Withextremelyfewexceptions，allanimalsandplantswhichhavebeenlongdomesticatedhavevariedgreatly。Itmattersnotunderwhatclimate，orforwhatpurposetheyarekept，whetherasfoodformanorbeast，fordraughtorhunting，forclothingormerepleasure，——underallthesecircumstancesraceshavebeenproducedwhichdiffermorefromoneanotherthandotheformswhichinastateofnaturearerankedasdifferentspecies。Whycertainanimalsandplantshavevariedmoreunderdomesticationthanotherswedonotknow，anymorethanwhysomearerenderedmoresterilethanothersunderchangedconditionsoflife。Butwehavetojudgeoftheamountofvariationwhichourdomesticproductionshaveundergone，chieflybythenumberandamountofdifferencebetweentheraceswhichhavebeenformed，andwecanoftenclearlyseewhymanyanddistinctraceshavenotbeenformed，namely，becauseslightsuccessivevariationshavenotbeensteadilyaccumulated;andsuchvariationswillneverbeaccumulatedifananimalorplantbenotcloselyobserved，muchvalued，andkeptinlargenumbers。
Thefluctuating，and，asfaraswecanjudge，never—endingvariabilityofourdomesticatedproductions，——theplasticityofalmosttheirwholeorganisation，—
—isoneofthemostimportantlessonswhichwelearnfromthenumerousdetailsgivenintheearlierchaptersofthiswork。Yetdomesticatedanimalsandplantscanhardlyhavebeenexposedtogreaterchangesintheirconditionsoflifethanhavemanynaturalspeciesduringtheincessantgeological，geographical，andclimatalchangestowhichtheworldhasbeensubject;butdomesticatedproductionswilloftenhavebeenexposedtomoresuddenchangesandtolesscontinuouslyuniformconditions。Asmanhasdomesticatedsomanyanimalsandplantsbelongingtowidelydifferentclasses，andashecertainlydidnotchoosewithpropheticinstinctthosespecieswhichwouldvarymost，wemayinferthatallnaturalspecies，ifexposedtoanalogousconditions，would，onanaverage，varytothesamedegree。Fewmenatthepresentdaywillmaintainthatanimalsandplantswerecreatedwithatendencytovary，whichlongremaineddormant，inorderthatfanciersinafteragesmightrear，forinstance，curiousbreedsofthefowl，pigeon，orcanary—bird。
Fromseveralcausesitisdifficulttojudgeoftheamountofmodificationwhichourdomesticproductionshaveundergone。Insomecasestheprimitiveparent—stockhasbecomeextinct;oritcannotberecognisedwithcertainty，owingtoitssupposeddescendantshavingbeensomuchmodified。Inothercasestwoormoreclosely—alliedforms，afterbeingdomesticated，havecrossed;andthenitisdifficulttoestimatehowmuchofthecharacterofthepresentdescendantsoughttobeattributedtovariation，andhowmuchtotheinfluenceoftheseveralparent—stocks。Butthedegreetowhichourdomesticatedbreedshavebeenmodifiedbythecrossingofdistinctspecieshasprobablybeenmuchexaggeratedbysomeauthors。Afewindividualsofoneformwouldseldompermanentlyaffectanotherformexistingingreaternumbers;for，withoutcarefulselection，thestainoftheforeignbloodwouldsoonbeobliterated，andduringearlyandbarbaroustimes，whenouranimalswerefirstdomesticated，suchcarewouldseldomhavebeentaken。
Thereisgoodreasontobelieveinthecaseofthedog，ox，pig，andofsomeotheranimals，thatseveralofourracesaredescendedfromdistinctwildprototypes;neverthelessthebeliefinthemultipleoriginofourdomesticatedanimalshasbeenextendedbysomefewnaturalistsandbymanybreederstoanunauthorisedextent。Breedersrefusetolookatthewholesubjectunderasinglepointofview;Ihavehearditsaidbyaman，whomaintainedthatourfowlsweredescendedfromatleasthalf—a—dozenaboriginalspecies，thattheevidenceofthecommonoriginofpigeons，ducksandrabbits，wasofnoavailwithrespecttofowls。Breedersoverlooktheimprobabilityofmanyspecieshavingbeendomesticatedatanearlyandbarbarousperiod。Theydonotconsidertheimprobabilityofspecieshavingexistedinastateofnaturewhich，iftheyresembledourpresentdomesticbreeds，wouldhavebeenhighlyabnormalincomparisonwithalltheircongeners。Theymaintainthatcertainspecies，whichformerlyexisted，havebecomeextinct，orarenowunknown，althoughformerlyknown。Theassumptionofsomuchrecentextinctionisnodifficultyintheireyes;fortheydonotjudgeofitsprobabilitybythefacilityordifficultyoftheextinctionofotherclosely—alliedwildforms。
Lastly，theyoftenignorethewholesubjectofgeographicaldistributionascompletelyasifitweretheresultofchance。
Althoughfromthereasonsjustassigneditisoftendifficulttojudgeaccuratelyoftheamountofchangewhichourdomesticatedproductionshaveundergone，yetthiscanbeascertainedinthecasesinwhichallthebreedsareknowntobedescendedfromasinglespecies，——aswiththepigeon，duck，rabbit，andalmostcertainlywiththefowl;andbytheaidofanalogythiscanbejudgedoftoacertainextentwithdomesticatedanimalsdescendedfromseveralwildstocks。Itisimpossibletoreadthedetailsgivenintheearlierchaptersandinmanypublishedworks，ortovisitourvariousexhibitions，withoutbeingdeeplyimpressedwiththeextremevariabilityofourdomesticatedanimalsandcultivatedplants。Nopartoftheorganisationescapesthetendencytovary。Thevariationsgenerallyaffectpartsofsmallvitalorphysiologicalimportance，butsoitiswiththedifferenceswhichexistbetweenclosely—alliedspecies。Intheseunimportantcharactersthereisoftenagreaterdifferencebetweenthebreedsofthesamespeciesthanbetweenthenaturalspeciesofthesamegenus，asIsidoreGeoffroyhasshowntobethecasewithsize，andasisoftenthecasewiththecolour，texture，form，etc。，ofthehair，feathers，horns，andotherdermalappendages。
