Howlongatimeitwouldrequirebeforesuchacrossedbodyofanimalswouldassumeauniformcharacterwithinalimitedarea，noonecansay;thattheywouldultimatelybecomeuniformfromfreeintercrossing，andfromthesurvivalofthefittest，wemayfeelassured;butthecharactersthusacquiredwouldrarelyornever，asmaybeinferredfromthepreviousconsiderations，beexactlyintermediatebetweenthoseofthetwoparent—breeds。Withrespecttotheveryslightdifferencesbywhichtheindividualsofthesamesub—variety，orevenofalliedvarieties，arecharacterised，itisobviousthatfreecrossingwouldsoonobliteratesuchsmalldistinctions。Theformationofnewvarieties，independentlyofselection，wouldalsothusbeprevented;exceptwhenthesamevariationcontinuallyrecurredfromtheactionofsomestronglypredisposingcause。Wemaythereforeconcludethatfreecrossinghasinallcasesplayedanimportantpartingivinguniformityofcharactertoallthemembersofthesamedomesticraceandofthesamenaturalspecies，thoughlargelygovernedbynaturalselectionandbythedirectactionofthesurroundingconditions。
ONTHEPOSSIBILITYOFALLORGANICBEINGSOCCASIONALLYINTERCROSSING。
Butitmaybeasked，canfreecrossingoccurwithhermaphroditeanimalsandplants？Allthehigheranimals，andthefewinsectswhichhavebeendomesticated，haveseparatesexes，andmustinevitablyuniteforeachbirth。
Withrespecttothecrossingofhermaphrodites，thesubjectistoolargeforthepresentvolume，butinthe’OriginofSpecies’Ihavegivenashortabstractofthereasonswhichinducemetobelievethatallorganicbeingsoccasionallycross，thoughperhapsinsomecasesonlyatlongintervalsoftime。（15/14。Withrespecttoplants，anadmirableessayonthissubject（DieGeschlechter—VertheilungbeidenPflanzen:1867）hasbeenpublishedbyDr。
HildebrandwhoarrivesatthesamegeneralconclusionsasIhavedone。Variousothertreatiseshavesinceappearedonthesamesubject，moreespeciallybyHermannMullerandDelpino。）Iwillmerelyrecallthefactthatmanyplants，thoughhermaphroditeinstructure，areunisexualinfunction;——suchasthosecalledbyC。K。SprengelDICHOGAMOUS，inwhichthepollenandstigmaofthesameflowerarematuredatdifferentperiods;orthosecalledbymeRECIPROCALLYDIMORPHIC，inwhichtheflower’sownpollenisnotfittedtofertiliseitsownstigma;oragain，themanykindsinwhichcuriousmechanicalcontrivancesexist，effectuallypreventingself—fertilisation。Thereare，however，manyhermaphroditeplantswhicharenotinanywayspeciallyconstructedtofavourintercrossing，butwhichneverthelesscomminglealmostasfreelyasanimalswithseparatedsexes。Thisisthecasewithcabbages，radishes，andonions，asIknowfromhavingexperimentedonthem:eventhepeasantsofLiguriasaythatcabbagesmustbeprevented"fromfallinginlove"
witheachother。Intheorangetribe，Gallesio（15/15。’TeoriadellaRiproduzioneVegetal’1816page12。）remarksthattheameliorationofthevariouskindsischeckedbytheircontinualandalmostregularcrossing。Soitiswithnumerousotherplants。
Ontheotherhand，somecultivatedplantsrarelyorneverintercross，forinstance，thecommonpeaandsweet—pea（Lathyrusodoratus）;yettheirflowersarecertainlyadaptedforcrossfertilisation。Thevarietiesofthetomatoandaubergine（Solanum）andthepimenta（Pimentavulgaris？）aresaid（15/16。
Verlot’DesVarietes’1865page72。）nevertocross，evenwhengrowingalongsideoneanother。Butitshouldbeobservedthattheseareallexoticplants，andwedonotknowhowtheywouldbehaveintheirnativecountrywhenvisitedbytheproperinsects。Withrespecttothecommonpea，Ihaveascertainedthatitisrarelycrossedinthiscountryowingtoprematurefertilisation。Thereexist，however，someplantswhichundertheirnaturalconditionsappeartobealwaysself—fertilised，suchastheBeeOphrys（Ophrysapifera）andafewotherOrchids;yettheseplantsexhibittheplainestadaptationsforcross—fertilisation。Again，somefewplantsarebelievedtoproduceonlyclosedflowers，calledcleistogene，whichcannotpossiblybecrossed。ThiswaslongthoughttobethecasewiththeLeersiaoryzoides（15/17。DuvalJouve’Bull。Soc。Bot。deFrance’tome101863page194。WithrespecttotheperfectflowerssettingseedseeDr。Aschersonin’Bot。
Zeitung’1864page350。），butthisgrassisnowknownoccasionallytoproduceperfectflowers，whichsetseed。
Althoughsomeplants，bothindigenousandnaturalised，rarelyorneverproduceflowers，oriftheyflowerneverproduceseeds，yetnoonedoubtsthatphanerogamicplantsareadaptedtoproduceflowers，andtheflowerstoproduceseed。Whentheyfail，webelievethatsuchplantsunderdifferentconditionswouldperformtheirproperfunction，orthattheyformerlydidso，andwilldosoagain。Onanalogousgrounds，Ibelievethattheflowersintheabovespecifiedanomalouscaseswhichdonotnowintercross，eitherwoulddosooccasionallyunderdifferentconditions，orthattheyformerlydidso——themeansforaffectingthisbeinggenerallystillretained——andwillagainintercrossatsomefutureperiod，unlessindeedtheybecomeextinct。Onthisviewalone，manypointsinthestructureandactionofthereproductiveorgansinhermaphroditeplantsandanimalsareintelligible，——forinstance，thefactofthemaleandfemaleorgansneverbeingsocompletelyenclosedastorenderaccessfromwithoutimpossible。Hencewemayconcludethatthemostimportantofallthemeansforgivinguniformitytotheindividualsofthesamespecies，namely，thecapacityofoccasionallyintercrossing，ispresent，orhasbeenformerlypresent，withallorganicbeings，except，perhaps，someofthelowest。
[ONCERTAINCHARACTERSNOTBLENDING。
Whentwobreedsarecrossedtheircharactersusuallybecomeintimatelyfusedtogether;butsomecharactersrefusetoblend，andaretransmittedinanunmodifiedstateeitherfrombothparentsorfromone。Whengreyandwhitemicearepaired，theyoungarepiebald，orpurewhiteorgrey，butnotofanintermediatetint;soitiswhenwhiteandcommoncollaredturtle—dovesarepaired。InbreedingGamefowls，agreatauthority，Mr。J。Douglas，remarks，"I
