconcursinregardtothevarietiesofthemelon。）
Inordertojudgehowfarmyownimpressiononthissubjectwascorrect，I
cultivatednumerousvarietiesofthesamespeciesclosetooneanother。Thecomparisonoftheamountofdifferencebetweenwidelydifferentorgansisnecessarilyvague;Iwillthereforegivetheresultsinonlyafewcases。Wehavepreviouslyseenintheninthchapterhowgreatlythevarietiesofthecabbagedifferintheirfoliageandstems，whicharetheselectedparts，andhowcloselytheyresembleoneanotherintheirflowers，capsules，andseeds。
Insevenvarietiesoftheradish，therootsdifferedgreatlyincolourandshape，butnodifferencewhatevercouldbedetectedintheirfoliage，flowers，orseeds。Nowwhatacontrastispresented，ifwecomparetheflowersofthevarietiesofthesetwoplantswiththoseofanyspeciescultivatedinourflower—gardensforornament;orifwecomparetheirseedswiththoseofthevarietiesofmaize，peas，beans，etc。，whicharevaluedandcultivatedfortheirseeds。Intheninthchapteritwasshownthatthevarietiesofthepeadifferbutlittleexceptinthetallnessoftheplant，moderatelyintheshapeofthepod，andgreatlyinthepeaitself，andtheseareallselectedpoints。
Thevarieties，however，ofthePoissansparchemindiffermuchmoreintheirpods，andtheseareeatenandvalued。Icultivatedtwelvevarietiesofthecommonbean;onealone，theDwarfFan，differedconsiderablyingeneralappearance;twodifferedinthecolouroftheirflowers，onebeinganalbino，andtheotherbeingwhollyinsteadofpartiallypurple;severaldifferedconsiderablyintheshapeandsizeofthepod，butfarmoreinthebeanitself，andthisisthevaluedandselectedpart。Toker’sbean，forinstance，istwice—and—a—halfaslongandbroadasthehorse—bean，andismuchthinnerandofadifferentshape。
Thevarietiesofthegooseberry，asformerlydescribed，differmuchintheirfruit，buthardlyperceptiblyintheirflowersororgansofvegetation。Withtheplum，thedifferenceslikewiseappeartobegreaterinthefruitthanintheflowersorleaves。Ontheotherhand，theseedofthestrawberry，whichcorrespondswiththefruitoftheplum，differshardlyatall;whilsteveryoneknowshowgreatlythefruit——thatis，theenlargedreceptacle——differsinseveralvarieties。Inapples，pears，andpeachestheflowersandleavesdifferconsiderably，butnot，asfarasIcanjudge，inproportionwiththefruit。
TheChinesedouble—floweringpeaches，ontheotherhand，showthatvarietiesofthistreehavebeenformed，whichdiffermoreinflowerthaninfruit。If，asishighlyprobable，thepeachisthemodifieddescentofthealmond，asurprisingamountofchangehasbeeneffectedinthesamespecies，inthefleshycoveringoftheformerandinthekernelsofthelatter。
Whenpartsstandincloserelationshiptoeachother，suchastheseedandthefleshycoveringofthefruit（whateveritshomologicalnaturemaybe），changesintheoneareusuallyaccompaniedbymodificationsintheother，thoughnotnecessarilytothesamedegree。Withtheplum—tree，forinstance，somevarietiesproduceplumswhicharenearlyalike，butincludestonesextremelydissimilarinshape;whilstconverselyothervarietiesproducedissimilarfruitwithbarelydistinguishablestones;andgenerallythestones，thoughtheyhaveneverbeensubjectedtoselection，differgreatlyintheseveralvarietiesoftheplum。Inothercasesorganswhicharenotmanifestlyrelated，throughsomeunknownbondvarytogether，andareconsequentlyliable，withoutanyintentiononman’spart，tobesimultaneouslyactedonbyselection。Thusthevarietiesofthestock（Matthiola）havebeenselectedsolelyforthebeautyoftheirflowers，buttheseedsdiffergreatlyincolourandsomewhatinsize。Varietiesofthelettucehavebeenselectedsolelyonaccountoftheirleaves，yetproduceseedswhichlikewisedifferincolour。Generally，throughthelawofcorrelation，whenavarietydiffersgreatlyfromitsfellow—varietiesinanyonecharacter，itdifferstoacertainextentinseveralothercharacters。IobservedthisfactwhenIcultivatedtogethermanyvarietiesofthesamespecies，forIusedfirsttomakealistofthevarietieswhichdifferedmostfromeachotherintheirfoliageandmannerofgrowth，afterwardsofthosethatdifferedmostintheirflowers，thenintheirseed—capsules，andlastlyintheirmatureseed;andIfoundthatthesamenamesgenerallyoccurredintwo，three，orfourofthesuccessivelists。
Neverthelessthegreatestamountofdifferencebetweenthevarietieswasalwaysexhibited，asfarasIcouldjudge，bythatpartororganforwhichtheplantwascultivated。
