第7章

Ithinknobetterevidenceofthiscanberequiredthanthatthetwomostexperiencedobserverswhohaveeverlived，namely，Kö。lreuterandGä。rtner，shouldhavearrivedatdiametricallyoppositeconclusionsinregardtotheverysamespecies。ItisalsomostinstructivetocomparebutIhavenotspaceheretoenterondetailstheevidenceadvancedbyourbestbotanistsonthequestionwhethercertaindoubtfulformsshouldberankedasspeciesorvarieties，withtheevidencefromfertilityadducedbydifferenthybridisers，orbythesameauthor，fromexperimentsmadeduringdifferentyears。Itcanthusbeshownthatneithersterilitynorfertilityaffordsanycleardistinctionbetweenspeciesandvarieties。butthattheevidencefromthissourcegraduatesaway，andisdoubtfulinthesamedegreeasistheevidencederivedfromotherconstitutionalandstructuraldifferences。

Inregardtothesterilityofhybridsinsuccessivegenerations。thoughGä。rtnerwasenabledtorearsomehybrids，carefullyguardingthemfromacrosswitheitherpureparent，forsixorseven，andinonecasefortengenerations，yetheassertspositivelythattheirfertilityneverincreased，butgenerallygreatlydecreased。Idonotdoubtthatthisisusuallythecase，andthatthefertilityoftensuddenlydecreasesinthefirstfewgenerations。NeverthelessIbelievethatinalltheseexperimentsthefertilityhasbeendiminishedbyanindependentcause，namely，fromcloseinterbreeding。Ihavecollectedsolargeabodyoffacts，showingthatcloseinterbreedinglessensfertility，and，ontheotherhand，thatanoccasionalcrosswithadistinctindividualorvarietyincreasesfertility，thatIcannotdoubtthecorrectnessofthisalmostuniversalbeliefamongstbreeders。Hybridsareseldomraisedbyexperimentalistsingreatnumbers。

andastheparent-species，orotheralliedhybrids，generallygrowinthesamegarden，thevisitsofinsectsmustbecarefullypreventedduringthefloweringseason：hencehybridswillgenerallybefertilisedduringeachgenerationbytheirownindividualpollen。andIamconvincedthatthiswouldbeinjurioustotheirfertility，alreadylessenedbytheirhybridorigin。IamstrengthenedinthisconvictionbyaremarkablestatementrepeatedlymadebyGä。rtner，namely，thatifeventhelessfertilehybridsbeartificiallyfertilisedwithhybridpollenofthesamekind，theirfertility，notwithstandingthefrequentilleffectsofmanipulation，sometimesdecidedlyincreases，andgoesonincreasing。Now，inartificialfertilisationpollenisasoftentakenbychance(asIknowfrommyownexperience)fromtheanthersofanotherflower，asfromtheanthersofthefloweritselfwhichistobefertilised。sothatacrossbetweentwoflowers，thoughprobablyonthesameplant，wouldbethuseffected。Moreover，whenevercomplicatedexperimentsareinprogress，socarefulanobserverasGä。rtnerwouldhavecastratedhishybrids，andthiswouldhaveinsuredineachgenerationacrosswiththepollenfromadistinctflower，eitherfromthesameplantorfromanotherplantofthesamehybridnature。Andthus，thestrangefactoftheincreaseoffertilityinthesuccessivegenerationsofartificiallyfertilisedhybridsmay，Ibelieve，beaccountedforbycloseinterbreedinghavingbeenavoided。

Nowletusturntotheresultsarrivedatbythethirdmostexperiencedhybridiser，namely，theHon。andRev。W。Herbert。Heisasemphaticinhisconclusionthatsomehybridsareperfectlyfertileasfertileasthepureparent-speciesasareKö。lreuterandGä。rtnerthatsomedegreeofsterilitybetweendistinctspeciesisauniversallawofnature。HeexperimentisedonsomeoftheverysamespeciesasdidGä。rtner。Thedifferenceintheirresultsmay，Ithink，beinpartaccountedforbyHerbert’sgreathorticulturalskill，andbyhishavinghothousesathiscommand。

OfhismanyimportantstatementsIwillheregiveonlyasingleoneasanexample，namely，that’everyovuleinapodofCrinumcapensefertilisedbyC。revolutumproducedaplant，which(hesays)Ineversawtooccurinacaseofitsnaturalfecundation。’Sothatweherehaveperfect，orevenmorethancommonlyperfect，fertilityinafirstcrossbetweentwodistinctspecies。