Ithasoftenbeenassertedthatimportantpartsnevervaryunderdomestication，butthisisacompleteerror。Lookattheskullofthepiginanyoneofthehighlyimprovedbreeds，withtheoccipitalcondylesandotherpartsgreatlymodified;orlookatthatoftheniataox。Or，again，intheseveralbreedsoftherabbit，observetheelongatedskull，withthedifferentlyshapedoccipitalforamen，atlas，andothercervicalvertebrae。Thewholeshapeofthebrain，togetherwiththeskull，hasbeenmodifiedinPolishfowls;inotherbreedsofthefowlthenumberofthevertebraeandtheformsofthecervicalvertebraehavebeenchanged。Incertainpigeonstheshapeofthelowerjaw，therelativelengthofthetongue，thesizeofthenostrilsandeyelids，thenumberandshapeoftheribs，theformandsizeoftheoesophagus，haveallvaried。Incertainquadrupedsthelengthoftheintestineshasbeenmuchincreasedordiminished。Withplantsweseewonderfuldifferencesinthestonesofvariousfruits。IntheCucurbitaceaeseveralhighlyimportantcharactershavevaried，suchasthesessilepositionofthestigmasontheovarium，thepositionofthecarpels，andtheprojectionoftheovariumoutofthereceptacle。Butitwouldbeuselesstorunthroughthemanyfactsgivenintheearlierchapters。
Itisnotorioushowgreatlythementaldisposition，tastes，habits，consensualmovements，loquacityorsilence，andtoneofvoicehavevariedandbeeninheritedinourdomesticatedanimals。Thedogoffersthemoststrikinginstanceofchangedmentalattributes，andthesedifferencescannotbeaccountedforbydescentfromdistinctwildtypes。
Newcharactersmayappearandoldonesdisappearatanystageofdevelopment，beinginheritedatacorrespondingstage。Weseethisinthedifferencebetweentheeggs，thedownonthechickensandthefirstplumageofthevariousbreedsofthefowl;andstillmoreplainlyinthedifferencesbetweenthecaterpillarsandcocoonsofthevariousbreedsofthesilk—moth。Thesefacts，simpleastheyappear，throwlightonthedifferencesbetweenthelarvalandadultstatesofalliednaturalspecies，andonthewholegreatsubjectofembryology。Newcharactersfirstappearinglateinlifeareapttobecomeattachedexclusivelytothatsexinwhichtheyfirstarose，ortheymaybedevelopedinamuchhigherdegreeinthisthanintheothersex;oragain，afterhavingbecomeattachedtoonesex，theymaybetransferredtotheoppositesex。Thesefacts，andmoreespeciallythecircumstancethatnewcharactersseemtobeparticularlyliable，fromsomeunknowncause，tobecomeattachedtothemalesex，haveanimportantbearingontheacquirementofsecondarysexualcharactersbyanimalsinastateofnature。
Ithassometimesbeensaidthatourdomesticracesdonotdifferinconstitutionalpeculiarities，butthiscannotbemaintained。Inourimprovedcattle，pigs，etc。，theperiodofmaturity，includingthatoftheseconddentition，hasbeenmuchhastened。Theperiodofgestationvariesmuch，andhasbeenmodifiedinafixedmannerinoneortwocases。Insomebreedsofpoultryandpigeonstheperiodatwhichthedownandthefirstplumageareacquired，differs。Thenumberofmoultsthroughwhichthelarvaeofsilk—mothspass，varies。Thetendencytofatten，toyieldmuchmilk，toproducemanyyoungoreggsatabirthorduringlife，differsindifferentbreeds。Wefinddifferentdegreesofadaptationtoclimate，anddifferenttendenciestocertaindiseases，totheattacksofparasites，andtotheactionofcertainvegetablepoisons。Withplants，adaptationtocertainsoils，thepowerofresistingfrost，theperiodoffloweringandfruiting，thedurationoflife，theperiodofsheddingtheleavesorofretainingthemthroughoutthewinter，theproportionandnatureofcertainchemicalcompoundsinthetissuesorseeds，allvary。
Thereis，however，oneimportantconstitutionaldifferencebetweendomesticracesandspecies;Irefertothesterilitywhichalmostinvariablyfollows，inagreaterorlessdegree，whenspeciesarecrossed，andtotheperfectfertilityofthemostdistinctdomesticraces，withtheexceptionofaveryfewplants，whensimilarlycrossed。Itiscertainlyamostremarkablefactthatmanyclosely—alliedspecies，whichinappearancedifferextremelylittle，shouldyieldwhencrossedonlyafewmoreorlesssterileoffspring，ornoneatall;whilstdomesticraceswhichdifferconspicuouslyfromeachotherare，whenunited，remarkablyfertile，andyieldperfectlyfertileoffspring。Butthisfactisnotinrealitysoinexplicableasitatfirstappears。Inthefirstplace，itwasclearlyshowninthenineteenthchapterthatthesterilityofcrossedspeciesdoesnotdependchieflyondifferencesintheirexternalstructureorgeneralconstitution，butondifferencesinthereproductivesystem，analogoustothosewhichcausethelessenedfertilityoftheillegitimateunionsofdimorphicandtrimorphicplants。Inthesecondplace，thePallasiandoctrine，thatspeciesafterhavingbeenlongdomesticatedlosetheirnaturaltendencytosterilitywhencrossed，hasbeenshowntobehighlyprobableoralmostcertain。Wecannotavoidthisconclusionwhenwereflectontheparentageandpresentfertilityoftheseveralbreedsofthedog，oftheIndianorhumpedandEuropeancattle，andofthetwochiefkindsofpigs。
Henceitwouldbeunreasonabletoexpectthatracesformedunderdomesticationshouldacquiresterilitywhencrossed，whilstatthesametimeweadmitthatdomesticationeliminatesthenormalsterilityofcrossedspecies。Whywithclosely—alliedspeciestheirreproductivesystemsshouldalmostinvariablyhavebeenmodifiedinsopeculiaramannerastobemutuallyincapableofactingoneachother——thoughinunequaldegreesinthetwosexes，asshownbythedifferenceinfertilitybetweenreciprocalcrossesofthesamespecies——wedonotknow，butmaywithmuchprobabilityinferthecausetobeasfollows。
Mostnaturalspecieshavebeenhabituatedtonearlyuniformconditionsoflifeforanincomparablylongertimethanhavedomesticraces;andwepositivelyknowthatchangedconditionsexertanespecialandpowerfulinfluenceonthereproductivesystem。Hencethisdifferencemaywellaccountforthedifferenceinthepowerofreproductionbetweendomesticraceswhencrossedandspecieswhencrossed。Itisprobablyinchiefpartowingtothesamecausethatdomesticracescanbesuddenlytransportedfromoneclimatetoanother，orplacedunderwidelydifferentconditions，andyetretaininmostcasestheirfertilityunimpaired;whilstamultitudeofspeciessubjectedtolesserchangesarerenderedincapableofbreeding。