Whenwebearinmindthateachplantwasatfirstcultivatedbecauseusefultoman，andthatitsvariationwasasubsequent，oftenalongsubsequent，event，wecannotexplainthegreateramountofdiversityinthevaluablepartsbysupposingthatspeciesendowedwithanespecialtendencytovaryinanyparticularmannerwereoriginallychosen。Wemustattributetheresulttothevariationsinthesepartshavingbeensuccessivelypreserved，andthuscontinuallyaugmented;whilstothervariations，exceptingsuchasinevitablyappearedthroughcorrelation，wereneglectedandlost。Wemaythereforeinferthatmostplantsmightbemade，throughlong—continuedselection，toyieldracesasdifferentfromoneanotherinanycharacterastheynowareinthosepartsforwhichtheyarevaluedandcultivated。
Withanimalsweseenothingofthesamekind;butasufficientnumberofspecieshavenotbeendomesticatedforafaircomparison。Sheeparevaluedfortheirwool，andthewooldiffersmuchmoreintheseveralracesthanthehairincattle。Neithersheep，goats，Europeancattle，norpigsarevaluedfortheirfleetnessorstrength;andwedonotpossessbreedsdifferingintheserespectsliketheracehorseanddray—horse。Butfleetnessandstrengtharevaluedincamelsanddogs;andwehavewiththeformertheswiftdromedaryandheavycamel;withthelatterthegreyhoundandmastiff。Butdogsarevaluedeveninahigherdegreefortheirmentalqualitiesandsenses;andeveryoneknowshowgreatlytheracesdifferintheserespects。Ontheotherhand，wherethedogiskeptsolelytoserveforfood，asinthePolynesianislandsandChina，itisdescribedasanextremelystupidanimal。（20/89。Godron’Del’Espece’tome2page27。）Blumenbachremarksthat\"manydogs，suchasthebadger—dog，haveabuildsomarkedandsoappropriateforparticularpurposes，thatIshouldfinditverydifficulttopersuademyselfthatthisastonishingfigurewasanaccidentalconsequenceofdegeneration。\"（20/90。’TheAnthropologicalTreatisesofBlumenbach’1856page292。）HadBlumenbachreflectedonthegreatprincipleofselection，hewouldnothaveusedthetermdegeneration，andhewouldnothavebeenastonishedthatdogsandotheranimalsshouldbecomeexcellentlyadaptedfortheserviceofman。
Onthewholewemayconcludethatwhateverpartorcharacterismostvalued——
whethertheleaves，stems，tubers，bulbs，flowers，fruit，orseedofplants，orthesize，strength，fleetness，hairycovering，orintellectofanimals——
thatcharacterwillalmostinvariablybefoundtopresentthegreatestamountofdifferencebothinkindanddegree。Andthisresultmaybesafelyattributedtomanhavingpreservedduringalongcourseofgenerationsthevariationswhichwereusefultohim，andneglectedtheothers。
Iwillconcludethischapterbysomeremarksonanimportantsubject。Withanimalssuchasthegiraffe，ofwhichthewholestructureisadmirablyco—
ordinatedforcertainpurposes，ithasbeensupposedthatallthepartsmusthavebeensimultaneouslymodified;andithasbeenarguedthat，ontheprincipleofnaturalselection，thisisscarcelypossible。Butinthusarguing，ithasbeentacitlyassumedthatthevariationsmusthavebeenabruptandgreat。Nodoubt，iftheneckofaruminantweresuddenlytobecomegreatlyelongated，theforelimbsandbackwouldhavetobesimultaneouslystrengthenedandmodified;butitcannotbedeniedthatananimalmighthaveitsneck，orhead，ortongue，orfore—limbselongatedaverylittlewithoutanycorrespondingmodificationinotherpartsofthebody;andanimalsthusslightlymodifiedwould，duringadearth，haveaslightadvantage，andbeenabledtobrowseonhighertwigs，andthussurvive。Afewmouthfulsmoreorlesseverydaywouldmakeallthedifferencebetweenlifeanddeath。Bytherepetitionofthesameprocess，andbytheoccasionalintercrossingofthesurvivors，therewouldbesomeprogress，slowandfluctuatingthoughitwouldbe，towardstheadmirablycoordinatedstructureofthegiraffe。Iftheshort—
facedtumbler—pigeon，withitssmallconicalbeak，globularhead，roundedbody，shortwings，andsmallfeet——characterswhichappearallinharmony——hadbeenanaturalspecies，itswholestructurewouldhavebeenviewedaswellfittedforitslife;butinthiscaseweknowthatinexperiencedbreedersareurgedtoattendtopointafterpoint，andnottoattemptimprovingthewholestructureatthesametime。Lookatthegreyhound，thatperfectimageofgrace，symmetry，andvigour;nonaturalspeciescanboastofamoreadmirablyco—ordinatedstructure，withitstaperinghead，slimbody，deepchest，tucked—
upabdomen，rat—liketail，andlongmuscularlimbs，alladaptedforextremefleetness，andforrunningdownweakprey。Now，fromwhatweseeofthevariabilityofanimals，andfromwhatweknowofthemethodwhichdifferentmenfollowinimprovingtheirstock——somechieflyattendingtoonepoint，otherstoanotherpoint，othersagaincorrectingdefectsbycrosses，andsoforth——wemayfeelassuredthatifwecouldseethelonglineofancestorsofafirst—rategreyhounduptoitswildwolf—likeprogenitor，weshouldbeholdaninfinitenumberofthefinestgradations，sometimesinonecharacterandsometimesinanother，butallleadingtowardsourpresentperfecttype。Bysmallanddoubtfulstepssuchasthese，nature，aswemayconfidentlybelieve，hasprogressed，onhergrandmarchofimprovementanddevelopment。