mayherestateastrangefact:ifyoucrossablackwithawhitegame，yougetbirdsofbothbreedsoftheclearestcolour。"SirR。Heroncrossedduringmanyyearswhite，black，brown，andfawn—colouredAngorarabbits，andneveroncegotthesecoloursmingledinthesameanimal，butoftenallfourcoloursinthesamelitter。（15/18。ExtractofaletterfromSirR。Heron1838givenmebyMr。Yarrell。Withrespecttomicesee’Annal。desSc。Nat。’tome1page180;andIhaveheardofothersimilarcases。Forturtle—dovesBoitardandCorbie’LesPigeons’etc。page238。FortheGamefowl’ThePoultryBook’1866
page128。ForcrossesoftaillessfowlsseeBechstein’Naturges。Deutsch。’b。
3s。403。Bronn’GeschichtederNatur’b。2s。170givesanalogousfactswithhorses。OnthehairlessconditionofcrossedSouthAmericandogsseeRengger’SaugethierevonParaguay’s。152;butIsawintheZoologicalGardensmongrels，fromasimilarcross，whichwerehairless，quitehairy，orhairyinpatches，thatis，piebaldwithhair。ForcrossesofDorkingandotherfowlssee’PoultryChronicle’volume2page355。Aboutthecrossedpigs，extractofletterfromSirR。HerontoMr。Yarrell。ForothercasesseeP。Lucas’L’Hered。Nat。’tome1page212。）Fromcaseslikethese，inwhichthecoloursofthetwoparentsaretransmittedquiteseparatelytotheoffspring，wehaveallsortsofgradations，leadingtocompletefusion。Iwillgiveaninstance:
agentlemanwithafaircomplexion，lighthairbutdarkeyes，marriedaladywithdarkhairandcomplexion:theirthreechildrenhaveverylighthair，butoncarefulsearchaboutadozenblackhairswerefoundscatteredinthemidstofthelighthairontheheadsofallthree。
Whenturnspitdogsandanconsheep，bothofwhichhavedwarfedlimbs，arecrossedwithcommonbreeds，theoffspringarenotintermediateinstructure，buttakeaftereitherparent。Whentaillessorhornlessanimalsarecrossedwithperfectanimals，itfrequently，butbynomeansinvariably，happensthattheoffspringareeitherfurnishedwiththeseorgansinaperfectstate，orarequitedestituteofthem。AccordingtoRengger，thehairlessconditionoftheParaguaydogiseitherperfectlyornotatalltransmittedtoitsmongreloffspring;butIhaveseenonepartialexceptioninadogofthisparentagewhichhadpartofitsskinhairy，andpartnaked，thepartsbeingdistinctlyseparatedasinapiebaldanimal。WhenDorkingfowlswithfivetoesarecrossedwithotherbreeds，thechickensoftenhavefivetoesononefootandfourontheother。SomecrossedpigsraisedbySirR。Heronbetweenthesolid—
hoofedandcommonpighadnotallfourfeetinanintermediatecondition，buttwofeetwerefurnishedwithproperlydivided，andtwowithunitedhoofs。
Analogousfactshavebeenobservedwithplants:MajorTrevorClarkecrossedthelittle，glabrous—leaved，annualstock（Matthiola），withpollenofalarge，red—flowered，rough—leaved，biennialstock，calledcocardeaubytheFrench，andtheresultwasthathalftheseedlingshadglabrousandtheotherhalfroughleaves，butnonehadleavesinanintermediatestate。Thattheglabrousseedlingsweretheproductoftherough—leavedvariety，andnotaccidentallyofthemother—plant’sownpollen，wasshownbytheirtallandstronghabitofgrowth。（15/19。’Internat。Hort。andBot。CongressofLondon’1866。）inthesucceedinggenerationsraisedfromtherough—leavedcrossedseedlings，someglabrousplantsappeared，showingthattheglabrouscharacter，thoughincapableofblendingwithandmodifyingtheroughleaves，wasallthetimelatentinthisfamilyofplants。Thenumerousplantsformerlyreferredto，whichIraisedfromreciprocalcrossesbetweenthepeloricandcommonAntirrhinum，offeranearlyparallelcase;forinthefirstgenerationalltheplantsresembledthecommonform，andinthenextgeneration，outofonehundredandthirty—sevenplants，twoalonewereinanintermediatecondition，theothersperfectlyresemblingeitherthepeloricorcommonform。MajorTrevorClarkealsofertilisedtheabove—mentionedred—floweredstockwithpollenfromthepurpleQueenstock，andabouthalftheseedlingsscarcelydifferedinhabit，andnotatallintheredcolouroftheflower，fromthemother—plant，theotherhalfbearingblossomsofarichpurple，closelylikethoseofthepaternalplant。Gartnercrossedmanywhiteandyellow—floweredspeciesandvarietiesofVerbascum;andthesecolourswereneverblended，buttheoffspringboreeitherpurewhiteorpureyellowblossoms;theformerinthelargerproportion。（15/20。’Bastarderzeugung’s。307。Kolreuter’DritteFortsetszung’s。34，39however，obtainedintermediatetintsfromsimilarcrossesinthegenusVerbascum。WithrespecttotheturnipsseeHerbert’Amaryllidaceae’1837page370。）Dr。Herbertraisedmanyseedlings，asheinformedme，fromSwedishturnipscrossedbytwoothervarieties，andtheseneverproducedflowersofanintermediatetint，butalwayslikeoneoftheirparents。Ifertilisedthepurplesweet—pea（Lathyrusodoratus），whichhasadarkreddish—purplestandard—petalandviolet—colouredwingsandkeel，withpollenofthepaintedladysweet—pea，whichhasapalecherry—colouredstandard，andalmostwhitewingsandkeel;andfromthesamepodItwiceraisedplantsperfectlyresemblingbothsorts;thegreaternumberresemblingthefather。Soperfectwastheresemblance，thatIshouldhavethoughttherehadbeensomemistake，iftheplantswhichwereatfirstidenticalwiththepaternalvariety，namely，thepainted—lady，hadnotlaterintheseasonproduced，asmentionedinaformerchapter，flowersblotchedandstreakedwithdarkpurple。Iraisedgrandchildrenandgreat—grandchildrenfromthesecrossedplants，andtheycontinuedtoresemblethepainted—lady，butduringlatergenerationsbecamerathermoreblotchedwithpurple，yetnonerevertedcompletelytotheoriginalmother—plant，thepurplesweet—pea。Thefollowingcaseisslightlydifferent，butstillshowsthesameprinciple:Naudin（15/21。