ThiscaseoftheCrinumleadsmetorefertoamostsingularfact，namely，thatthereareindividualplants，aswithcertainspeciesofLobelia，andwithallthespeciesofthegenusHippeastrum，whichcanbefarmoreeasilyfertilisedbythepollenofanotheranddistinctspecies，thanbytheirownpollen。Fortheseplantshavebeenfoundtoyieldseedtothepollenofadistinctspecies，thoughquitesterilewiththeirownpollen，notwithstandingthattheirownpollenwasfoundtobeperfectlygood，foritfertiliseddistinctspecies。Sothatcertainindividualplantsandalltheindividualsofcertainspeciescanactuallybehybridisedmuchmorereadilythantheycanbeself-fertilised！Forinstance，abulbofHippeastrumaulicumproducedfourflowers。threewerefertilisedbyHerbertwiththeirownpollen，andthefourthwassubsequentlyfertilisedbythepollenofacompoundhybriddescendedfromthreeotheranddistinctspecies：theresultwasthat’theovariesofthethreefirstflowerssoonceasedtogrow，andafterafewdaysperishedentirely，whereasthepodimpregnatedbythepollenofthehybridmadevigorousgrowthandrapidprogresstomaturity，andboregoodseed，whichvegetatedfreely。’Inalettertome，in1839，MrHerberttoldmethathehadthentriedtheexperimentduringfiveyears，andhecontinuedtotryitduringseveralsubsequentyears，andalwayswiththesameresult。

Thisresulthas，also，beenconfirmedbyotherobserversinthecaseofHippeastrumwithitssub-genera，andinthecaseofsomeothergenera，asLobelia，PassifloraandVerbascum。Althoughtheplantsintheseexperimentsappearedperfectlyhealthy，andalthoughboththeovulesandpollenofthesameflowerwereperfectlygoodwithrespecttootherspecies，yetastheywerefunctionallyimperfectintheirmutualself-action，wemustinferthattheplantswereinanunnaturalstate。Neverthelessthesefactsshowonwhatslightandmysteriouscausesthelesserorgreaterfertilityofspecieswhencrossed，incomparisonwiththesamespecieswhenself-fertilised，sometimesdepends。

Thepracticalexperimentsofhorticulturists，thoughnotmadewithscientificprecision，deservesomenotice。ItisnotoriousinhowcomplicatedamannerthespeciesofPelargonium，Fuchsia，Calceolaria，Petunia，Rhododendron，&。c。，havebeencrossed，yetmanyofthesehybridsseedfreely。Forinstance，HerbertassertsthatahybridfromCalceolariaintegrifoliaandplantaginea，speciesmostwidelydissimilaringeneralhabit，’reproduceditselfasperfectlyasifithadbeenanaturalspeciesfromthemountainsofChile。’IhavetakensomepainstoascertainthedegreeoffertilityofsomeofthecomplexcrossesofRhododendrons，andIamassuredthatmanyofthemareperfectlyfertile。MrC。Noble，forinstance，informsmethatheraisesstocksforgraftingfromahybridbetweenRhod。PonticumandCatawbiense，andthatthishybrid’seedsasfreelyasitispossibletoimagine。’Hadhybrids，whenfairlytreated，goneondecreasinginfertilityineachsuccessivegeneration，asGä。rtnerbelievestobethecase，thefactwouldhavebeennotorioustonurserymen。Horticulturistsraiselargebedsofthesamehybrids，andsuchalonearefairlytreated，forbyinsectagencytheseveralindividualsofthesamehybridvarietyareallowedtofreelycrosswitheachother，andtheinjuriousinfluenceofcloseinterbreedingisthusprevented。Anyonemayreadilyconvincehimselfoftheefficiencyofinsect-agencybyexaminingtheflowersofthemoresterilekindsofhybridrhododendrons，whichproducenopollen，forhewillfindontheirstigmasplentyofpollenbroughtfromotherflowers。