Asimilarlineofreasoningisasapplicabletoseparateorgansastothewholeorganisation。Awriter（20/91。Mr。J。J。MurphyinhisopeningaddresstotheBelfastNat。Hist。Soc。asgiveninthe’BelfastNorthernWhig’November19，1866。Mr。MurphyherefollowsthelineofargumentagainstmyviewspreviouslyandmorecautiouslygivenbytheRev。C。Pritchard，Pres。RoyalAstronomicalSoc。，inhissermonAppendixpage33preachedbeforetheBritishAssociationatNottingham1866。）hasrecentlymaintainedthat\"itisprobablynoexaggerationtosupposethatinordertoimprovesuchanorganastheeyeatall，itmustbeimprovedintendifferentwaysatonce。Andtheimprobabilityofanycomplexorganbeingproducedandbroughttoperfectioninanysuchwayisanimprobabilityofthesamekindanddegreeasthatofproducingapoemoramathematicaldemonstrationbythrowinglettersatrandomonatable。\"Iftheeyewereabruptlyandgreatlymodified，nodoubtmanypartswouldhavetobesimultaneouslyaltered，inorderthattheorganshouldremainserviceable。
Butisthisthecasewithsmallerchanges？Therearepersonswhocanseedistinctlyonlyinadulllight，andthisconditiondepends，Ibelieve，ontheabnormalsensitivenessoftheretina，andisknowntobeinherited。Nowifabird，forinstance，receivesomegreatadvantagefromseeingwellinthetwilight，alltheindividualswiththemostsensitiveretinawouldsucceedbestandbethemostlikelytosurvive;andwhyshouldnotallthosewhichhappenedtohavetheeyeitselfalittlelarger，orthepupilcapableofgreaterdilatation，belikewisepreserved，whetherornotthesemodificationswerestrictlysimultaneous？Theseindividualswouldsubsequentlyintercrossandblendtheirrespectiveadvantages。Bysuchslightsuccessivechanges，theeyeofadiurnalbirdwouldbebroughtintotheconditionofthatofanowl，whichhasoftenbeenadvancedasanexcellentinstanceofadaptation。Short—
sight，whichisofteninherited，permitsapersontoseedistinctlyaminuteobjectatsonearadistancethatitwouldbeindistincttoordinaryeyes;andherewehaveacapacitywhichmightbeserviceableundercertainconditions，abruptlygained。TheFuegiansonboardtheBeaglecouldcertainlyseedistantobjectsmoredistinctlythanoursailorswithalltheirlongpractice;Idonotknowwhetherthisdependsuponsensitivenessoronthepowerofadjustmentinthefocus;butthiscapacityfordistantvisionmight，itisprobable，beslightlyaugmentedbysuccessivemodificationsofeitherkind。Amphibiousanimalswhichareenabledtoseebothinthewaterandintheair，requireandpossess，asM。Plateauhasshown（20/92。OntheVisionofFishesandAmphibia，translatedin’AnnalsandMag。ofNat。Hist。’volume181866page469。），eyesconstructedonthefollowingplan:\"thecorneaisalwaysflat，oratleastmuchflattenedinthefrontofthecrystallineandoveraspaceequaltothediameterofthatlens，whilstthelateralportionsmaybemuchcurved。\"Thecrystallineisverynearlyasphere，andthehumourshavenearlythesamedensityaswater。Nowasaterrestrialanimalbecamemoreandmoreaquaticinitshabits，veryslightchanges，firstinthecurvatureofthecorneaorcrystalline，andtheninthedensityofthehumours，orconversely，mightsuccessivelyoccur，andwouldbeadvantageoustotheanimalwhilstunderwater，withoutseriousdetrimenttoitspowerofvisionintheair。ItisofcourseimpossibletoconjecturebywhatstepsthefundamentalstructureoftheeyeintheVertebratawasoriginallyacquired，forweknownothingaboutthisorganinthefirstprogenitorsoftheclass。Withrespecttothelowestanimalsinthescale，thetransitionalstatesthroughwhichtheeyeatfirstprobablypassed，canbytheaidofanalogybeindicated，asIhaveattemptedtoshowinmy’OriginofSpecies。’（20/93。Sixthedition1872page144。）
CHAPTER2。XXI。
SELECTION，continued。
NATURALSELECTIONASAFFECTINGDOMESTICPRODUCTIONS。
CHARACTERSWHICHAPPEAROFTRIFLINGVALUEOFTENOFREALIMPORTANCE。
CIRCUMSTANCESFAVOURABLETOSELECTIONBYMAN。
FACILITYINPREVENTINGCROSSES，ANDTHENATUREOFTHECONDITIONS。
CLOSEATTENTIONANDPERSEVERANCEINDISPENSABLE。
THEPRODUCTIONOFALARGENUMBEROFINDIVIDUALSESPECIALLYFAVOURABLE。
WHENNOSELECTIONISAPPLIED，DISTINCTRACESARENOTFORMED。
HIGHLY—BREDANIMALSLIABLETODEGENERATION。
TENDENCYINMANTOCARRYTHESELECTIONOFEACHCHARACTERTOANEXTREMEPOINT，LEADINGTODIVERGENCEOFCHARACTER，RARELYTOCONVERGENCE。