Theoffspringofcrosseddomesticracesandofcrossedspeciesresembleeachotherinmostrespects，withtheoneimportantexceptionoffertility;theyoftenpartakeinthesameunequaldegreeofthecharactersoftheirparents，oneofwhichisoftenprepotentovertheother;andtheyareliabletoreversionofthesamekind。Bysuccessivecrossesonespeciesmaybemadetoabsorbcompletelyanother，andsoitnotoriouslyiswithraces。Thelatterresemblespeciesinmanyotherways。Theysometimesinherittheirnewly—
acquiredcharactersalmostorevenquiteasfirmlyasspecies。Theconditionsleadingtovariabilityandthelawsgoverningitsnatureappeartobethesameinboth。Varietiescanbeclassedingroupsundergroups，likespeciesundergenera，andtheseunderfamiliesandorders;andtheclassificationmaybeeitherartificial，——thatis，foundedonanyarbitrarycharacter，——ornatural。
Withvarietiesanaturalclassificationiscertainlyfounded，andwithspeciesisapparentlyfounded，oncommunityofdescent，togetherwiththeamountofmodificationwhichtheformshaveundergone。Thecharactersbywhichdomesticvarietiesdifferfromoneanotheraremorevariablethanthosedistinguishingspecies，thoughhardlymoresothanwithcertainpolymorphicspecies;butthisgreaterdegreeofvariabilityisnotsurprising，asvarietieshavegenerallybeenexposedwithinrecenttimestofluctuatingconditionsoflife，andaremuchmoreliabletohavebeencrossed;theyarealsoinmanycasesstillundergoing，orhaverecentlyundergone，modificationbyman’smethodicalorunconsciousselection。
Domesticvarietiesasageneralrulecertainlydifferfromoneanotherinlessimportantpartsthandospecies;andwhenimportantdifferencesoccur，theyareseldomfirmlyfixed;butthisfactisintelligible，ifweconsiderman’smethodofselection。Inthelivinganimalorplanthecannotobserveinternalmodificationsinthemoreimportantorgans;nordoesheregardthemaslongastheyarecompatiblewithhealthandlife。Whatdoesthebreedercareaboutanyslightchangeinthemolarteethofhispigs，orforanadditionalmolartoothinthedog;orforanychangeintheintestinalcanalorotherinternalorgan？
Thebreedercaresforthefleshofhiscattlebeingwellmarbledwithfat，andforanaccumulationoffatwithintheabdomenofhissheep，andthishehaseffected。Whatwouldthefloriculturistcareforanychangeinthestructureoftheovariumoroftheovules？Asimportantinternalorgansarecertainlyliabletonumerousslightvariations，andasthesewouldprobablybetransmitted，formanystrangemonstrositiesareinherited，mancouldundoubtedlyeffectacertainamountofchangeintheseorgans。Whenhehasproducedanymodificationinanimportantpart，hehasgenerallydonesounintentionally，incorrelationwithsomeotherconspicuouspart。Forinstance，hehasgivenridgesandprotuberancestotheskullsoffowls，byattendingtotheformofthecomb，ortotheplumeoffeathersonthehead。Byattendingtotheexternalformofthepouter—pigeon，hehasenormouslyincreasedthesizeoftheoesophagus，andhasaddedtothenumberoftheribs，andgiventhemgreaterbreadth。Withthecarrier—pigeon，byincreasingthroughsteadyselectionthewattlesontheuppermandible，hehasgreatlymodifiedtheformofthelowermandible;andsoinmanyothercases。Naturalspecies，ontheotherhand，havebeenmodifiedexclusivelyfortheirowngood，tofitthemforinfinitelydiversifiedconditionsoflife，toavoidenemiesofallkinds，andtostruggleagainstahostofcompetitors。Hence，undersuchcomplexconditions，itwouldoftenhappenthatmodificationsofthemostvariedkinds，inimportantaswellasinunimportantparts，wouldbeadvantageousorevennecessary;andtheywouldslowlybutsurelybeacquiredthroughthesurvivalofthefittest。Stillmoreimportantisthefactthatvariousindirectmodificationswouldlikewisearisethroughthelawofcorrelatedvariation。
Domesticbreedsoftenhaveanabnormalorsemi—monstrouscharacter，asamongstdogs，theItaliangreyhound，bulldog，Blenheimspaniel，andbloodhound，——somebreedsofcattleandpigs，——severalbreedsofthefowl，——andthechiefbreedsofthepigeon。Insuchabnormalbreeds，partswhichdifferbutslightlyornotatallinthealliednaturalspecies，havebeengreatlymodified。Thismaybeaccountedforbyman’softenselecting，especiallyatfirst，conspicuousandsemi—monstrousdeviationsofstructure。Weshould，however，becautiousindecidingwhatdeviationsoughttobecalledmonstrous:therecanhardlybeadoubtthat，ifthebrushofhorse—likehaironthebreastoftheturkey—cockhadfirstappearedinthedomesticatedbird，itwouldhavebeenconsideredasamonstrosity;thegreatplumeoffeathersontheheadofthePolishcockhasbeenthusdesignated，thoughplumesarecommonontheheadsofmanykindsofbirds;wemightcallthewattleorcorrugatedskinroundthebaseofthebeakoftheEnglishcarrier—pigeonamonstrosity，butwedonotthusspeakoftheglobularfleshyexcrescenceatthebaseofthebeakoftheCarpophagaoceanica。
Someauthorshavedrawnawidedistinctionbetweenartificialandnaturalbreeds;althoughinextremecasesthedistinctionisplain，inmanyothercasesitisarbitrary;thedifferencedependingchieflyonthekindofselectionwhichhasbeenapplied。Artificialbreedsarethosewhichhavebeenintentionallyimprovedbyman;theyfrequentlyhaveanunnaturalappearance，andareespeciallyliabletolosetheircharactersthroughreversionandcontinuedvariability。Theso—callednaturalbreeds，ontheotherhand，arethosewhicharefoundinsemi—civilisedcountries，andwhichformerlyinhabitedseparatedistrictsinnearlyalltheEuropeankingdoms。Theyhavebeenrarelyactedonbyman’sintentionalselection;morefrequentlybyunconsciousselection，andpartlybynaturalselection，foranimalskeptinsemi—civilisedcountrieshavetoprovidelargelyfortheirownwants。Suchnaturalbreedswillalsohavebeendirectlyactedonbythedifferences，thoughslight，inthesurroundingconditions。