Inregardtoanimals，muchfewerexperimentshavebeencarefullytriedthanwithplants。Ifoursystematicarrangementscanbetrusted，thatisifthegeneraofanimalsareasdistinctfromeachother，asarethegeneraofplants，thenwemayinferthatanimalsmorewidelyseparatedinthescaleofnaturecanbemoreeasilycrossedthaninthecaseofplants。

butthehybridsthemselvesare，Ithink，moresterile。Idoubtwhetheranycaseofaperfectlyfertilehybridanimalcanbeconsideredasthoroughlywellauthenticated。Itshould，however，beborneinmindthat，owingtofewanimalsbreedingfreelyunderconfinement，fewexperimentshavebeenfairlytried：forinstance，thecanary-birdhasbeencrossedwithnineotherfinches，butasnotoneoftheseninespeciesbreedsfreelyinconfinement，wehavenorighttoexpectthatthefirstcrossesbetweenthemandthecanary，orthattheirhybrids，shouldbeperfectlyfertile。Again，withrespecttothefertilityinsuccessivegenerationsofthemorefertilehybridanimals，Ihardlyknowofaninstanceinwhichtwofamiliesofthesamehybridhavebeenraisedatthesametimefromdifferentparents，soastoavoidtheilleffectsofcloseinterbreeding。Onthecontrary，brothersandsistershaveusuallybeencrossedineachsuccessivegeneration，inoppositiontotheconstantlyrepeatedadmonitionofeverybreeder。Andinthiscase，itisnotatallsurprisingthattheinherentsterilityinthehybridsshouldhavegoneonincreasing。Ifweweretoactthus，andpairbrothersandsistersinthecaseofanypureanimal，whichfromanycausehadtheleasttendencytosterility，thebreedwouldassuredlybelostinaveryfewgenerations。

AlthoughIdonotknowofanythoroughlywell-authenticatedcasesofperfectlyfertilehybridanimals，IhavesomereasontobelievethatthehybridsfromCervulusvaginalisandReevesii，andfromPhasianuscolchicuswithp。torquatusandwithp。versicolorareperfectlyfertile。ThehybridsfromthecommonandChinesegeese(A。cygnoides)，specieswhicharesodifferentthattheyaregenerallyrankedindistinctgenera，haveoftenbredinthiscountrywitheitherpureparent，andinonesingleinstancetheyhavebredinterse。ThiswaseffectedbyMrEyton，whoraisedtwohybridsfromthesameparentsbutfromdifferenthatches。andfromthesetwobirdsheraisednolessthaneighthybrids(grandchildrenofthepuregeese)fromonenest。InIndia，however，thesecross-bredgeesemustbefarmorefertile。forIamassuredbytwoeminentlycapablejudges，namelyMrBlythandCapt。Hutton，thatwholeflocksofthesecrossedgeesearekeptinvariouspartsofthecountry。andastheyarekeptforprofit，whereneitherpureparent-speciesexists，theymustcertainlybehighlyfertile。

AdoctrinewhichoriginatedwithPallas，hasbeenlargelyacceptedbymodernnaturalists。namely，thatmostofourdomesticanimalshavedescendedfromtwoormoreaboriginalspecies，sincecommingledbyintercrossing。

Onthisview，theaboriginalspeciesmusteitheratfirsthaveproducedquitefertilehybrids，orthehybridsmusthavebecomeinsubsequentgenerationsquitefertileunderdomestication。Thislatteralternativeseemstomethemostprobable，andIaminclinedtobelieveinitstruth，althoughitsrestsonnodirectevidence。Ibelieve，forinstance，thatourdogshavedescendedfromseveralwildstocks。yet，withperhapstheexceptionofcertainindigenousdomesticdogsofSouthAmerica，allarequitefertiletogether。andanalogymakesmegreatlydoubt，whethertheseveralaboriginalspecieswouldatfirsthavefreelybredtogetherandhaveproducedquitefertilehybrids。SoagainthereisreasontobelievethatourEuropeanandthehumpedIndiancattlearequitefertiletogether。butfromfactscommunicatedtomebyMrBlyth，Ithinktheymustbeconsideredasdistinctspecies。Onthisviewoftheoriginofmanyofourdomesticanimals，wemusteithergiveupthebeliefofthealmostuniversalsterilityofdistinctspeciesofanimalswhencrossed。orwemustlookatsterility，notasanindeliblecharacteristic，butasonecapableofbeingremovedbydomestication。

Finally，lookingtoalltheascertainedfactsontheintercrossingofplantsandanimals，itmaybeconcludedthatsomedegreeofsterility，bothinfirstcrossesandinhybrids，isanextremelygeneralresult。butthatitcannot，underourpresentstateofknowledge，beconsideredasabsolutelyuniversal。