’NouvellesArchivesduMuseum’tome1page100。）raisednumeroushybridsbetweentheyellowLinariavulgarisandthepurpleL。purpurea，andduringthreesuccessivegenerationsthecolourskeptdistinctindifferentpartsofthesameflower。
Fromcasessuchastheforegoing，inwhichtheoffspringofthefirstgenerationperfectlyresembleeitherparent，wecomebyasmallsteptothosecasesinwhichdifferentlycolouredflowersborneonthesamerootresemblebothparents，andbyanothersteptothoseinwhichthesameflowerorfruitisstripedorblotchedwiththetwoparentalcolours，orbearsasinglestripeofthecolourorothercharacteristicqualityofoneoftheparent—forms。Withhybridsandmongrelsitfrequentlyorevengenerallyhappensthatonepartofthebodyresemblesmoreorlesscloselyoneparentandanotherparttheotherparent;andhereagainsomeresistencetofusion，or，whatcomestothesamething，somemutualaffinitybetweentheorganicatomsofthesamenature，apparentlycomesintoplay，forotherwiseallpartsofthebodywouldbeequallyintermediateincharacter。Soagain，whentheoffspringofhybridsormongrels，whicharethemselvesnearlyintermediateincharacter，reverteitherwhollyorbysegmentstotheirancestors，theprincipleoftheaffinityofsimilar，ortherepulsionofdissimilaratoms，mustcomeintoaction。Tothisprinciple，whichseemstobeextremelygeneral，weshallrecurinthechapteronpangenesis。
Itisremarkable，ashasbeenstronglyinsisteduponbyIsidoreGeoffroySt。
Hilaireinregardtoanimals，thatthetransmissionofcharacterswithoutfusionoccursveryrarelywhenspeciesarecrossed;Iknowofoneexceptionalone，namely，withthehybridsnaturallyproducedbetweenthecommonandhoodedcrow（Corvuscoroneandcornix），which，however，arecloselyalliedspecies，differinginnothingexceptcolour。NorhaveImetwithanywell—
ascertainedcasesoftransmissionofthiskind，evenwhenoneformisstronglyprepotentoveranother，whentworacesarecrossedwhichhavebeenslowlyformedbyman’sselection，andthereforeresembletoacertainextentnaturalspecies。Suchcasesaspuppiesinthesamelittercloselyresemblingtwodistinctbreeds，areprobablyduetosuperfoetation，——thatis，totheinfluenceoftwofathers。Allthecharactersaboveenumerated，whicharetransmittedinaperfectstatetosomeoftheoffspringandnottoothers，——
suchasdistinctcolours，nakednessofskin，smoothnessofleaves，absenceofhornsortail，additionaltoes，pelorism，dwarfedstructure，etc。，——haveallbeenknowntoappearsuddenlyinindividualanimalsandplants。Fromthisfact，andfromtheseveralslight，aggregateddifferenceswhichdistinguishdomesticracesandspeciesfromoneanother，notbeingliabletothispeculiarformoftransmission，wemayconcludethatitisinsomewayconnectedwiththesuddenappearanceofthecharactersinquestion。]
ONTHEMODIFICATIONOFOLDRACESANDTHEFORMATIONOFNEWRACESBYCROSSING。
Wehavehithertochieflyconsideredtheeffectsofcrossingingivinguniformityofcharacter;wemustnowlooktoanoppositeresult。Therecanbenodoubtthatcrossing，withtheaidofrigorousselectionduringseveralgenerations，hasbeenapotentmeansinmodifyingoldraces，andinformingnewones。LordOrfordcrossedhisfamousstudofgreyhoundsoncewiththebulldog，inordertogivethemcourageandperseverance。Certainpointershavebeencrossed，asIhearfromtheRev。W。D。Fox，withthefoxhound，togivethemdashandspeed。CertainstrainsofDorkingfowlshavehadaslightinfusionofGameblood;andIhaveknownagreatfancierwhoonasingleoccasioncrossedhisturbit—pigeonswithbarbs，forthesakeofgaininggreaterbreadthofbeak。
Intheforegoingcasesbreedshavebeencrossedonce，forthesakeofmodifyingsomeparticularcharacter;butwithmostoftheimprovedracesofthepig，whichnowbreedtrue，therehavebeenrepeatedcrosses，——forinstance，theimprovedEssexowesitsexcellencetorepeatedcrosseswiththeNeapolitan，togetherprobablywithsomeinfusionofChineseblood。（15/22。
Richardson’Pigs’1847pages37，42;S。Sidney’seditionof’YouattonthePig’1860page3。）SowithourBritishsheep:almostalltheraces，excepttheSouthdown，havebeenlargelycrossed;"this，infact，hasbeenthehistoryofourprincipalbreeds。"（15/23。SeeMr。W。C。Spooner’sexcellentpaperonCross—Breeding’JournalRoyalAgricult。Soc。’volume20part2:seealsoanequallygoodarticlebyMr。Ch。Howardin’Gardener’sChronicle’1860page320。）Togiveanexample，the"OxfordshireDowns"nowrankasanestablishedbreed。（15/24。’Gardener’sChronicle’1857pages649，652。）Theywereproducedabouttheyear1830bycrossing"HampshireandinsomeinstancesSouthdowneweswithCotswoldrams:"nowtheHampshireramwasitselfproducedbyrepeatedcrossesbetweenthenativeHampshiresheepandSouthdowns;andthelong—woolledCotswoldwereimprovedbycrosseswiththeLeicester，whichlatteragainisbelievedtohavebeenacrossbetweenseverallong—woolledsheep。Mr。Spooner，afterconsideringthevariouscaseswhichhavebeencarefullyrecorded，concludes，"thatfromajudiciouspairingofcross—bredanimalsitispracticabletoestablishanewbreed。"Onthecontinentthehistoryofseveralcrossedracesofcattleandofotheranimalshasbeenwellascertained。Togiveoneinstance:theKingofWurtemburg，aftertwenty—fiveyears’carefulbreeding，thatis，aftersixorsevengenerations，madeanewbreedofcattlefromacrossbetweenaDutchandaSwissbreed，combinedwithotherbreeds。（15/25。’BulletindeLaSoc。d’Acclimat。’1862tome9page463。
SeealsoforothercasesMM。MollandGayot’DuBoeuf’1860page32。）TheSebrightbantam，whichbreedsastrueasanyotherkindoffowl，wasformedaboutsixtyyearsagobyacomplicatedcross。（15/26。’PoultryChronicle’