CHARACTERSCONTINUINGTOVARYINTHESAMEDIRECTIONINWHICHTHEYHAVE
ALREADYVARIED。
DIVERGENCEOFCHARACTER，WITHTHEEXTINCTIONOFINTERMEDIATEVARIETIES，LEADS
TODISTINCTNESSINOURDOMESTICRACES。
LIMITTOTHEPOWEROFSELECTION。
LAPSEOFTIMEIMPORTANT。
MANNERINWHICHDOMESTICRACESHAVEORIGINATED。
SUMMARY。
NATURALSELECTION，ORTHESURVIVALOFTHEFITTEST，ASAFFECTINGDOMESTIC
PRODUCTIONS。
Weknowlittleonthishead。Butasanimalskeptbysavageshavetoprovidethroughouttheyeartheirownfoodeitherentirelyortoalargeextent，itcanhardlybedoubtedthatindifferentcountries，varietiesdifferinginconstitutionandinvariouscharacterswouldsucceedbest，andsobenaturallyselected。Henceperhapsitisthatthefewdomesticatedanimalskeptbysavagespartake，ashasbeenremarkedbymorethanonewriter，ofthewildappearanceoftheirmasters，andlikewiseresemblenaturalspecies。Eveninlong—civilisedcountries，atleastinthewilderparts，naturalselectionmustactonourdomesticraces。Itisobviousthatvarietieshavingverydifferenthabits，constitution，andstructure，wouldsucceedbestonmountainsandonrichlowlandpastures。Forexample，theimprovedLeicestersheepwereformerlytakentotheLammermuirHills;butanintelligentsheep—masterreportedthat\"ourcoarseleanpastureswereunequaltothetaskofsupportingsuchheavy—
bodiedsheep;andtheygraduallydwindledawayintolessandlessbulk:eachgenerationwasinferiortotheprecedingone;andwhenthespringwassevere，seldommorethantwo—thirdsofthelambssurvivedtheravagesofthestorms。\"
（21/1。QuotedbyYouatton’Sheep’page325。SeealsoYouatton’Cattle’pages62，69。）SowiththemountaincattleofNorthWalesandtheHebrides，ithasbeenfoundthattheycouldnotwithstandbeingcrossedwiththelargerandmoredelicatelowlandbreeds。TwoFrenchnaturalists，indescribingthehorsesofCircassia，remarkthat，subjectedastheyaretoextremevicissitudesofclimate，havingtosearchforscantypasture，andexposedtoconstantdangerfromwolves，thestrongestandmostvigorousalonesurvive。（21/2。MM。
LherbetteandDeQuatrefagesin’Bull。Soc。d’Acclimat。’tome81861page311。）
Everyonemusthavebeenstruckwiththesurpassinggrace，strength，andvigouroftheGame—cock，withitsboldandconfidentair，itslong，yetfirmneck，compactbody，powerfulandcloselypressedwings，muscularthighs，strongbeakmassiveatthebase，denseandsharpspurssetlowonthelegsfordeliveringthefatalblow，anditscompact，glossy，andmail—likeplumageservingasadefence。NowtheEnglishgame—cockhasnotonlybeenimprovedduringmanyyearsbyman’scarefulselection，butinaddition，asMr。
Tegetmeierhasremarked（21/3。’ThePoultryBook’1866page123。Mr。
Tegetmeier，’TheHomingorCarrierPigeon’1871pages45—58。），byakindofnaturalselection，forthestrongest，mostactiveandcourageousbirdshavestrickendowntheirantagonistsinthecockpit，generationaftergeneration，andhavesubsequentlyservedastheprogenitorsoftheirrace。Thesamekindofdoubleselectionhascomeintoplaywiththecarrierpigeon，forduringtheirtrainingtheinferiorbirdsfailtoreturnhomeandarelost，sothatevenwithoutselectionbymanonlythesuperiorbirdspropagatetheirrace。
InGreatBritain，informertimes，almosteverydistricthaditsownbreedofcattleandsheep;\"theywereindigenoustothesoil，climate，andpasturageofthelocalityonwhichtheygrazed:theyseemedtohavebeenformedforitandbyit。\"（21/4。’YouattonSheep’page312。）Butinthiscasewearequiteunabletodisentangletheeffectsofthedirectactionoftheconditionsoflife，——ofuseorhabit——ofnaturalselection——andofthatkindofselectionwhichwehaveseenisoccasionallyandunconsciouslyfollowedbymanevenduringtherudestperiodsofhistory。
Letusnowlooktotheactionofnaturalselectiononspecialcharacters。
Althoughnatureisdifficulttoresist，yetmanoftenstrivesagainstherpower，andsometimeswithsuccess。Fromthefactstobegiven，itwillalsobeseenthatnaturalselectionwouldpowerfullyaffectmanyofourdomesticproductionsifleftunprotected。Thisisapointofmuchinterest，forwethuslearnthatdifferencesapparentlyofveryslightimportancewouldcertainlydeterminethesurvivalofaformwhenforcedtostruggleforitsownexistence。Itmayhaveoccurredtosomenaturalists，asitformerlydidtome，that，thoughselectionactingundernaturalconditionswoulddeterminethestructureofallimportantorgans，yetthatitcouldnotaffectcharacterswhichareesteemedbyusoflittleimportance;butthisisanerrortowhichweareeminentlyliable，fromourignoranceofwhatcharactersareofrealvaluetoeachlivingcreature。