Thereisamuchmoreimportantdistinctionbetweenourseveralbreeds，namely，insomehavingoriginatedfromastrongly—markedorsemi—monstrousdeviationofstructure，which，however，maysubsequentlyhavebeenaugmentedbyselection;whilstothershavebeenformedinsoslowandinsensibleamanner，thatifwecouldseetheirearlyprogenitorsweshouldhardlybeabletosaywhenorhowthebreedfirstarose。Fromthehistoryoftheracehorse，greyhound，gamecock，etc。，andfromtheirgeneralappearance，wemayfeelnearlyconfidentthattheywereformedbyaslowprocessofimprovement;andweknowthatthishasbeenthecasewiththecarrier—pigeon，aswellaswithsomeotherpigeons。Ontheotherhand，itiscertainthattheanconandmauchampbreedsofsheep，andalmostcertainthattheniatacattle，turnspit，andpug—dogs，jumperandfrizzledfowls，short—facedtumblerpigeons，hook—
billedducks，etc。，suddenlyappearedinnearlythesamestateaswenowseethem。Soithasbeenwithmanycultivatedplants。Thefrequencyofthesecasesislikelytoleadtothefalsebeliefthatnaturalspecieshaveoftenoriginatedinthesameabruptmanner。Butwehavenoevidenceoftheappearance，oratleastofthecontinuedprocreation，undernature，ofabruptmodificationsofstructure;andvariousgeneralreasonscouldbeassignedagainstsuchabelief。
Ontheotherhand，wehaveabundantevidenceoftheconstantoccurrenceundernatureofslightindividualdifferencesofthemostdiversifiedkinds;andwearethusledtoconcludethatspecieshavegenerallyoriginatedbythenaturalselectionofextremelyslightdifferences。Thisprocessmaybestrictlycomparedwiththeslowandgradualimprovementoftheracehorse，greyhound，andgamecock。Aseverydetailofstructureineachspecieshastobecloselyadaptedtoitshabitsoflife，itwillrarelyhappenthatonepartalonewillbemodified;but，aswasformerlyshown，theco—adaptedmodificationsneednotbeabsolutelysimultaneous。Manyvariations，however，arefromthefirstconnectedbythelawofcorrelation。Henceitfollowsthatevenclosely—alliedspeciesrarelyorneverdifferfromoneanotherbyonecharacteralone;andthesameremarkistoacertainextentapplicabletodomesticraces;forthese，iftheydiffermuch，generallydifferinmanyrespects。
Somenaturalistsboldlyinsist（28/1。Godron’Del’Espece’1859tome2page44
etc。）thatspeciesareabsolutelydistinctproductions，neverpassingbyintermediatelinksintooneanother;whilsttheymaintainthatdomesticvarietiescanalwaysbeconnectedeitherwithoneanotherorwiththeirparent—forms。Butifwecouldalwaysfindthelinksbetweentheseveralbreedsofthedog，horse，cattle，sheep，pigs，etc。，therewouldnothavebeensuchincessantdoubtswhethertheyweredescendedfromoneorseveralspecies。Thegreyhoundgenus，ifsuchatermmaybeused，cannotbecloselyconnectedwithanyotherbreed，unless，perhaps，wegobacktotheancientEgyptianmonuments。OurEnglishbulldogalsoformsaverydistinctbreed。Inallthesecasescrossedbreedsmustofcoursebeexcluded，fordistinctnaturalspeciescanthusbelikewiseconnected。BywhatlinkscantheCochinfowlbecloselyunitedwithothers？Bysearchingforbreedsstillpreservedindistantlands，andbygoingbacktohistoricalrecords，tumbler—pigeons，carriers，andbarbscanbecloselyconnectedwiththeparentrock—pigeon;butwecannotthusconnecttheturbitorthepouter。Thedegreeofdistinctnessbetweenthevariousdomesticbreedsdependsontheamountofmodificationwhichtheyhaveundergone，andmoreespeciallyontheneglectandfinalextinctionofintermediateandless—valuedforms。
Ithasoftenbeenarguedthatnolightisthrownonthechangeswhichnaturalspeciesarebelievedtoundergofromtheadmittedchangesofdomesticraces，asthelatteraresaidtobemeretemporaryproductions，alwaysreverting，assoonastheybecomeferal，totheirpristineform。ThisargumenthasbeenwellcombatedbyMr。Wallace（28/2。’JournalProc。Linn。Soc。’1858volume3page60。）andfulldetailsweregiveninthethirteenthchapter，showingthatthetendencytoreversioninferalanimalsandplantshasbeengreatlyexaggerated，thoughnodoubtitexiststoacertainextent。Itwouldbeopposedtoalltheprinciplesinculcatedinthiswork，ifdomesticanimals，whenexposedtonewconditionsandcompelledtostrugglefortheirownwantsagainstahostofforeigncompetitors，werenotmodifiedinthecourseoftime。Itshouldalsoberememberedthatmanycharacterslielatentinallorganicbeings，readytobeevolvedunderfittingconditions;andinbreedsmodifiedwithinrecenttimes，thetendencytoreversionisparticularlystrong。Buttheantiquityofsomeofourbreedsclearlyprovesthattheyremainnearlyconstantaslongastheirconditionsofliferemainthesame。
Ithasbeenboldlymaintainedbysomeauthorsthattheamountofvariationtowhichourdomesticproductionsareliableisstrictlylimited;butthisisanassertionrestingonlittleevidence。Whetherornottheamountofchangeinanyparticulardirectionislimited，thetendencytogeneralvariabilityis，asfaraswecanjudge，unlimited。Cattle，sheep，andpigshavevariedunderdomesticationfromtheremotestperiod，asshownbytheresearchesofRutimeyerandothers;yettheseanimalshavebeenimprovedtoanunparalleleddegree，withinquiterecenttimes，andthisimpliescontinuedvariabilityofstructure。Wheat，asweknowfromtheremainsfoundintheSwisslake—
dwellings，isoneofthemostancientlycultivatedplants，yetatthepresentdaynewandbettervarietiesfrequentlyarise。Itmaybethatanoxwillneverbeproducedoflargersizeandfinerproportions，oraracehorsefleeter，thanourpresentanimals，oragooseberrylargerthantheLondonvariety;buthewouldbeaboldmanwhowouldassertthattheextremelimitintheserespectshasbeenfinallyattained。Withflowersandfruitithasrepeatedlybeenassertedthatperfectionhasbeenreached，butthestandardhassoonbeenexcelled。Abreedofpigeonsmayneverbeproducedwithabeakshorterthanthatofthepresentshort—facedtumbler，orwithonelongerthanthatoftheEnglishcarrier，forthesebirdshaveweakconstitutionsandarebadbreeders;