volume21854page36。）DarkBrahmas，whicharebelievedbysomefancierstoconstituteadistinctspecies，wereundoubtedlyformed（15/27。’ThePoultryBook’byW。B。Tegetmeier1866page58。）intheUnitedStates，withinarecentperiod，byacrossbetweenChittagongsandCochins。WithplantsthereislittledoubtthattheSwede—turniporiginatedfromacross;andthehistoryofavarietyofwheat，raisedfromtwoverydistinctvarieties，andwhichaftersixyears’culturepresentedanevensample，hasbeenrecordedongoodauthority。（15/28。’Gardener’sChronicle’1852page765。）
Untillately，cautiousandexperiencedbreeders，thoughnotaversetoasingleinfusionofforeignblood，werealmostuniversallyconvincedthattheattempttoestablishanewrace，intermediatebetweentwowidelydistinctraces，washopeless"theyclungwithsuperstitioustenacitytothedoctrineofpurityofblood，believingittobethearkinwhichalonetruesafetycouldbefound。"
（15/29。Spoonerin’JournalRoyalAgricult。Soc。’volume20part2）Norwasthisconvictionunreasonable:whentwodistinctracesarecrossed，theoffspringofthefirstgenerationaregenerallynearlyuniformincharacter;
buteventhissometimesfailstobethecase，especiallywithcrosseddogsandfowls，theyoungofwhichfromthefirstaresometimesmuchdiversified。Ascross—bredanimalsaregenerallyoflargesizeandvigorous，theyhavebeenraisedingreatnumbersforimmediateconsumption。Butforbreedingtheyarefoundutterlyuseless;forthoughtheymaythemselvesbeuniformincharacter，theyyieldduringmanygenerationsastonishinglydiversifiedoffspring。Thebreederisdriventodespair，andconcludesthathewillneverformanintermediaterace。Butfromthecasesalreadygiven，andfromotherswhichhavebeenrecorded，itappearsthatpatiencealoneisnecessary;asMr。
Spoonerremarks，"natureopposesnobarriertosuccessfuladmixture;inthecourseoftime，bytheaidofselectionandcarefulweeding，itispracticabletoestablishanewbreed。"Aftersixorsevengenerationsthehoped—forresultwillinmostcasesbeobtained;buteventhenanoccasionalreversion，orfailuretokeeptrue，maybeexpected。Theattempt，however，willassuredlyfailiftheconditionsoflifebedecidedlyunfavourabletothecharactersofeitherparent—breed。（15/30。SeeColin’TraitedePhys。Comp。desAnimauxDomestiques’tome2page536，wherethissubjectiswelltreated。）
Althoughthegrandchildrenandsucceedinggenerationsofcross—bredanimalsaregenerallyvariableinanextremedegree，somecuriousexceptionstotherulehavebeenobservedbothwithcrossedracesandspecies。ThusBoitardandCorbie（15/31。’LesPigeons’page37。）assertthatfromaPouterandaRunt"aCavalierwillappear，whichwehaveclassedamongstpigeonsofpurerace，becauseittransmitsallitsqualitiestoitsposterity。"Theeditorofthe’PoultryChronicle’（15/32。Volume11854page101。）bredsomebluishfowlsfromablackSpanishcockandaMalayhen;andtheseremainedtruetocolour"generationaftergeneration。"TheHimalayanbreedofrabbitswascertainlyformedbycrossingtwosub—varietiesofthesilver—greyrabbit;althoughitsuddenlyassumeditspresentcharacter，whichdiffersmuchfromthatofeitherparent—breed，yetithaseversincebeeneasilyandtrulypropagated。I
crossedsomeLabradorandPenguinducks，andrecrossedthemongrelswithPenguins;afterwardsmostoftheducksrearedduringthreegenerationswerenearlyuniformincharacter，beingbrownwithawhitecrescenticmarkonthelowerpartofthebreast，andwithsomewhitespotsatthebaseofthebeak;
sothatbytheaidofalittleselectionanewbreedmighteasilyhavebeenformed。Withregardtocrossedvarietiesofplants，Mr。Beaton（15/33。
’CottageGardener’1856page110。）remarksthat"Melville’sextraordinarycrossbetweentheScotchkaleandanearlycabbageisastrueandgenuineasanyonrecord;"butinthiscasenodoubtselectionwaspractised。Gartner（15/34。’Bastarderzeugung’s。553。）hasgivenfivecasesofhybrids，inwhichtheprogenykeptconstant;andhybridsbetweenDianthusarmeriaanddeltoidesremainedtrueanduniformtothetenthgeneration。Dr。HerbertlikewiseshowedmeahybridfromtwospeciesofLoasawhichfromitsfirstproductionhadkeptconstantduringseveralgenerations。
Wehaveseeninthefirstchapter，thattheseveralkindsofdogsarealmostcertainlydescendedfrommorethanonespecies，andsoitiswithcattle，pigsandsomeotherdomesticatedanimals。Hencethecrossingofaboriginallydistinctspeciesprobablycameintoplayatanearlyperiodintheformationofourpresentraces。FromRutimeyer’sobservationstherecanbelittledoubtthatthisoccurredwithcattle;butinmostcasesoneformwillprobablyhaveabsorbedandobliteratedtheother，foritisnotlikelythatsemi—civilisedmenwouldhavetakenthenecessarypainstomodifybyselectiontheircommingled，crossed，andfluctuatingstock。Nevertheless，thoseanimalswhichwerebestadaptedtotheirconditionsoflifewouldhavesurvivedthroughnaturalselection;andbythismeanscrossingwilloftenhaveindirectlyaidedintheformationofprimevaldomesticatedbreeds。Withinrecenttimes，asfarasanimalsareconcerned，thecrossingofdistinctspecieshasdonelittleornothingtowardstheformationormodificationofourraces。Itisnotyetknownwhethertheseveralspeciesofsilk—mothwhichhavebeenrecentlycrossedinFrancewillyieldpermanentraces。Withplantswhichcanbemultipliedbybudsandcuttings，hybridisationhasdonewonders，aswithmanykindsofRoses，Rhododendrons，Pelargoniums，Calceolarias，andPetunias。
Nearlyalltheseplantscanbepropagatedbyseed，mostofthemfreely;butextremelyfewornonecometruebyseed。
Someauthorsbelievethatcrossingisthechiefcauseofvariability，——thatis，oftheappearanceofabsolutelynewcharacters。Somehavegonesofarastolookatitasthesolecause;butthisconclusionisdisprovedbythefactsgiveninthechapteronBud—variation。Thebeliefthatcharactersnotpresentineitherparentorintheirancestorsfrequentlyoriginatefromcrossingisdoubtful;thattheyoccasionallydosoisprobable;butthissubjectwillbemoreconvenientlydiscussedinafuturechapteronthecausesofVariability。