LawsgoverningtheSterilityoffirstCrossesandofHybrids。Wewillnowconsideralittlemoreindetailthecircumstancesandrulesgoverningthesterilityoffirstcrossesandofhybrids。Ourchiefobjectwillbetoseewhetherornottherulesindicatethatspecieshavespeciallybeenendowedwiththisquality，inordertopreventtheircrossingandblendingtogetherinutterconfusion。ThefollowingrulesandconclusionsarechieflydrawnupfromGä。rtner’sadmirableworkonthehybridisationofplants。Ihavetakenmuchpainstoascertainhowfartherulesapplytoanimals，andconsideringhowscantyourknowledgeisinregardtohybridanimals，Ihavebeensurprisedtofindhowgenerallythesamerulesapplytobothkingdoms。

Ithasbeenalreadyremarked，thatthedegreeoffertility，bothoffirstcrossesandofhybrids，graduatesfromzerotoperfectfertility。

Itissurprisinginhowmanycuriouswaysthisgradationcanbeshowntoexist。butonlythebarestoutlineofthefactscanherebegiven。Whenpollenfromaplantofonefamilyisplacedonthestigmaofaplantofadistinctfamily，itexertsnomoreinfluencethansomuchinorganicdust。

Fromthisabsolutezerooffertility，thepollenofdifferentspeciesofthesamegenusappliedtothestigmaofsomeonespecies，yieldsaperfectgradationinthenumberofseedsproduced，uptonearlycompleteorevenquitecompletefertility。and，aswehaveseen，incertainabnormalcases，eventoanexcessoffertility，beyondthatwhichtheplant’sownpollenwillproduce。Soinhybridsthemselves，therearesomewhichneverhaveproduced，andprobablyneverwouldproduce，evenwiththepollenofeitherpureparent，asinglefertileseed：butinsomeofthesecasesafirsttraceoffertilitymaybedetected，bythepollenofoneofthepureparent-speciescausingtheflowerofthehybridtowitherearlierthanitotherwisewouldhavedone。andtheearlywitheringofthefloweriswellknowntobeasignofincipientfertilisation。Fromthisextremedegreeofsterilitywehaveself-fertilisedhybridsproducingagreaterandgreaternumberofseedsuptoperfectfertility。

Hybridsfromtwospecieswhichareverydifficulttocross，andwhichrarelyproduceanyoffspring，aregenerallyverysterile。buttheparallelismbetweenthedifficultyofmakingafirstcross，andthesterilityofthehybridsthusproducedtwoclassesoffactswhicharegenerallyconfoundedtogetherisbynomeansstrict。Therearemanycases，inwhichtwopurespeciescanbeunitedwithunusualfacility，andproducenumeroushybrid-offspring，yetthesehybridsareremarkablysterile。Ontheotherhand，therearespecieswhichcanbecrossedveryrarely，orwithextremedifficulty，butthehybrids，whenatlastproduced，areveryfertile。Evenwithinthelimitsofthesamegenus，forinstanceinDianthus，thesetwooppositecasesoccur。

Thefertility，bothoffirstcrossesandofhybrids，ismoreeasilyaffectedbyunfavourableconditions，thanisthefertilityofpurespecies。

Butthedegreeoffertilityislikewiseinnatelyvariable。foritisnotalwaysthesamewhenthesametwospeciesarecrossedunderthesamecircumstances，butdependsinpartupontheconstitutionoftheindividualswhichhappentohavebeenchosenfortheexperiment。Soitiswithhybrids，fortheirdegreeoffertilityisoftenfoundtodiffergreatlyintheseveralindividualsraisedfromseedoutofthesamecapsuleandexposedtoexactlythesameconditions。

Bythetermsystematicaffinityismeant，theresemblancebetweenspeciesinstructureandinconstitution，moreespeciallyinthestructureofpartswhichareofhighphysiologicalimportanceandwhichdifferlittleinthealliedspecies。Nowthefertilityoffirstcrossesbetweenspecies，andofthehybridsproducedfromthem，islargelygovernedbytheirsystematicaffinity。Thisisclearlyshownbyhybridsneverhavingbeenraisedbetweenspeciesrankedbysystematistsindistinctfamilies。andontheotherhand，byverycloselyalliedspeciesgenerallyunitingwithfacility。Butthecorrespondencebetweensystematicaffinityandthefacilityofcrossingisbynomeansstrict。Amultitudeofcasescouldbegivenofverycloselyalliedspecieswhichwillnotunite，oronlywithextremedifficulty。andontheotherhandofverydistinctspecieswhichunitewiththeutmostfacility。Inthesamefamilytheremaybeagenus，asDianthus，inwhichverymanyspeciescanmostreadilybecrossed。andanothergenus，asSilene，inwhichthemostperseveringeffortshavefailedtoproducebetweenextremelyclosespeciesasinglehybrid。Evenwithinthelimitsofthesamegenus，wemeetwiththissamedifference。forinstance，themanyspeciesofNicotianahavebeenmorelargelycrossedthanthespeciesofalmostanyothergenus。

butGä。rtnerfoundthatN。acuminata，whichisnotaparticularlydistinctspecies，obstinatelyfailedtofertilise，ortobefertilisedby，nolessthaneightotherspeciesofNicotiana。Verymanyanalogousfactscouldbegiven。