butshortnessandlengthofbeakarethepointswhichhavebeensteadilyimprovedduringthelast150years，andsomeofthebestjudgesdenythatthegoalhasyetbeenreached。Fromreasonswhichcouldbeassigned，itisprobablethatpartswhichhavenowreachedtheirmaximumdevelopment，might，afterremainingconstantduringalongperiod，varyagaininthedirectionofincreaseundernewconditionsoflife。Buttheremustbe，asMr。Wallacehasremarkedwithmuchtruth（28/3。’TheQuarterlyJournalofScience’October1867page486。），alimittochangeincertaindirectionsbothwithnaturalanddomesticproductions;forinstance，theremustbealimittothefleetnessofanyterrestrialanimal，asthiswillbedeterminedbythefrictiontobeovercome，theweighttobecarried，andthepowerofcontractioninthemuscularfibres。TheEnglishracehorsemayhavereachedthislimit;butitalreadysurpassesinfleetnessitsownwildprogenitorandallotherequinespecies。Theshort—facedtumbler—pigeonhasabeakshorter，andthecarrierabeaklonger，relativelytothesizeoftheirbodies，thanthatofanynaturalspeciesofthefamily。Ourapples，pearsandgooseberriesbearlargerfruitthanthoseofanynaturalspeciesofthesamegenera;andsoinmanyothercases。
Itisnotsurprising，seeingthegreatdifferencebetweenmanydomesticbreeds，thatsomefewnaturalistshaveconcludedthateachisdescendedfromadistinctaboriginalstock，moreespeciallyastheprincipleofselectionhasbeenignored，andthehighantiquityofman，asabreederofanimals，hasonlyrecentlybecomeknown。Mostnaturalists，however，freelyadmitthatourvariousbreeds，howeverdissimilar，aredescendedfromasinglestock，althoughtheydonotknowmuchabouttheartofbreeding，cannotshowtheconnectinglinks，norsaywhereandwhenthebreedsarose。Yetthesesamenaturalistsdeclare，withanairofphilosophicalcaution，thattheywillneveradmitthatonenaturalspecieshasgivenbirthtoanotheruntiltheybeholdallthetransitionalsteps。Fanciersuseexactlythesamelanguagewithrespecttodomesticbreeds;thus，anauthorofanexcellenttreatiseonpigeonssayshewillneverallowthatthecarrierandfantailarethedescendantsofthewildrock—pigeon，untilthetransitionshave\"actuallybeenobserved，andcanberepeatedwhenevermanchoosestosetaboutthetask。\"Nodoubtitisdifficulttorealisethatslightchangesaddedupduringlongcenturiescanproducesuchgreatresults;buthewhowishestounderstandtheoriginofdomesticbreedsorofnaturalspeciesmustovercomethisdifficulty。
Thecauseswhichexciteandthelawswhichgovernvariabilityhavebeendiscussedsolately，thatIneedhereonlyenumeratetheleadingpoints。Asdomesticatedorganismsaremuchmoreliabletoslightdeviationsofstructureandtomonstrositiesthanspecieslivingundertheirnaturalconditions，andaswidely—rangingspeciesgenerallyvarymorethanthosewhichinhabitrestrictedareas，wemayinferthatvariabilitymainlydependsonchangedconditionsoflife。Wemustnotoverlooktheeffectsoftheunequalcombinationofthecharactersderivedfrombothparents，orreversiontoformerprogenitors。Changedconditionshaveanespecialtendencytorenderthereproductiveorgansmoreorlessimpotent，asshowninthechapterdevotedtothissubject;andtheseorgansconsequentlyoftenfailtotransmitfaithfullytheparentalcharacters。Changedconditionsalsoactdirectlyanddefinitelyontheorganisation，sothatallornearlyalltheindividualsofthesamespeciesthusexposedbecomemodifiedinthesamemanner;butwhythisorthatpartisespeciallyaffectedwecanseldomoreversay。Inmostcases，however，achangeintheconditionsseemstoactindefinitely，causingdiversifiedvariationsinnearlythesamemannerasexposuretocoldortheabsorptionofthesamepoisonaffectsdifferentindividualsindifferentways。Wehavereasontosuspectthatanhabitualexcessofhighly—nutritiousfood，oranexcessrelativelytothewearandtearoftheorganisationfromexercise，isapowerfulexcitingcauseofvariability。Whenweseethesymmetricalandcomplexoutgrowths，causedbyaminutedropofthepoisonofagall—insect，wemaybelievethatslightchangesinthechemicalnatureofthesaporbloodwouldleadtoextraordinarymodificationsofstructure。
Theincreaseduseofamusclewithitsvariousattachedparts，andtheincreasedactivityofaglandorotherorgan，leadtotheirincreaseddevelopment。Disusehasacontraryeffect。Withdomesticatedproductions，althoughtheirorganssometimesbecomerudimentarythroughabortion，wehavenoreasontosupposethatthishaseverfollowedsolelyfromdisuse。Withnaturalspecies，onthecontrary，manyorgansappeartohavebeenrenderedrudimentarythroughdisuse，aidedbytheprincipleoftheeconomyofgrowthtogetherwithintercrossing。Completeabortioncanbeaccountedforonlybythehypothesisgiveninthelastchapter，namely，thefinaldestructionofthegermsorgemmulesofuselessparts。Thisdifferencebetweenspeciesanddomesticvarietiesmaybepartlyaccountedforbydisusehavingactedonthelatterforaninsufficientlengthoftime，andpartlyfromtheirexemptionfromanyseverestruggleforexistenceentailingrigideconomyinthedevelopmentofeachpart，towhichallspeciesundernaturearesubjected。
Neverthelessthelawofcompensationorbalancement，whichlikewisedependsontheeconomyofgrowth，apparentlyhasaffectedtoacertainextentourdomesticatedproductions。
Asalmosteverypartoftheorganisationbecomeshighlyvariableunderdomestication，andasvariationsareeasilyselectedbothconsciouslyandunconsciously，itisverydifficulttodistinguishbetweentheeffectsoftheselectionofindefinitevariationsandthedirectactionoftheconditionsoflife。Forinstance，itispossiblethatthefeetofourwater—dogsandoftheAmericandogswhichhavetotravelmuchoverthesnow，mayhavebecomepartiallywebbedfromthestimulusofwidelyextendingtheirtoes;butitismoreprobablethatthewebbing，likethemembranebetweenthetoesofcertainpigeons，spontaneouslyappearedandwasafterwardsincreasedbythebestswimmersandthebestsnow—travellersbeingpreservedduringmanygenerations。