Acondensedsummaryofthisandofthethreefollowingchapters，togetherwithsomeremarksonHybridism，willbegiveninthenineteenthchapter。
CHAPTER2。XVI。
CAUSESWHICHINTERFEREWITHTHEFREECROSSINGOFVARIETIES——INFLUENCEOF
DOMESTICATIONONFERTILITY。
DIFFICULTIESINJUDGINGOFTHEFERTILITYOFVARIETIESWHENCROSSED。
VARIOUSCAUSESWHICHKEEPVARIETIESDISTINCT，ASTHEPERIODOFBREEDINGAND
SEXUALPREFERENCE。
VARIETIESOFWHEATSAIDTOBESTERILEWHENCROSSED。
VARIETIESOFMAIZE，VERBASCUM，HOLLYHOCK，GOURDS，MELONS，ANDTOBACCO，RENDEREDINSOMEDEGREEMUTUALLYSTERILE。
DOMESTICATIONELIMINATESTHETENDENCYTOSTERILITYNATURALTOSPECIESWHEN
CROSSED。
ONTHEINCREASEDFERTILITYOFUNCROSSEDANIMALSANDPLANTSFROMDOMESTICATION
ANDCULTIVATION。
Thedomesticatedracesofbothanimalsandplants，whencrossed，are，withextremelyfewexceptions，quiteprolific，——insomecasesevenmoresothanthepurely—bredparent—races。Theoffspring，also，raisedfromsuchcrossesarelikewise，asweshallseeinthefollowingchapter，generallymorevigorousandfertilethantheirparents。Ontheotherhand，specieswhencrossed，andtheirhybridoffspring，arealmostinvariablyinsomedegreesterile;andherethereseemstoexistabroadandinsuperabledistinctionbetweenracesandspecies。Theimportanceofthissubjectasbearingontheoriginofspeciesisobvious;andweshallhereafterrecurtoit。
Itisunfortunatehowfewpreciseobservationshavebeenmadeonthefertilityofmongrelanimalsandplantsduringseveralsuccessivegenerations。Dr。Broca（16/1。’JournaldePhysiolog。’tome21859page385。）hasremarkedthatnoonehasobservedwhether，forinstance，mongreldogs，bredinterse，areindefinitelyfertile;yet，ifashadeofinfertilitybedetectedbycarefulobservationintheoffspringofnaturalformswhencrossed，itisthoughtthattheirspecificdistinctionisproved。Butsomanybreedsofsheep，cattle，pigs，dogs，andpoultry，havebeencrossedandrecrossedinvariousways，thatanysterility，ifithadexisted，wouldfrombeinginjuriousalmostcertainlyhavebeenobserved。Ininvestigatingthefertilityofcrossedvarietiesmanysourcesofdoubtoccur。Whenevertheleasttraceofsterilitybetweentwoplants，howevercloselyallied，wasobservedbyKolreuter，andmoreespeciallybyGartner，whocountedtheexactnumberofseedineachcapsule，thetwoformswereatoncerankedasdistinctspecies;andifthisrulebefollowed，assuredlyitwillneverbeprovedthatvarietieswhencrossedareinanydegreesterile。Wehaveformerlyseenthatcertainbreedsofdogsdonotreadilypairtogether;butnoobservationshavebeenmadewhether，whenpaired，theyproducethefullnumberofyoung，andwhetherthelatterareperfectlyfertileinterse;but，supposingthatsomedegreeofsterilitywerefoundtoexist，naturalistswouldsimplyinferthatthesebreedsweredescendedfromaboriginallydistinctspecies;anditwouldbescarcelypossibletoascertainwhetherornotthisexplanationwasthetrueone。
TheSebrightBantamismuchlessprolificthananyotherbreedoffowls，andisdescendedfromacrossbetweentwoverydistinctbreeds，recrossedbyathirdsub—variety。Butitwouldbeextremelyrashtoinferthatthelossoffertilitywasinanymannerconnectedwithitscrossedorigin，foritmaywithmoreprobabilitybeattributedeithertolong—continuedcloseinterbreeding，ortoaninnatetendencytosterilitycorrelatedwiththeabsenceofhacklesandsickletail—feathers。
Beforegivingthefewrecordedcasesofforms，whichmustberankedasvarieties，beinginsomedegreesterilewhencrossed，Imayremarkthatothercausessometimesinterferewithvarietiesfreelyintercrossing。Thustheymaydiffertoogreatlyinsize，aswithsomekindsofdogsandfowls:forinstance，theeditorofthe’JournalofHorticulture，etc。’（16/2。December1863page484。）saysthathecankeepBantamswiththelargerbreedswithoutmuchdangeroftheircrossing，butnotwiththesmallerbreeds，suchasGames，Hamburghs，etc。Withplantsadifferenceintheperiodoffloweringservestokeepvarietiesdistinct，aswiththevariouskindsofmaizeandwheat:thusColonelLeCouteur（16/3。On’TheVarietiesofWheat’page66。）remarks，"theTalaverawheat，fromfloweringmuchearlierthananyotherkind，issuretocontinuepure。"IndifferentpartsoftheFalklandIslandsthecattlearebreakingupintoherdsofdifferentcolours;andthoseonthehigherground，whicharegenerallywhite，usuallybreed，asIaminformedbySirJ。Sulivan，threemonthsearlierthanthoseonthelowland;andthiswouldmanifestlytendtokeeptheherdsfromblending。
Certaindomesticracesseemtopreferbreedingwiththeirownkind;andthisisafactofsomeimportance，foritisasteptowardsthatinstinctivefeelingwhichhelpstokeepcloselyalliedspeciesinastateofnaturedistinct。Wehavenowabundantevidencethat，ifitwerenotforthisfeeling，manymorehybridswouldbenaturallyproducedthaninthiscase。WehaveseeninthefirstchapterthatthealcodogofMexicodislikesdogsofotherbreeds;andthehairlessdogofParaguaymixeslessreadilywiththeEuropeanraces，thanthelatterdowitheachother。InGermanythefemaleSpitz—dogissaidtoreceivethefoxmorereadilythanwillotherdogs;afemaleAustralianDingoinEnglandattractedthewildmalefoxes。Butthesedifferencesinthesexualinstinctandattractivepowerofthevariousbreedsmaybewhollyduetotheirdescentfromdistinctspecies。InParaguaythehorseshavemuchfreedom，andanexcellentobserver（16/4。Rengger’SaugethierevonParaguay’
s。336。）believesthatthenativehorsesofthesamecolourandsizepreferassociatingwitheachother，andthatthehorseswhichhavebeenimportedfromEntreRiosandBandaOrientalintoParaguaylikewisepreferassociatingtogether。InCircassiasixsub—racesofthehorsehavereceiveddistinctnames;andanativeproprietorofrank（16/5。SeeamemoirbyMM。LherbetteandDeQuatrefagesin’Bull。Soc。d’Acclimat。’tome8July1861page312。）