Noonehasbeenabletopointoutwhatkind，orwhatamount，ofdifferenceinanyrecognisablecharacterissufficienttopreventtwospeciescrossing。

Itcanbeshownthatplantsmostwidelydifferentinhabitandgeneralappearance，andhavingstronglymarkeddifferencesineverypartoftheflower，eveninthepollen，inthefruit，andinthecotyledons，canbecrossed。Annualandperennialplants，deciduousandevergreentrees，plantsinhabitingdifferentstationsandfittedforextremelydifferentclimates，canoftenbecrossedwithease。

Byareciprocalcrossbetweentwospecies，Imeanthecase，forinstance，ofastallion-horsebeingfirstcrossedwithafemale-ass，andthenamale-asswithamare：thesetwospeciesmaythenbesaidtohavebeenreciprocallycrossed。Thereisoftenthewidestpossibledifferenceinthefacilityofmakingreciprocalcrosses。Suchcasesarehighlyimportant，fortheyprovethatthecapacityinanytwospeciestocrossisoftencompletelyindependentoftheirsystematicaffinity，orofanyrecognisabledifferenceintheirwholeorganisation。Ontheotherhand，thesecasesclearlyshowthatthecapacityforcrossingisconnectedwithconstitutionaldifferencesimperceptiblebyus，andconfinedtothereproductivesystem。ThisdifferenceintheresultofreciprocalcrossesbetweenthesametwospecieswaslongagoobservedbyKö。lreuter。Togiveaninstance：MirabilisjalappacaneasilybefertilisedbythepollenofM。longiflora，andthehybridsthusproducedaresufficientlyfertile。butKö。lreutertriedmorethantwohundredtimes，duringeightfollowingyears，tofertilisereciprocallyM。longiflorawiththepollenofM。jalappa，andutterlyfailed。Severalotherequallystrikingcasescouldbegiven。Thurethasobservedthesamefactwithcertainsea-weedsorFuci。Gä。rtner，moreover，foundthatthisdifferenceoffacilityinmakingreciprocalcrossesisextremelycommoninalesserdegree。Hehasobserveditevenbetweenformssocloselyrelated(asMatthiolaannuaandglabra)thatmanybotanistsrankthemonlyasvarieties。

Itisalsoaremarkablefact，thathybridsraisedfromreciprocalcrosses，thoughofcoursecompoundedoftheverysametwospecies，theonespecieshavingfirstbeenusedasthefatherandthenasthemother，generallydifferinfertilityinasmall，andoccasionallyinahighdegree。

SeveralothersingularrulescouldbegivenfromGä。rtner：forinstance，somespecieshavearemarkablepowerofcrossingwithotherspecies。otherspeciesofthesamegenushavearemarkablepowerofimpressingtheirlikenessontheirhybridoffspring。butthesetwopowersdonotatallnecessarilygotogether。Therearecertainhybridswhichinsteadofhaving，asisusual，anintermediatecharacterbetweentheirtwoparents，alwayscloselyresembleoneofthem。andsuchhybrids，thoughexternallysolikeoneoftheirpureparent-species，arewithrareexceptionsextremelysterile。Soagainamongsthybridswhichareusuallyintermediateinstructurebetweentheirparents，exceptionalandabnormalindividualssometimesareborn，whichcloselyresembleoneoftheirpureparents。andthesehybridsarealmostalwaysutterlysterile，evenwhentheotherhybridsraisedfromseedfromthesamecapsulehaveaconsiderabledegreeoffertility。Thesefactsshowhowcompletelyfertilityinthehybridisindependentofitsexternalresemblancetoeitherpureparent。