Whenmanattemptstomakeabreedwithsomeseriousdefectinstructure，orinthemutualrelationoftheseveralparts，hewillpartlyorcompletelyfail，orencountermuchdifficulty;heisinfactresistedbyaformofnaturalselection。WehaveseenthatanattemptwasoncemadeinYorkshiretobreedcattlewithenormousbuttocks，butthecowsperishedsoofteninbringingforththeircalves，thattheattempthadtobegivenup。Inrearingshort—
facedtumblers，Mr。Eatonsays（21/5。’TreatiseontheAlmondTumbler’1851
page33。），\"Iamconvincedthatbetterheadandbeakbirdshaveperishedintheshellthaneverwerehatched;thereasonbeingthattheamazinglyshort—
facedbirdcannotreachandbreaktheshellwithitsbeak，andsoperishes。\"
Hereisamorecuriouscase，inwhichnaturalselectioncomesintoplayonlyatlongintervalsoftime:duringordinaryseasonstheNiatacattlecangrazeaswellasothers，butoccasionally，asfrom1827to1830theplainsofLaPlatasufferfromlong—continueddroughtsandthepastureisburntup;atsuchtimescommoncattleandhorsesperishbythethousand，butmanysurvivebybrowsingontwigs，reeds，etc。;thistheNiatacattlecannotsowelleffectfromtheirupturnedjawsandtheshapeoftheirlips;consequently，ifnotattendedto，theyperishbeforetheothercattle。InColumbia，accordingtoRoulin，thereisabreedofnearlyhairlesscattle，calledPelones;thesesucceedintheirnativehotdistrict，butarefoundtootenderfortheCordillera;inthiscase，however，naturalselectiondeterminesonlytherangeofthevariety。Itisobviousthatahostofartificialracescouldneversurviveinastateofnature;——suchasItaliangreyhounds，——hairlessandalmosttoothlessTurkishdogs，——fantailpigeons，whichcannotflywellagainstastrongwind，——barbsandPolishfowls，withtheirvisionimpededbytheireyewattlesandgreattopknots，——hornlessbullsandrams，whichconsequentlycannotcopewithothermales，andthushaveapoorchanceofleavingoffspring，——seedlessplants，andmanyothersuchcases。
Colourisgenerallyesteemedbythesystematicnaturalistasunimportant:letus，therefore，seehowfaritindirectlyaffectsourdomesticproductions，andhowfaritwouldaffectthemiftheywereleftexposedtothefullforceofnaturalselection。InafuturechapterIshallhavetoshowthatconstitutionalpeculiaritiesofthestrangestkind，entailingliabilitytotheactionofcertainpoisons，arecorrelatedwiththecolouroftheskin。Iwillheregiveasinglecase，onthehighauthorityofProfessorWyman;heinformsmethat，beingsurprisedatallthepigsinapartofVirginiabeingblack，hemadeinquiries，andascertainedthattheseanimalsfeedontherootsoftheLachnanthestinctoria，whichcolourstheirbonespink，and，exceptinginthecaseoftheblackvarieties，causesthehoofstodropoff。Hence，asoneofthesquattersremarked，\"weselecttheblackmembersofthelitterforraising，astheyalonehaveagoodchanceofliving。\"Sothatherewehaveartificialandnaturalselectionworkinghandinhand。ImayaddthatintheTarentinotheinhabitantskeepblacksheepalone，becausetheHypericumcrispumaboundsthere;andthisplantdoesnotinjureblacksheep，butkillsthewhiteonesinaboutafortnight’stime。（21/6。Dr。Heusinger’WochenschriftfurdieHei1kunde’Berlin1846s。279。）
Complexion，andliabilitytocertaindiseases，arebelievedtoruntogetherinmanandtheloweranimals。Thuswhiteterrierssuffermorethanthoseofanyothercolourfromthefataldistemper。（21/7。Youattonthe’Dog’page232。）
InNorthAmericaplum—treesareliabletoadiseasewhichDowning（21/8。’TheFruit—treesofAmerica’1845page270:forpeachespage466。）believesisnotcausedbyinsects;thekindsbearingpurplefruitaremostaffected，\"andwehaveneverknownthegreenoryellowfruitedvarietiesinfecteduntiltheothersortshadfirstbecomefilledwiththeknots。\"Ontheotherhand，peachesinNorthAmericasuffermuchfromadiseasecalledthe\"yellows，\"
whichseemstobepeculiartothatcontinent，andmorethannine—tenthsofthevictims，\"whenthediseasefirstappeared，weretheyellow—fleshedpeaches。
Thewhite—fleshedkindsaremuchmorerarelyattacked;insomepartsofthecountrynever。\"InMauritius，thewhitesugar—caneshaveoflateyearsbeensoseverelyattackedbyadisease，thatmanyplantershavebeencompelledtogiveupgrowingthisvariety（althoughfreshplantswereimportedfromChinafortrial），andcultivateonlyredcanes。（21/9。’Proc。RoyalSoc。ofArtsandSciencesofMauritius’1852page135。）Now，iftheseplantshadbeenforcedtostrugglewithothercompetingplantsandenemies，therecannotbeadoubtthatthecolourofthefleshorskinofthefruit，unimportantasthesecharactersareconsidered，wouldhaverigorouslydeterminedtheirexistence。