Afancierwhowishedtodecreasethesizeofhisbantamsortumbler—pigeonswouldneverthinkofstarvingthem，butwouldselectthesmallestindividualswhichspontaneouslyappeared。Quadrupedsaresometimesborndestituteofhairandhairlessbreedshavebeenformed，butthereisnoreasontobelievethatthisiscausedbyahotclimate。Withinthetropicsheatoftencausessheeptolosetheirfleeces;ontheotherhand，wetandcoldactasadirectstimulustothegrowthofhair;butwhowillpretendtodecidehowfarthethickfurofarcticanimals，ortheirwhitecolour，isduetothedirectactionofasevereclimate，andhowfartothepreservationofthebest—protectedindividualsduringalongsuccessionofgenerations？
Ofallthelawsgoverningvariability，thatofcorrelationisoneofthemostimportant。Inmanycasesofslightdeviationsofstructureaswellasofgravemonstrosities，wecannotevenconjecturewhatisthenatureofthebondofconnexion。Butbetweenhomologousparts——betweentheforeandhindlimbs——
betweenthehair，hoofs，horns，andteeth——whicharecloselysimilarduringtheirearlydevelopmentandwhichareexposedtosimilarconditions，wecanseethattheywouldbeeminentlyliabletobemodifiedinthesamemanner。
Homologousparts，fromhavingthesamenature，areapttoblendtogether，and，whenmanyexist，tovaryinnumber。
Althougheveryvariationiseitherdirectlyorindirectlycausedbysomechangeinthesurroundingconditions，wemustneverforgetthatthenatureoftheorganisationwhichisactedon，isbyfarthemoreimportantfactorintheresult。Weseethisindifferentorganisms，whichwhenplacedundersimilarconditionsvaryinadifferentmanner，whilstclosely—alliedorganismsunderdissimilarconditionsoftenvaryinnearlythesamemanner。Weseethis，inthesamemodificationfrequentlyreappearinginthesamevarietyatlongintervalsoftime，andlikewiseintheseveralstrikingcasesgivenofanalogousorparallelvariations。Althoughsomeoftheselattercasesareduetoreversion，otherscannotthusbeaccountedfor。
Fromtheindirectactionofchangedconditionsontheorganisation，owingtothereproductiveorgansbeingthusaffected——fromthedirectactionofsuchconditions，andthesewillcausetheindividualsofthesamespecieseithertovaryinthesamemanner，ordifferentlyinaccordancewithslightdifferencesintheirconstitution——fromtheeffectsoftheincreasedordecreaseduseofparts——andfromcorrelation，——thevariabilityofourdomesticatedproductionsiscomplicatedtoanextremedegree。Thewholeorganisationbecomesslightlyplastic。Althougheachmodificationmusthaveitsownexcitingcause，andthougheachissubjectedtolaw，yetwecansorarelytracethepreciserelationbetweencauseandeffect，thatwearetemptedtospeakofvariationsasiftheyarosespontaneously。Wemayevencallthemaccidental，butthismustbeonlyinthesenseinwhichwesaythatafragmentofrockdroppedfromaheightowesitsshapetoaccident。
Itmaybeworthwhilebrieflytoconsidertheresultoftheexposuretounnaturalconditionsofalargenumberofanimalsofthesamespeciesandallowedtocrossfreelywithnoselectionofanykind，andafterwardstoconsidertheresultwhenselectionisbroughtintoplay。Letussupposethat500wildrock—pigeonswereconfinedintheirnativelandinanaviaryandfedinthesamemanneraspigeonsusuallyare;andthattheywerenotallowedtoincreaseinnumber。Aspigeonspropagatesorapidly，Isupposethatathousandorfifteenhundredbirdswouldhavetobeannuallykilled。Afterseveralgenerationshadbeenthusreared，wemayfeelsurethatsomeoftheyoungbirdswouldvary，andthevariationswouldtendtobeinherited;foratthepresentdayslightdeviationsofstructureoftenoccurandareinherited。Itwouldbetediouseventoenumeratethemultitudeofpointswhichstillgoonvaryingorhaverecentlyvaried。Manyvariationswouldoccurincorrelationwithoneanother，asthelengthofthewingandtailfeathers——thenumberoftheprimarywing—feathers，aswellasthenumberandbreadthoftheribs，incorrelationwiththesizeandformofthebody——thenumberofthescutellaewiththesizeofthefeet——thelengthofthetonguewiththelengthofthebeak——thesizeofthenostrilsandeyelidsandtheformoflowerjawincorrelationwiththedevelopmentofwattle——thenakednessoftheyoungwiththefuturecolouroftheplumage——thesizeofthefeetwiththatofthebeak，andothersuchpoints。Lastly，asourbirdsaresupposedtobeconfinedinanaviary，theywouldusetheirwingsandlegsbutlittle，andcertainpartsoftheskeleton，suchasthesternum，scapulaeandfeet，wouldinconsequencebecomeslightlyreducedinsize。
Asinourassumedcasemanybirdshavetobeindiscriminatelykilledeveryyear，thechancesareagainstanynewvarietysurvivinglongenoughtobreed。
Andasthevariationswhichariseareofanextremelydiversifiednature，thechancesareverygreatagainsttwobirdspairingwhichhavevariedinthesamemanner;nevertheless，avaryingbirdevenwhennotthuspairedwouldoccasionallytransmititscharactertoitsyoung;andthesewouldnotonlybeexposedtothesameconditionswhichfirstcausedthevariationinquestiontoappear，butwouldinadditioninheritfromtheirmodifiedparentatendencyagaintovaryinthesamemanner。Sothat，iftheconditionsdecidedlytendedtoinducesomeparticularvariation，allthebirdsmightinthecourseoftimebecomesimilarlymodified。Butafarcommonerresultwouldbe，thatonebirdwouldvaryinonewayandanotherbirdinanotherway;onewouldbebornwithabeakalittlelonger，andanotherwithashorterbeak;onewouldgainsomeblackfeathers，anothersomewhiteorredfeathers。Andasthesebirdswouldbecontinuallyintercrossing，thefinalresultwouldbeabodyofindividualsdifferingfromeachotherinmanyways，butonlyslightly;yetmorethandidtheoriginalrock—pigeons。Buttherewouldnotbetheleasttendencytowardstheformationofseveraldistinctbreeds。