Consideringtheseveralrulesnowgiven，whichgovernthefertilityoffirstcrossesandofhybrids，weseethatwhenforms，whichmustbeconsideredasgoodanddistinctspecies，areunited，theirfertilitygraduatesfromzerotoperfectfertility，oreventofertilityundercertainconditionsinexcess。Thattheirfertility，besidesbeingeminentlysusceptibletofavourableandunfavourableconditions，isinnatelyvariable。Thatitisbynomeansalwaysthesameindegreeinthefirstcrossandinthehybridsproducedfromthiscross。Thatthefertilityofhybridsisnotrelatedtothedegreeinwhichtheyresembleinexternalappearanceeitherparent。

Andlastly，thatthefacilityofmakingafirstcrossbetweenanytwospeciesisnotalwaysgovernedbytheirsystematicaffinityordegreeofresemblancetoeachother。Thislatterstatementisclearlyprovedbyreciprocalcrossesbetweenthesametwospecies，foraccordingastheonespeciesortheotherisusedasthefatherorthemother，thereisgenerallysomedifference，andoccasionallythewidestpossibledifference，inthefacilityofeffectinganunion。Thehybrids，moreover，producedfromreciprocalcrossesoftendifferinfertility。

Nowdothesecomplexandsingularrulesindicatethatspecieshavebeenendowedwithsterilitysimplytopreventtheirbecomingconfoundedinnature？

Ithinknot。Forwhyshouldthesterilitybesoextremelydifferentindegree，whenvariousspeciesarecrossed，allofwhichwemustsupposeitwouldbeequallyimportanttokeepfromblendingtogether？Whyshouldthedegreeofsterilitybeinnatelyvariableintheindividualsofthesamespecies？Whyshouldsomespeciescrosswithfacility，andyetproduceverysterilehybrids。andotherspeciescrosswithextremedifficulty，andyetproducefairlyfertilehybrids？Whyshouldthereoftenbesogreatadifferenceintheresultofareciprocalcrossbetweenthesametwospecies？

Why，itmayevenbeasked，hastheproductionofhybridsbeenpermitted？

Togranttospeciesthespecialpowerofproducinghybrids，andthentostoptheirfurtherpropagationbydifferentdegreesofsterility，notstrictlyrelatedtothefacilityofthefirstunionbetweentheirparents，seemstobeastrangearrangement。

Theforegoingrulesandfacts，ontheotherhand，appeartomeclearlytoindicatethatthesterilitybothoffirstcrossesandofhybridsissimplyincidentalordependentonunknowndifferences，chieflyinthereproductivesystems，ofthespecieswhicharecrossed。Thedifferencesbeingofsopeculiarandlimitedanature，that，inreciprocalcrossesbetweentwospeciesthemalesexualelementoftheonewilloftenfreelyactonthefemalesexualelementoftheother，butnotinareverseddirection。ItwillbeadvisabletoexplainalittlemorefullybyanexamplewhatImeanbysterilitybeingincidentalonotherdifferences，andnotaspeciallyendowedquality。Asthecapacityofoneplanttobegraftedorbuddedonanotherissoentirelyunimportantforitswelfareinastateofnature，Ipresumethatnoonewillsupposethatthiscapacityisaspeciallyendowedquality，butwilladmitthatitisincidentalondifferencesinthelawsofgrowthofthetwoplants。Wecansometimesseethereasonwhyonetreewillnottakeonanother，fromdifferencesintheirrateofgrowth，inthehardnessoftheirwood，intheperiodoftheflowornatureoftheirsap，&。c……butinamultitudeofcaseswecanassignnoreasonwhatever。

Greatdiversityinthesizeoftwoplants，onebeingwoodyandtheotherherbaceous，onebeingevergreenandtheotherdeciduous，andadaptationtowidelydifferentclimates，doesnotalwayspreventthetwograftingtogether。Asinhybridisation，sowithgrafting，thecapacityislimitedbysystematicaffinity，fornoonehasbeenabletografttreestogetherbelongingtoquitedistinctfamilies。and，ontheotherhand，closelyalliedspecies，andvarietiesofthesamespecies，canusually，butnotinvariably，begraftedwithease。Butthiscapacity，asinhybridisation，isbynomeansabsolutelygovernedbysystematicaffinity。Althoughmanydistinctgenerawithinthesamefamilyhavebeengraftedtogether，inothercasesspeciesofthesamegenuswillnottakeoneachother。Thepearcanbegraftedfarmorereadilyonthequince，whichisrankedasadistinctgenus，thanontheapple，whichisamemberofthesamegenus。Evendifferentvarietiesofthepeartakewithdifferentdegreesoffacilityonthequince。

sododifferentvarietiesoftheapricotandpeachoncertainvarietiesoftheplum。

AsGä。rtnerfoundthattherewassometimesaninnatedifferenceindifferentindividualsofthesametwospeciesincrossing。soSagaretbelievesthistobethecasewithdifferentindividualsofthesametwospeciesinbeinggraftedtogether。Asinreciprocalcrosses，thefacilityofeffectinganunionisoftenveryfarfromequal，soitsometimesisingrafting。thecommongooseberry，forinstance，cannotbegraftedonthecurrant，whereasthecurrantwilltake，thoughwithdifficulty，onthegooseberry。