Liabilitytotheattacksofparasitesisalsoconnectedwithcolour。Whitechickensarecertainlymoresubjectthandark—colouredchickenstothe\"gapes，\"whichiscausedbyaparasiticworminthetrachea。（21/10。
’Gardener’sChronicle’1856page379。）Ontheotherhand，experiencehasshownthatinFrancethecaterpillarswhichproducewhitecocoonsresistthedeadlyfungusbetterthanthoseproducingyellowcocoons。（21/11。Quatrefages’MaladiesActuellesduVeraSoie’1859pages12，214。）Analogousfactshavebeenobservedwithplants:anewandbeautifulwhiteonion，importedfromFrance，thoughplantedclosetootherkinds，wasaloneattackedbyaparasiticfungus。（21/12。’Gardener’sChronicle’1851page595。）Whiteverbenasareespeciallyliabletomildew。（21/13。’JournalofHorticulture’1862page476。）
NearMalaga，duringanearlyperiodofthevine—disease，thegreensortssufferedmost;\"andredandblackgrapes，evenwheninterwovenwiththesickplants，sufferednotatall。\"InFrancewholegroupsofvarietieswerecomparativelyfree，andothers，suchastheChasselas，didnotaffordasinglefortunateexception;butIdonotknowwhetheranycorrelationbetweencolourandliabilitytodiseasewashereobserved。（21/14。’Gardener’sChronicle’
1852pages435，691。）Inaformerchapteritwasshownhowcuriouslyliableonevarietyofthestrawberryistomildew。
Itiscertainthatinsectsregulateinmanycasestherangeandeventheexistenceofthehigheranimals，whilstlivingundertheirnaturalconditions。
Underdomesticationlight—colouredanimalssuffermost:inThuringia（21/15。
Bechstein’Naturgesch。Deutschlands’1801b。1s。310。）theinhabitantsdonotlikegrey，white，orpalecattle，becausetheyaremuchmoretroubledbyvariouskindsoffliesthanthebrown，red，orblackcattle。AnAlbinonegro，ithasbeenremarked（21/16。Prichard’Phys。Hist。ofMankind’1851volume1
page224。），waspeculiarlysensitivetothebitesofinsects。IntheWestIndies（21/17。G。Lewis’JournalofResidenceinWestIndies’’HomeandCol。
Library’page100。）itissaidthat\"theonlyhornedcattlefitforworkarethosewhichhaveagooddealofblackinthem。Thewhiteareterriblytormentedbytheinsects;andtheyareweakandsluggishinproportiontothewhite。\"
InDevonshirethereisaprejudiceagainstwhitepigs，becauseitisbelievedthatthesunblistersthemwhenturnedout（21/18。Sidney’seditionofYouattonthe’Pig’page24。Ihavegivenanalogousfactsinthecaseofmankindinmy’DescentofMan’2ndeditionpage195。）;andIknewamanwhowouldnotkeepwhitepigsinKent，forthesamereason。Thescorchingofflowersbythesunseemslikewisetodependmuchoncolour;thus，darkpelargoniumssuffermost;andfromvariousaccountsitisclearthatthecloth—of—goldvarietywillnotwithstandadegreeofexposuretosunshinewhichothervarietiesenjoy。Anotheramateurassertsthatnotonlyalldark—colouredverbenas，butlikewisescarlets，sufferfromthesun:\"thepalerkindsstandbetter，andpaleblueisperhapsthebestofall。\"Soagainwiththeheartsease（Violatricolor）;hotweathersuitstheblotchedsorts，whilstitdestroysthebeautifulmarkingsofsomeotherkinds。（21/19。’JournalofHorticulture’1862
pages476，498;1865page460。Withrespecttotheheartsease’Gardener’sChronicle’1863page628。）DuringoneextremelycoldseasoninHollandallred—floweredhyacinthswereobservedtobeveryinferiorinquality。Itisbelievedbymanyagriculturiststhatredwheatishardierinnorthernclimatesthanwhitewheat。（21/20。’DesJacinthes，deleurCulture’1768page53:onwheat’Gardener’sChronicle’1846page653。）
Withanimals，whitevarietiesfrombeingconspicuousarethemostliabletobeattackedbybeastsandbirdsofprey。InpartsofFranceandGermanywherehawksabound，personsareadvisednottokeepwhitepigeons;for，asParmentiersays，\"itiscertainthatinaflockthewhitealwaysfirstfallvictimstothekite。\"InBelgium，wheresomanysocietieshavebeenestablishedfortheflightofcarrier—pigeons，whiteistheonecolourwhichforthesamereasonisdisliked。（20/21。W。B。Tegetmeier’TheField’February25，1865。WithrespecttoblackfowlsseeaquotationinThompson’Nat。Hist。
ofIreland’1849volume1page22。）Prof。G。Jaeger（21/22。’InSachenDarwin’scontraWigand’1874page70。）whilstfishingfoundfourpigeonswhichhadbeenkilledbyhawks，andallwerewhite;onanotheroccasionheexaminedtheeyrieofahawk，andthefeathersofthepigeonswhichhadbeencaughtwereallofawhiteoryellowcolour。Ontheotherhand，itissaidthatthesea—eagle（Falcoossifragus，Linn。）onthewestcoastofIrelandpicksouttheblackfowls，sothat\"thevillagersavoidasmuchaspossiblerearingbirdsofthatcolour。\"M。Daudin（20/23。’Bull。delaSoc。d’Acclimat。’tome71860