Iftwoseparatelotsofpigeonsweretreatedinthemannerjustdescribed，oneinEnglandandtheotherinatropicalcountry，thetwolotsbeingsuppliedwithdifferentkindsoffood，wouldtheyaftermanygenerationsdiffer？Whenwereflectonthecasesgiveninthetwenty—thirdchapter，andonsuchfactsasthedifferenceinformertimesbetweenthebreedsofcattle，sheep，etc。，inalmosteverydistrictofEurope，wearestronglyinclinedtoadmitthatthetwolotswouldbedifferentlymodifiedthroughtheinfluenceofclimateandfood。Buttheevidenceonthedefiniteactionofchangedconditionsisinmostcasesinsufficient;and，withrespecttopigeons，Ihavehadtheopportunityofexaminingalargecollectionofdomesticatedkinds，senttomebySirW。
ElliotfromIndia，andtheyvariedinaremarkablysimilarmannerwithourEuropeanbirds。
Iftwodistinctbreedsweremingledtogetherinequalnumbers，thereisreasontosuspectthattheywouldtoacertainextentpreferpairingwiththeirownkind;buttheywouldoftenintercross。Fromthegreatervigourandfertilityofthecrossedoffspring，thewholebodywouldbythismeansbecomeinterblendedsoonerthanwouldotherwisehaveoccurred。Fromcertainbreedsbeingprepotentoverothers，itdoesnotfollowthattheinterblendedprogenywouldbestrictlyintermediateincharacter。Ihave，also，provedthattheactofcrossinginitselfgivesastrongtendencytoreversion，sothatthecrossedoffspringwouldtendtoreverttothestateoftheaboriginalrock—
pigeon;andinthecourseoftimetheywouldprobablybenotmuchmoreheterogeneousincharacterthaninourfirstcase，whenbirdsofthesamebreedwereconfinedtogether。
Ihavejustsaidthatthecrossedoffspringwouldgaininvigourandfertility。Fromthefactsgivenintheseventeenthchaptertherecanbenodoubtofthisfact;andtherecanbelittledoubt，thoughtheevidenceonthisheadisnotsoeasilyacquired，thatlong—continuedcloseinterbreedingleadstoevilresults。Withhermaphroditesofallkinds，ifthesexualelementsofthesameindividualhabituallyactedoneachother，theclosestpossibleinterbreedingwouldbeperpetual。Butweshouldbearinmindthatthestructureofallhermaphroditeanimals，asfarasIcanlearn，permitsandfrequentlynecessitatesacrosswithadistinctindividual。Withhermaphroditeplantsweincessantlymeetwithelaborateandperfectcontrivancesforthissameend。Itisnoexaggerationtoassertthat，iftheuseofthetalonsandtusksofacarnivorousanimal，oroftheplumesandhooksonaseed，maybesafelyinferredfromtheirstructure，wemaywithequalsafetyinferthatmanyflowersareconstructedfortheexpresspurposeofensuringacrosswithadistinctplant。Fromthesevariousconsiderations，nottomentiontheresultofalongseriesofexperimentswhichIhavetried，theconclusionarrivedatinthechapterjustreferredto——namely，thatgreatgoodofsomekindisderivedfromthesexualconcourseofdistinctindividuals——mustbeadmitted。
Toreturntoourillustration:wehavehithertoassumedthatthebirdswerekeptdowntothesamenumberbyindiscriminateslaughter;butiftheleastchoicebepermittedintheirpreservation，thewholeresultwillbechanged。
Shouldtheownerobserveanyslightvariationinoneofhisbirds，andwishtoobtainabreedthuscharacterised，hewouldsucceedinasurprisinglyshorttimebycarefulselection。Asanypartwhichhasoncevariedgenerallygoesonvaryinginthesamedirection，itiseasy，bycontinuallypreservingthemoststronglymarkedindividuals，toincreasetheamountofdifferenceuptoahigh，predeterminedstandardofexcellence。Thisismethodicalselection。
Iftheowneroftheaviary，withoutanythoughtofmakinganewbreed，simplyadmired，forinstance，short—beakedmorethanlong—beakedbirds，hewould，whenhehadtoreducethenumber，generallykillthelatter;andtherecanbenodoubtthathewouldthusinthecourseoftimesensiblymodifyhisstock。
Itisimprobable，iftwomenweretokeeppigeonsandactinthismanner，thattheywouldpreferexactlythesamecharacters;theywould，asweknow，oftenpreferdirectlyoppositecharacters，andthetwolotswouldultimatelycometodiffer。Thishasactuallyoccurredwithstrainsorfamiliesofcattle，sheep，andpigeons，whichhavebeenlongkeptandcarefullyattendedtobydifferentbreeders，withoutanywishontheirparttoformnewanddistinctsub—breeds。
Thisunconsciouskindofselectionwillmoreespeciallycomeintoactionwithanimalswhicharehighlyserviceabletoman;foreveryonetriestogetthebestdogs，horses，cows，orsheep，withoutthinkingabouttheirfutureprogeny，yettheseanimalswouldtransmitmoreorlesssurelytheirgoodqualitiestotheiroffspring。Norisanyonesocarelessastobreedfromhisworstanimals。Evensavages，whencompelledfromextremewanttokillsomeoftheiranimals，woulddestroytheworstandpreservethebest。Withanimalskeptforuseandnotformereamusement，differentfashionsprevailindifferentdistricts，leadingtothepreservation，andconsequentlytothetransmission，ofallsortsoftriflingpeculiaritiesofcharacter。Thesameprocesswillhavebeenpursuedwithourfruit—treesandvegetables，forthebestwillalwayshavebeenthemostlargelycultivated，andwilloccasionallyhaveyieldedseedlingsbetterthantheirparents。
Thedifferentstrains，justalludedto，whichhavebeenactuallyproducedbybreederswithoutanywishontheirparttoobtainsucharesult，affordexcellentevidenceofthepowerofunconsciousselection。Thisformofselectionhasprobablyledtofarmoreimportantresultsthanmethodicalselection，andislikewisemoreimportantunderatheoreticalpointofviewfromcloselyresemblingnaturalselection。Forduringthisprocessthebestormostvaluedindividualsarenotseparatedandpreventedfromcrossingwithothersofthesamebreed，butaresimplypreferredandpreserved;yetthisinevitablyleadstotheirgradualmodificationandimprovement;sothatfinallytheyprevail，totheexclusionoftheoldparent—form。
Withourdomesticatedanimalsnaturalselectioncheckstheproductionofraceswithanyinjuriousdeviationofstructure。Inthecaseofanimalswhich，frombeingkeptbysavagesorsemi—civilisedpeople，havetoprovidelargelyfortheirownwantsunderdifferentcircumstances，naturalselectionwillhaveplayedamoreimportantpart。Henceitprobablyisthattheyoftencloselyresemblenaturalspecies。