Wehaveseenthatthesterilityofhybrids，whichhavetheirreproductiveorgansinanimperfectcondition，isaverydifferentcasefromthedifficultyofunitingtwopurespecies，whichhavetheirreproductiveorgansperfect。

yetthesetwodistinctcasesruntoacertainextentparallel。Somethinganalogousoccursingrafting。forThouinfoundthatthreespeciesofRobinia，whichseededfreelyontheirownroots，andwhichcouldbegraftedwithnogreatdifficultyonanotherspecies，whenthusgraftedwererenderedbarren。Ontheotherhand，certainspeciesofSorbus，whengraftedonotherspecies，yieldedtwiceasmuchfruitaswhenontheirownroots。WeareremindedbythislatterfactoftheextraordinarycaseofHippeastrum，Lobelia，&。c。，whichseededmuchmorefreelywhenfertilisedwiththepollenofdistinctspecies，thanwhenself-fertilisedwiththeirownpollen。

Wethussee，thatalthoughthereisaclearandfundamentaldifferencebetweenthemereadhesionofgraftedstocks，andtheunionofthemaleandfemaleelementsintheactofreproduction，yetthatthereisarudedegreeofparallelismintheresultsofgraftingandofcrossingdistinctspecies。Andaswemustlookatthecuriousandcomplexlawsgoverningthefacilitywithwhichtreescanbegraftedoneachotherasincidentalonunknowndifferencesintheirvegetativesystems，soIbelievethatthestillmorecomplexlawsgoverningthefacilityoffirstcrosses，areincidentalonunknowndifferences，chieflyintheirreproductivesystems。Thesedifferences，inbothcases，followtoacertainextent，asmighthavebeenexpected，systematicaffinity，bywhicheverykindofresemblanceanddissimilaritybetweenorganicbeingsisattemptedtobeexpressed。Thefactsbynomeansseemtometoindicatethatthegreaterorlesserdifficultyofeithergraftingorcrossingtogethervariousspecieshasbeenaspecialendowment。

althoughinthecaseofcrossing，thedifficultyisasimportantfortheenduranceandstabilityofspecificforms，asinthecaseofgraftingitisunimportantfortheirwelfare。

CausesoftheSterilityoffirstCrossesandofHybrids。Wemaynowlookalittlecloserattheprobablecausesofthesterilityoffirstcrossesandofhybrids。Thesetwocasesarefundamentallydifferent，for，asjustremarked，intheunionoftwopurespeciesthemaleandfemalesexualelementsareperfect，whereasinhybridstheyareimperfect。Eveninfirstcrosses，thegreaterorlesserdifficultyineffectingaunionapparentlydependsonseveraldistinctcauses。Theremustsometimesbeaphysicalimpossibilityinthemaleelementreachingtheovule，aswouldbethecasewithaplanthavingapistiltoolongforthepollen-tubestoreachtheovarium。Ithasalsobeenobservedthatwhenpollenofonespeciesisplacedonthestigmaofadistantlyalliedspecies，thoughthepollen-tubesprotrude，theydonotpenetratethestigmaticsurface。Again，themaleelementmayreachthefemaleelement，butbeincapableofcausinganembryotobedeveloped，asseemstohavebeenthecasewithsomeofThuret’sexperimentsonFuci。Noexplanationcanbegivenofthesefacts，anymorethanwhycertaintreescannotbegraftedonothers。Lastly，anembryomaybedeveloped，andthenperishatanearlyperiod。Thislatteralternativehasnotbeensufficientlyattendedto。butIbelieve，fromobservationscommunicatedtomebyMr。Hewitt，whohashadgreatexperienceinhybridisinggallinaceousbirds，thattheearlydeathoftheembryoisaveryfrequentcauseofsterilityinfirstcrosses。Iwasatfirstveryunwillingtobelieveinthisview。ashybrids，whenonceborn，aregenerallyhealthyandlong-lived，asweseeinthecaseofthecommonmule。Hybrids，however，aredifferentlycircumstancedbeforeandafterbirth：whenbornandlivinginacountrywheretheirtwoparentscanlive，theyaregenerallyplacedundersuitableconditionsoflife。Butahybridpartakesofonlyhalfofthenatureandconstitutionofitsmother，andthereforebeforebirth，aslongasitisnourishedwithinitsmother’swomborwithintheeggorseedproducedbythemother，itmaybeexposedtoconditionsinsomedegreeunsuitable，andconsequentlybeliabletoperishatanearlyperiod。moreespeciallyasallveryyoungbeingsseememinentlysensitivetoinjuriousorunnaturalconditionsoflife。