page359。），speakingofwhiterabbitskeptinwarrensinRussia，remarksthattheircolourisagreatdisadvantage，astheyarethusmoreexposedtoattack，andcanbeseenduringbrightnightsfromadistance。AgentlemaninKent，whofailedtostockhiswoodswithanearlywhiteandhardykindofrabbit，accountedinthesamemannerfortheirearlydisappearance。Anyonewhowillwatchawhitecatprowlingafterherpreywillsoonperceiveunderwhatadisadvantageshelies。
ThewhiteTartariancherry，\"owingeithertoitscolourbeingsomuchlikethatoftheleaves，ortothefruitalwaysappearingfromadistanceunripe，\"
isnotsoreadilyattackedbybirdsasothersorts。Theyellow—fruitedraspberry，whichgenerallycomesnearlytruebyseed，\"isverylittlemolestedbybirds，whoevidentlyarenotfondofit;sothatnetsmaybedispensedwithinplaceswherenothingelsewillprotecttheredfruit。\"（21/24。’Transact。
Hort。Soc。’volume12ndseries1835page275。Forraspberriessee’Gardener’sChronicle’1855page154and1863page245。）Thisimmunity，thoughabenefittothegardener，wouldbeadisadvantageinastateofnaturebothtothecherryandraspberry，asdisseminationdependsonbirds。Inoticedduringseveralwintersthatsometreesoftheyellow—berriedholly，whichwereraisedfromseedfromatreefoundwildbymyfatherremainedcoveredwithfruit，whilstnotascarletberrycouldbeseenontheadjoiningtreesofthecommonkind。Afriendinformsmethatamountain—ash（Pyrusaucuparia）growinginhisgardenbearsberrieswhich，thoughnotdifferentlycoloured，arealwaysdevouredbybirdsbeforethoseontheothertrees。Thisvarietyofthemountain—ashwouldthusbemorefreelydisseminated，andtheyellow—berriedvarietyofthehollylessfreely，thanthecommonvarietiesofthesetwotrees。
Independentlyofcolour，triflingdifferencesaresometimesfoundtobeofimportancetoplantsundercultivation，andwouldbeofparamountimportanceiftheyhadtofighttheirownbattleandtostrugglewithmanycompetitors。
Thethin—shelledpeas，calledpoissansparchemin，areattackedbybirds（21/25。’Gardener’sChronicle’1843page806。）muchmorecommonlythanordinarypeas。Ontheotherhand，thepurple—poddedpea，whichhasahardshell，escapedtheattacksoftomtits（Parusmajor）inmygardenfarbetterthananyotherkind。Thethin—shelledwalnutlikewisesuffersgreatlyfromthetomtit。（21/26。Ibid1850page732。）Thesesamebirdshavebeenobservedtopassoverandthusfavourthefilbert，destroyingonlytheotherkindsofnutswhichgrewinthesameorchard。（21/27。Ibid1860page956。）
Certainvarietiesofthepearhavesoftbark，andthesesufferseverelyfromwood—boringbeetles;whilstothervarietiesareknowntoresisttheirattacksmuchbetter。（21/28。J。DeJonghein’Gardener’sChronicle’1860page120。）InNorthAmericathesmoothness，orabsenceofdownonthefruit，makesagreatdifferenceintheattacksoftheweevil，\"whichistheuncompromisingfoeofallsmoothstone—fruits;\"andthecultivator\"hasthefrequentmortificationofseeingnearlyall，orindeedoftenthewholecrop，fallfromthetreeswhenhalfortwo—thirdsgrown。\"Hencethenectarinesuffersmorethanthepeach。A
particularvarietyoftheMorellocherry，raisedinNorthAmerica，is，withoutanyassignablecause，moreliabletobeinjuredbythissameinsectthanothercherry—trees。（21/29。Downing’Fruit—treesofNorthAmerica’pages266，501:
inregardtothecherrypage198。）Fromsomeunknowncause，certainvarietiesoftheappleenjoy，aswehaveseen，thegreatadvantageinvariouspartsoftheworldofnotbeinginfestedbythecoccus。Ontheotherhand，aparticularcasehasbeenrecordedinwhichaphidesconfinedthemselvestotheWinterNelispearandtouchednootherkindinanextensiveorchard。（21/30。
’Gardener’sChronicle’1849page755。）Theexistenceofminuteglandsontheleavesofpeaches，nectarines，andapricots，wouldnotbeesteemedbybotanistsasacharacteroftheleastimportancefortheyarepresentorabsentinclosely—relatedsub—varieties，descendedfromthesameparent—tree;
yetthereisgoodevidence（21/31。’JournalofHorticulture’September26，1865page254;seeotherreferencesgiveninchapter10。）thattheabsenceofglandsleadstomildew，whichishighlyinjurioustothesetrees。
Adifferenceeitherinflavourorintheamountofnutrimentincertainvarietiescausesthemtobemoreeagerlyattackedbyvariousenemiesthanothervarietiesofthesamespecies。Bullfinches（Pyrrhulavulgaris）injureourfruit—treesbydevouringtheflower—buds，andapairofthesebirdshavebeenseen\"todenudealargeplum—treeinacoupleofdaysofalmosteverybud;\"butcertainvarieties（21/32。Mr。Selbyin’Mag。ofZoologyandBotany’