assertsthathorsesofthreeoftheseraces，whilstlivingafreelife，almostalwaysrefusetomingleandcross，andwillevenattackoneanother。
Ithasbeenobserved，inadistrictstockedwithheavyLincolnshireandlightNorfolksheep，thatbothkinds;thoughbredtogether，whenturnedout，"inashorttimeseparatetoasheep;"theLincolnshiresdrawingofftotherichsoil，andtheNorfolkstotheirowndrylightsoil;andaslongasthereisplentyofgrass，"thetwobreedskeepthemselvesasdistinctasrooksandpigeons。"Inthiscasedifferenthabitsoflifetendtokeeptheracesdistinct。OnoneoftheFaroeislands，notmorethanhalfamileindiameter，thehalf—wildnativeblacksheeparesaidnottohavereadilymixedwiththeimportedwhitesheep。Itisamorecuriousfactthatthesemi—monstrousanconsheepofmodernorigin"havebeenobservedtokeeptogether，separatingthemselvesfromtherestoftheflock，whenputintoenclosureswithothersheep。"（16/6。FortheNorfolksheepseeMarshall’RuralEconomyofNorfolk’
volume2page136。SeeRev。L。Landt’DescriptionofFaroe’page66。Fortheanconsheepsee’Phil。Transact。’1813page90。）Withrespecttofallow—deer，whichliveinasemi—domesticatedcondition，Mr。Bennett（16/7。White’Nat。
Hist。ofSelbourne’editedbyBennettpage39。Withrespecttotheoriginofthedark—coloureddeersee’SomeAccountofEnglishDeerParks’byE。P。
Shirley，Esq。）statesthatthedarkandpalecolouredherds，whichhavelongbeenkepttogetherintheForestofDean，inHighMeadowWoods，andintheNewForest，haveneverbeenknowntomingle:thedark—coloureddeer，itmaybeadded，arebelievedtohavebeenfirstbroughtbyJamesI。fromNorway，onaccountoftheirgreaterhardiness。IimportedfromtheislandofPortoSantotwooftheferalrabbits，whichdiffer，asdescribedinthefourthchapter，fromcommonrabbits;bothprovedtobemales，and，thoughtheylivedduringsomeyearsintheZoologicalGardens，thesuperintendent，Mr。Bartlett，invainendeavouredtomakethembreedwithvarioustamekinds;butwhetherthisrefusaltobreedwasduetoanychangeintheinstinct，orsimplytotheirextremewildness，orwhetherconfinementhadrenderedthemsterile，asoftenoccurs，cannotbedetermined。
Whilstmatchingforthesakeofexperimentmanyofthemostdistinctbreedsofpigeons，itfrequentlyappearedtomethatthebirds，thoughfaithfultotheirmarriagevow，retainedsomedesireaftertheirownkind。AccordinglyIaskedMr。Wicking，whohaskeptalargerstockofvariousbreedstogetherthananymaninEngland，whetherhethoughtthattheywouldpreferpairingwiththeirownkind，supposingthatthereweremalesandfemalesenoughofeach;andhewithouthesitationansweredthathewasconvincedthatthiswasthecase。Ithasoftenbeennoticedthatthedovecotepigeonseemstohaveanactualaversiontowardstheseveralfancybreeds（16/8。’TheDovecote’bytheRev。
E。S。Dixonpage155;Bechstein’Naturgesch。Deutschlands’b。41795page17。）
yetallhavecertainlysprungfromacommonprogenitor。TheRev。W。D。FoxinformsmethathisflocksofwhiteandcommonChinesegeesekeptdistinct。
Thesefactsandstatements，thoughsomeofthemareincapableofproof，restingonlyontheopinionofexperiencedobservers，showthatsomedomesticracesareledbydifferenthabitsoflifetokeeptoacertainextentseparate，andthatothersprefercouplingwiththeirownkind，inthesamemannerasspeciesinastateofnature，thoughinamuchlessdegree。
[Withrespecttosterilityfromthecrossingofdomesticraces，Iknowofnowell—ascertainedcasewithanimals。Thisfact，seeingthegreatdifferenceinstructurebetweensomebreedsofpigeons，fowls，pigs，dogs，etc。，isextraordinary，incontrastwiththesterilityofmanycloselyalliednaturalspecieswhencrossed;butweshallhereafterattempttoshowthatitisnotsoextraordinaryasitatfirstappears。Anditmaybewellheretorecalltomindthattheamountofexternaldifferencebetweentwospeciesisnotasafeguideforpredictingwhetherornottheywillbreedtogether，——somecloselyalliedspecieswhencrossedbeingutterlysterile，andotherswhichareextremelyunlikebeingmoderatelyfertile。Ihavesaidthatnocaseofsterilityincrossedracesrestsonsatisfactoryevidence;buthereisonewhichatfirstseemstrustworthy。Mr。Youatt（16/9。’Cattle’page202。）andabetterauthoritycannotbequoted，states，thatformerlyinLancashirecrosseswerefrequentlymadebetweenlonghornandshorthorncattle;thefirstcrosswasexcellent，buttheproducewasuncertain;inthethirdorfourthgenerationthecowswerebadmilkers;"inadditiontowhich，therewasmuchuncertaintywhetherthecowswouldconceive;andfullone—thirdofthecowsamongsomeofthesehalf—bredsfailedtobeincalf。"Thisatfirstseemsagoodcase:butMr。Wilkinsonstates（16/10。Mr。J。Wilkinsonin’RemarksaddressedtoSirJ。Sebright’1820page38。），thatabreedderivedfromthissamecrosswasactuallyestablishedinanotherpartofEngland;andifithadfailedinfertility，thefactwouldsurelyhavebeennoticed。Moreover，supposingthatMr。Youatthadprovedhiscase，itmightbearguedthatthesterilitywaswhollyduetothetwoparent—breedsbeingdescendedfromprimordiallydistinctspecies。
InthecaseofplantsGartnerstatesthathefertilisedthirteenheads（andsubsequentlynineothers）onadwarfmaizebearingyellowseed（16/11。
’Bastarderzeugung’s。87，169。SeealsotheTableattheendofvolume。）withpollenofatallmaizehavingredseed;andoneheadaloneproducedgoodseed，butonlyfiveinnumber。Thoughtheseplantsaremonoecious，andthereforedonotrequirecastration，yetIshouldhavesuspectedsomeaccidentinthemanipulation，hadnotGartnerexpresslystatedthathehadduringmanyyearsgrownthesetwovarietiestogether，andtheydidnotspontaneouslycross;andthis，consideringthattheplantsaremonoeciousandaboundwithpollen，andarewellknowngenerallytocrossfreely，seemsexplicableonlyonthebeliefthatthesetwovarietiesareinsomedegreemutuallyinfertile。Thehybridplantsraisedfromtheabovefiveseedswereintermediateinstructure，extremelyvariable，andperfectlyfertile。（16/12。’Bastarderzeugung’s。87，577。）InlikemannerProf。Hildebrand（16/13。’Bot。Zeitung’1868page327。）