Asthereisnolimittoman’sdesiretopossessanimalsandplantsmoreandmoreusefulinanyrespect，andasthefancieralwayswishes，owingtofashionsrunningintoextremes，toproduceeachcharactermoreandmorestronglypronounced，thereis，throughtheprolongedactionofmethodicalandunconsciousselection，aconstanttendencyineverybreedtobecomemoreandmoredifferentfromitsparent—stock;andwhenseveralbreedshavebeenproducedandarevaluedfordifferentqualities，todiffermoreandmorefromeachother。ThisleadstoDivergenceofCharacter。Asimprovedsub—varietiesandracesareslowlyformed，theolderandlessimprovedbreedsareneglectedanddecreaseinnumber。Whenfewindividualsofanybreedexistwithinthesamelocality，closeinterbreeding，bylesseningtheirvigourandfertility，aidsintheirfinalextinction。Thustheintermediatelinksarelost，andtheremainingbreedsgaininDistinctnessofCharacter。
InthechaptersonthePigeon，itwasprovedbyhistoricalevidenceandbytheexistenceofconnectingsub—varietiesindistantlandsthatseveralbreedshavesteadilydivergedincharacter，andthatmanyoldandintermediatesub—
breedshavebeenlost。Othercasescouldbeadducedoftheextinctionofdomesticbreeds，asoftheIrishwolf—dog，theoldEnglishhound，andoftwobreedsinFrance，oneofwhichwasformerlyhighlyvalued。（28/4。M。RufzdeLavisonin’Bull。Soc。Imp。d’Acclimat。’December1862page1009。）Mr。
Pickeringremarks（28/5。’RacesofMan’1850page315。）that\"thesheepfiguredonthemostancientEgyptianmonumentsisunknownatthepresentday;
andatleastonevarietyofthebullock，formerlyknowninEgypt，hasinlikemannerbecomeextinct。\"SoithasbeenwithsomeanimalsandwithseveralplantscultivatedbytheancientinhabitantsofEuropeduringtheneolithicperiod。InPeru，VonTschudi（28/6。’TravelsinPeru’Englishtranslationpage177。）foundincertaintombs，apparentlypriortothedynastyoftheIncas，twokindsofmaizenotnowknowninthecountry。Withourflowersandculinaryvegetables，theproductionofnewvarietiesandtheirextinctionhasincessantlyrecurred。Atthepresenttimeimprovedbreedssometimesdisplaceolderbreedsatanextraordinarilyrapidrate;ashasrecentlyoccurredthroughoutEnglandwithpigs。TheLonghorncattleintheirnativehomewere\"suddenlysweptawayasifbysomemurderouspestilence，\"bytheintroductionofShorthorns。（28/7。Youatton’Cattle’1834page200。OnPigssee’Gardener’sChronicle’1854page410。）
Whatgrandresultshavefollowedfromthelong—continuedactionofmethodicalandunconsciousselection，regulatedtoacertainextentbynaturalselection，weseeoneverysideofus。Comparethemanyanimalsandplantswhicharedisplayedatourexhibitionswiththeirparent—formswhentheseareknown，orconsultoldhistoricalrecordswithrespecttotheirformerstate。Mostofourdomesticatedanimalshavegivenrisetonumerousanddistinctraces，butthosewhichcannotbeeasilysubjectedtoselectionmustbeexcepted——suchascats，thecochinealinsect，andthehive—bee。Inaccordancewithwhatweknowoftheprocessofselection，theformationofourmanyraceshasbeenslowandgradual。Themanwhofirstobservedandpreservedapigeonwithitsoesophagusalittleenlarged，itsbeakalittlelonger，oritstailalittlemoreexpandedthanusual，neverdreamedthathehadmadethefirststepinthecreationofapouter，carrier，andfantail—pigeon。Mancancreatenotonlyanomalousbreeds，butothershavingtheirwholestructureadmirablyco—
ordinatedforcertainpurposes，suchastheracehorseanddray—horse，orthegreyhoundandbulldog。Itisbynomeansnecessarythateachsmallchangeofstructurethroughoutthebody，leadingtowardsexcellence，shouldsimultaneouslyariseandbeselected。Althoughmanseldomattendstodifferencesinorganswhichareimportantunderaphysiologicalpointofview，yethehassoprofoundlymodifiedsomebreeds，thatassuredly，iffoundwild，theywouldberankedasdistinctgenera。
Thebestproofofwhatselectionhaseffectedisperhapsaffordedbythefactthatwhateverpartorqualityinanyanimal，andmoreespeciallyinanyplant，ismostvaluedbyman，thatpartorqualitydiffersmostintheseveralraces。
Thisresultiswellseenbycomparingtheamountofdifferencebetweenthefruitsproducedbytheseveralvarietiesoffruit—trees，betweentheflowersofourflower—gardenplants，betweentheseeds，roots，orleavesofourculinaryandagriculturalplants，incomparisonwiththeotherandnotvaluedpartsofthesamevarieties。StrikingevidenceofadifferentkindisaffordedbythefactascertainedbyOswaldHeer（28/8。’DiePflanzenderPfahlbauten’
1865。）namely，thattheseedsofalargenumberofplants，——wheat，barley，oats，peas，beans，lentils，poppies，——cultivatedfortheirseedbytheancientLake—inhabitantsofSwitzerland，wereallsmallerthantheseedsofourexistingvarieties。RutimeyerhasshownthatthesheepandcattlewhichwerekeptbytheearlierLake—inhabitantswerelikewisesmallerthanourpresentbreeds。InthemiddensofDenmark，theearliestdogofwhichtheremainshavebeenfoundwastheweakest;thiswassucceededduringtheBronzeagebyastrongerkind，andthisagainduringtheIronagebyonestillstronger。ThesheepofDenmarkduringtheBronzeperiodhadextraordinarilyslenderlimbs，andthehorsewassmallerthanourpresentanimal。（28/9。Morlot’Soc。Vaud。
desScien。Nat。’Mars1860page298。）Nodoubtinmostofthesecasesthenewandlargerbreedswereintroducedfromforeignlandsbytheimmigrationofnewhordesofmen。Butitisnotprobablethateachlargerbreed，whichinthecourseoftimehassupplantedapreviousandsmallerbreed，wasthedescendantofadistinctandlargerspecies;itisfarmoreprobablethatthedomesticracesofourvariousanimalsweregraduallyimprovedindifferentpartsofthegreatEuropaeo—Asiaticcontinent，andthencespreadtoothercountries。Thisfactofthegradualincreaseinsizeofourdomesticanimalsisallthemorestrikingascertainwildorhalf—wildanimals，suchasred—deer，aurochs，park—cattle，andboars（28/10。Rutimeyer’DieFaunaderPfahlbauten’1861s。
30。）havewithinnearlythesameperioddecreasedinsize。