Edinburghvolume21838page393。）oftheappleandthorn（Crataegusoxyacantha）aremoreespeciallyliabletobeattacked。AstrikinginstanceofthiswasobservedinMr。Rivers’sgarden，inwhichtworowsofaparticularvarietyofplum（21/33。TheReineClaudedeBavay’JournalofHorticulture’
December27，1864page511。）hadtobecarefullyprotected，astheywereusuallystrippedofalltheirbudsduringthewinter，whilstothersortsgrowingnearthemescaped。Theroot（orenlargedstem）ofLaing’sSwedishturnipispreferredbyhares，andthereforesuffersmorethanothervarieties。
HaresandrabbitseatdowncommonryebeforeSt。John’s—day—rye，whenbothgrowtogether。（21/34。Mr。Puseyin’JournalofR。Agricult。Soc。’volume6
page179。ForSwedishturnipssee’Gardener’sChronicle’1847page91。）InthesouthofFrance，whenanorchardofalmond—treesisformed，thenutsofthebittervarietyaresown，\"inorderthattheymaynotbedevouredbyfield—
mice\"（21/35。Godron’Del’Espece’tome2page98。）;soweseetheuseofthebitterprincipleinalmonds。
Otherslightdifferences，whichwouldbethoughtquiteunimportant，arenodoubtsometimesofgreatservicebothtoplantsandanimals。TheWhitesmith’sgooseberry，asformerlystated，producesitsleaveslaterthanothervarieties，and，astheflowersarethusleftunprotected，thefruitoftenfails。Inonevarietyofthecherry，accordingtoMr。Rivers（21/36。
’Gardener’sChronicle’1866page732。），thepetalsaremuchcurledbackwards，andinconsequenceofthisthestigmaswereobservedtobekilledbyaseverefrost;whilstatthesametime，inanothervarietywithincurvedpetals，thestigmaswerenotintheleastinjured。ThestrawoftheFentonwheatisremarkablyunequalinheight;andacompetentobserverbelievesthatthisvarietyishighlyproductive，partlybecausetheearsfrombeingdistributedatvariousheightsabovethegroundarelesscrowdedtogether。Thesameobservermaintainsthatintheuprightvarietiesthedivergentawnsareserviceablebybreakingtheshockswhentheearsaredashedtogetherbythewind。（21/37。’Gardener’sChronicle’1862pages820，821。）Ifseveralvarietiesofaplantaregrowntogether，andtheseedisindiscriminatelyharvested，itisclearthatthehardierandmoreproductivekindswill，byasortofnaturalselection，graduallyprevailovertheothers;thistakesplace，asColonelLeCouteurbelieves（21/38。’OntheVarietiesofWheat’page59。），inourwheat—fields，for，asformerlyshown，novarietyisquiteuniformincharacter。Thesamething，asIamassuredbynurserymen，wouldtakeplaceinourflower—gardens，iftheseedofthedifferentvarietieswerenotseparatelysaved。Whentheeggsofthewildandtameduckarehatchedtogether，theyoungwildducksalmostinvariablyperish，frombeingofsmallersizeandnotgettingtheirfairshareoffood。（21/39。Mr。Hewittandothers，in’JournalofHort。’1862page773。）
Factsinsufficientnumberhavenowbeengivenshowingthatnaturalselectionoftenchecks，butoccasionallyfavours，man’spowerofselection。Thesefactsteachus，inaddition，avaluablelesson，namely，thatweoughttobeextremelycautiousinjudgingwhatcharactersareofimportanceinastateofnaturetoanimalsandplants，whichhavetostruggleforexistencefromthehouroftheirbirthtothatoftheirdeath，——theirexistencedependingonconditions，aboutwhichweareprofoundlyignorant。
CIRCUMSTANCESFAVOURABLETOSELECTIONBYMAN。
Thepossibilityofselectionrestsonvariability，andthis，asweshallseeinthefollowingchapters，mainlydependsonchangedconditionsoflife，butisgovernedbyinfinitelycomplexandunknownlaws。Domestication，evenwhenlongcontinued，occasionallycausesbutasmallamountofvariability，asinthecaseofthegooseandturkey。Theslightdifferences，however，whichcharacteriseeachindividualanimalandplantwouldinmost，probablyinall，casessufficefortheproductionofdistinctracesthroughcarefulandprolongedselection。Weseewhatselection，thoughactingonmereindividualdifferences，caneffectwhenfamiliesofcattle，sheep，pigeons，etc。，ofthesamerace，havebeenseparatelybredduringanumberofyearsbydifferentmenwithoutanywishontheirparttomodifythebreed。Weseethesamefactinthedifferencebetweenhoundsbredforhuntingindifferentdistricts（21/40。
’Encyclop。ofRuralSports’page405。），andinmanyothersuchcases。
Inorderthatselectionshouldproduceanyresult，itismanifestthatthecrossingofdistinctracesmustbeprevented;hencefacilityinpairing，aswiththepigeon，ishighlyfavourableforthework;anddifficultyinpairing，aswithcats，preventstheformationofdistinctbreeds。OnnearlythesameprinciplethecattleofthesmallislandofJerseyhavebeenimprovedintheirmilkingqualities\"witharapiditythatcouldnothavebeenobtainedinawidelyextendedcountrylikeFrance。\"（21/41。Col。LeCouteur’JournalRoy。