couldnotsucceedinfertilisingthefemaleflowersofaplantbearingbrowngrainswithpollenfromacertainkindbearingyellowgrains;althoughotherflowersonthesameplant，whichwerefertilisedwiththeirownpollen，yieldedgoodseed。Noone，Ibelieve，evensuspectsthatthesevarietiesofmaizearedistinctspecies;buthadthehybridsbeenintheleaststerile，nodoubtGartnerwouldatoncehavesoclassedthem。Imayhereremark，thatwithundoubtedspeciesthereisnotnecessarilyanycloserelationbetweenthesterilityofafirstcrossandthatofthehybridoffspring。Somespeciescanbecrossedwithfacility，butproduceutterlysterilehybrids;otherscanbecrossedwithextremedifficulty，butthehybridswhenproducedaremoderatelyfertile。Iamnotaware，however，ofanyinstancequitelikethisofthemaize，namely，ofafirstcrossmadewithdifficulty，butyieldingperfectlyfertilehybrids。（16/14。Mr。Shirreffformerlythought（’Gardener’sChronicle’
1858page771）thattheoffspringfromacrossbetweencertainvarietiesofwheatbecamesterileinthefourthgeneration;buthenowadmits（’ImprovementoftheCereals’1873）thatthiswasanerror。）
Thefollowingcaseismuchmoreremarkable，andevidentlyperplexedGartner，whosestrongwishitwastodrawabroadlineofdistinctionbetweenspeciesandvarieties。InthegenusVerbascum，hemade，duringeighteenyears，avastnumberofexperiments，andcrossednolessthan1085flowersandcountedtheirseeds。ManyoftheseexperimentsconsistedincrossingwhiteandyellowvarietiesofbothV。lychnitisandV。blattariawithnineotherspeciesandtheirhybrids。Thatthewhiteandyellowfloweredplantsofthesetwospeciesarereallyvarieties，noonehasdoubted;andGartneractuallyraisedinthecaseofbothspeciesonevarietyfromtheseedoftheother。Nowintwoofhisworks（16/15。’KenntnissderBefruchtung’s。137;’Bastarderzeugung’s。92，181。Onraisingthetwovarietiesfromseedsees。307。）hedistinctlyassertsthatcrossesbetweensimilarly—colouredflowersyieldmoreseedthanbetweendissimilarly—coloured;sothattheyellow—floweredvarietyofeitherspecies（andconverselywiththewhite—floweredvariety），whencrossedwithpollenofitsownkind，yieldsmoreseedthanwhencrossedwiththatofthewhitevariety;andsoitiswhendifferentlycolouredspeciesarecrossed。ThegeneralresultsmaybeseenintheTableattheendofhisvolume。Inoneinstancehegives（16/16。’Bastarderzeugung’s。216。）thefollowingdetails;
butImustpremisethatGartner，toavoidexaggeratingthedegreeofsterilityinhiscrosses，alwayscomparestheMAXIMUMnumberobtainedfromacrosswiththeAVERAGEnumbernaturallygivenbythepuremother—plant。ThewhitevarietyofV。lychnitis，naturallyfertilisedbyitsownpollen，gavefromanAVERAGE
oftwelvecapsulesninety—sixgoodseedsineach;whilsttwentyflowersfertilisedwithpollenfromtheyellowvarietyofthissamespecies，gaveastheMAXIMUMonlyeighty—ninegoodseeds;sothatwehavetheproportionof1000to908，accordingtoGartner’susualscale。Ishouldhavethoughtitpossiblethatsosmalladifferenceinfertilitymighthavebeenaccountedforbytheevileffectsofthenecessarycastration;butGartnershowsthatthewhitevarietyofV。lychnitis，whenfertilisedfirstbythewhitevarietyofV。blattaria，andthenbytheyellowvarietyofthisspecies，yieldedseedintheproportionof622to438;andinboththesecasescastrationwasperformed。Nowthesterilitywhichresultsfromthecrossingofthedifferentlycolouredvarietiesofthesamespecies，isfullyasgreatasthatwhichoccursinmanycaseswhendistinctspeciesarecrossed。UnfortunatelyGartnercomparedtheresultsofthefirstunionsalone，andnotthesterilityofthetwosetsofhybridsproducedfromthewhitevarietyofV。lychnitiswhenfertilisedbythewhiteandyellowvarietiesofV。blattaria，foritisprobablethattheywouldhavedifferedinthisrespect。
Mr。J。ScotthasgivenmetheresultsofaseriesofexperimentsonVerbascum，madebyhimintheBotanicGardensofEdinburgh。（16/17。Theresultshavesincebeenpublishedin’Journ。AsiaticSoc。ofBengal’1867page145。）HerepeatedsomeofGartner’sexperimentsondistinctspecies，butobtainedonlyfluctuatingresults，someconfirmatory，thegreaternumbercontradictory;
neverthelesstheseseemhardlysufficienttooverthrowtheconclusionarrivedatbyGartnerfromexperimentstriedonalargerscale。Mr。Scottalsoexperimentedontherelativefertilityofunionsbetweensimilarlyanddissimilarly—colouredvarietiesofthesamespecies。ThushefertilisedsixflowersoftheyellowvarietyofV。lychnitisbyitsownpollen，andobtainedsixcapsules;andcalling，forthesakeofcomparison，theaveragenumberofgoodseedineachoftheircapsulesonehundred，hefoundthatthissameyellowvariety，whenfertilisedbythewhitevariety，yieldedfromsevencapsulesanaverageofninety—fourseed。Onthesameprinciple，thewhitevarietyofV。lychnitisbyitsownpollen（fromsixcapsules），andbythepollenoftheyellowvariety（eightcapsules），yieldedseedintheproportionof100to82。TheyellowvarietyofV。thapsusbyitsownpollen（eightcapsules），andbythatofthewhitevariety（onlytwocapsules），yieldedseedintheproportionof100to94。Lastly，thewhitevarietyofV。blattariabyitsownpollen（eightcapsules），andbythatoftheyellowvariety（fivecapsules），yieldedseedintheproportionof100to79。Sothatineverycasetheunionsofsimilarly—colouredvarietiesofthesamespeciesweremorefertilethantheunionsofdissimilarly—colouredvarieties;whenallthecasesaregroupedtogether，thedifferenceoffertilityisas100to86。Someadditionaltrialsweremade，andaltogetherthirty—sixsimilarly—colouredunionsyieldedthirty—fivegoodcapsules;whilstthirty—fivedissimilarly—
colouredunionsyieldedonlytwenty—sixgoodcapsules。Besidestheforegoingexperiments，thepurpleV。phoeniceumwascrossedbyarose—colouredandawhitevarietyofthesamespecies;thesetwovarietieswerealsocrossedtogether，andtheseseveralunionsyieldedlessseedthanV。phoeniceumbyitsownpollen。HenceitfollowsfromMr。Scott’sexperiments，thatinthegenusVerbascumthesimilarlyanddissimilarly—colouredvarietiesofthesamespeciesbehave，whencrossed，likecloselyalliedbutdistinctspecies。
（16/18。Thefollowingfacts，givenbyKolreuterinhis’DritteFortsetzung’
ss。34，39，appearatfirstsightstronglytoconfirmMr。Scott’sandGartner’sstatements;andtoacertainlimitedextenttheydoso。Kolreuterasserts，frominnumerableobservations，thatinsectsincessantlycarrypollenfromonespeciesandvarietyofVerbascumtoanother;andIcanconfirmthisassertion;yethefoundthatthewhiteandyellowvarietiesofVerbascumlychnitisoftengrewwildmingledtogether:moreover，hecultivatedthesetwovarietiesinconsiderablenumbersduringfouryearsinhisgarden，andtheykepttruebyseed;butwhenhecrossedthem，theyproducedflowersofanintermediatetint。Henceitmighthavebeenthoughtthatbothvarietiesmusthaveastrongerelectiveaffinityforthepollenoftheirownvarietythanforthatoftheother;thiselectiveaffinity，Imayaddofeachspeciesforitsownpollen（Kolreuter’DritteForts。’s。39andGartner’Bastarderz。’passim）
beingaperfectlywell—ascertainedpower。ButtheforceoftheforegoingfactsismuchlessenedbyGartner’snumerousexperiments，for，differentlyfromKolreuter，heneveroncegot（’Bastarderz。’s。307）anintermediatetintwhenhecrossedtheyellowandwhitefloweredvarietiesofVerbascum。Sothatthefactofthewhiteandyellowvarietieskeepingtruetotheircolourbyseeddoesnotprovethattheywerenotmutuallyfertilisedbythepollencarriedbyinsectsfromonetotheother。）
Thisremarkablefactofthesexualaffinityofsimilarly—colouredvarieties，asobservedbyGartnerandMr。Scott，maynotbeofveryrareoccurrence;forthesubjecthasnotbeenattendedtobyothers。Thefollowingcaseisworthgiving，partlytoshowhowdifficultitistoavoiderror。Dr。Herbert（16/19。
’Amaryllidaceae’1837page366。Gartnerhasmadeasimilarobservation。）hasremarkedthatvariously—coloureddoublevarietiesoftheHollyhock（Althearosea）mayberaisedwithcertaintybyseedfromplantsgrowingclosetogether。Ihavebeeninformedthatnurserymenwhoraiseseedforsaledonotseparatetheirplants;accordinglyIprocuredseedofeighteennamedvarieties;ofthese，elevenvarietiesproducedsixty—twoplantsallperfectlytruetotheirkind;andsevenproducedforty—nineplants，halfofwhichweretrueandhalffalse。Mr。MastersofCanterburyhasgivenmeamorestrikingcase;hesavedseedfromagreatbedoftwenty—fournamedvarietiesplantedincloselyadjoiningrows，andeachvarietyreproduceditselftrulywithonlysometimesashadeofdifferenceintint。Nowinthehollyhockthepollen，whichisabundant，ismaturedandnearlyallshedbeforethestigmaofthesameflowerisreadytoreceiveit（16/20。Kolreuterfirstobservedthisfact，’Mem。del’Acad。deSt。Petersburg’volume3page127。SeealsoC。K。Sprengel’DasEntdeckteGeheimniss’s。345。）;andasbeescoveredwithpollenincessantlyflyfromplanttoplant，itwouldappearthatadjoiningvarietiescouldnotescapebeingcrossed。As，however，thisdoesnotoccur，itappearedtomeprobablethatthepollenofeachvarietywasprepotentonitsownstigmaoverthatofallothervarieties，butIhavenoevidenceonthispoint。Mr。C。
TurnerofSlough，wellknownforhissuccessinthecultivationofthisplant，informsmethatitisthedoublenessoftheflowerswhichpreventsthebeesgainingaccesstothepollenandstigma;andhefindsthatitisdifficulteventocrossthemartificially。Whetherthisexplanationwillfullyaccountforvarietiesincloseproximitypropagatingthemselvessotrulybyseed，Idonotknow。
Thefollowingcasesareworthgiving，astheyrelatetomonoeciousforms，whichdonotrequire，andconsequentlycannothavebeeninjuredby，castration。GiroudeBuzareinguescrossedwhathedesignatesthreevarietiesofgourd（16/21。NamelyBarbarines，Pastissons，Giraumous:’Annal。desSc。
Nat。’tome301833pages398and405。），andassertsthattheirmutualfertilisationislesseasyinproportiontothedifferencewhichtheypresent。
Iamawarehowimperfectlytheformsinthisgroupwereuntilrecentlyknown;
butSageret（16/22。’MemoiresurlesCucurbitaceae’1826pages46，55。），whorankedthemaccordingtotheirmutualfertility，considersthethreeformsabovealludedtoasvarieties，asdoesafarhigherauthority，namely，M。
Naudin。（16/23。’AnnalesdesSc。Nat。’4thseriestome6。M。NaudinconsiderstheseformsasundoubtedlyvarietiesofCucurbitapepo。）Sageret（16/24。’Mem。
Cucurb。’page8。）hasobservedthatcertainmelonshaveagreatertendency，whateverthecausemaybe，tokeeptruethanothers;andM。Naudin，whohashadsuchimmenseexperienceinthisgroup，informsmethathebelievesthatcertainvarietiesintercrossmorereadilythanothersofthesamespecies;buthehasnotprovedthetruthofthisconclusion;thefrequentabortionofthepollennearParisbeingonegreatdifficulty。Nevertheless，hehasgrownclosetogether，duringsevenyears，certainformsofCitrullus，which，astheycouldbeartificiallycrossedwithperfectfacilityandproducedfertileoffspring，arerankedasvarieties;buttheseformswhennotartificiallycrossedkepttrue。Manyothervarieties，ontheotherhand，inthesamegroupcrosswithsuchfacility，asM。Naudinrepeatedlyinsists，thatwithoutbeinggrownfaraparttheycannotbekeptintheleasttrue。