Inregardtothesterilityofhybrids，inwhichthesexualelementsareimperfectlydeveloped，thecaseisverydifferent。Ihavemorethanoncealludedtoalargebodyoffacts，whichIhavecollected，showingthatwhenanimalsandplantsareremovedfromtheirnaturalconditions，theyareextremelyliabletohavetheirreproductivesystemsseriouslyaffected。This，infact，isthegreatbartothedomesticationofanimals。

Betweenthesterilitythussuperinducedandthatofhybrids，therearemanypointsofsimilarity。Inbothcasesthesterilityisindependentofgeneralhealth，andisoftenaccompaniedbyexcessofsizeorgreatluxuriance。

Inbothcases，thesterilityoccursinvariousdegrees。inboth，themaleelementisthemostliabletobeaffected。butsometimesthefemalemorethanthemale。Inboth，thetendencygoestoacertainextentwithsystematicaffinity，orwholegroupsofanimalsandplantsarerenderedimpotentbythesameunnaturalconditions。andwholegroupsofspeciestendtoproducesterilehybrids。Ontheotherhand，onespeciesinagroupwillsometimesresistgreatchangesofconditionswithunimpairedfertility。andcertainspeciesinagroupwillproduceunusuallyfertilehybrids。Noonecantell，tillhetries，whetheranyparticularanimalwillbreedunderconfinementoranyplantseedfreelyunderculture。norcanhetell，tillhetries，whetheranytwospeciesofagenuswillproducemoreorlesssterilehybrids。

Lastly，whenorganicbeingsareplacedduringseveralgenerationsunderconditionsnotnaturaltothem，theyareextremelyliabletovary，whichisdue，asIbelieve，totheirreproductivesystemshavingbeenspeciallyaffected，thoughinalesserdegreethanwhensterilityensues。Soitiswithhybrids，forhybridsinsuccessivegenerationsareeminentlyliabletovary，aseveryexperimentalisthasobserved。

Thusweseethatwhenorganicbeingsareplacedundernewandunnaturalconditions，andwhenhybridsareproducedbytheunnaturalcrossingoftwospecies，thereproductivesystem，independentlyofthegeneralstateofhealth，isaffectedbysterilityinaverysimilarmanner。Intheonecase，theconditionsoflifehavebeendisturbed，thoughofteninsoslightadegreeastobeinappreciablebyus。intheothercase，orthatofhybrids，theexternalconditionshaveremainedthesame，buttheorganisationhasbeendisturbedbytwodifferentstructuresandconstitutionshavingbeenblendedintoone。Foritisscarcelypossiblethattwoorganisationsshouldbecompoundedintoone，withoutsomedisturbanceoccurringinthedevelopment，orperiodicalaction，ormutualrelationofthedifferentpartsandorgansonetoanother，ortotheconditionsoflife。Whenhybridsareabletobreedinterse，theytransmittotheiroffspringfromgenerationtogenerationthesamecompoundedorganisation，andhenceweneednotbesurprisedthattheirsterility，thoughinsomedegreevariable，rarelydiminishes。

Itmust，however，beconfessedthatwecannotunderstand，exceptingonvaguehypotheses，severalfactswithrespecttothesterilityofhybrids。

forinstance，theunequalfertilityofhybridsproducedfromreciprocalcrosses。ortheincreasedsterilityinthosehybridswhichoccasionallyandexceptionallyresemblecloselyeitherpureparent。NordoIpretendthattheforegoingremarksgototherootofthematter：noexplanationisofferedwhyanorganism，whenplacedunderunnaturalconditions，isrenderedsterile。AllthatIhaveattemptedtoshow，isthatintwocases，insomerespectsallied，sterilityisthecommonresult，intheonecasefromtheconditionsoflifehavingbeendisturbed，intheothercasefromtheorganisationhavingbeendisturbedbytwoorganisationshavingbeencompoundedintoone。

Itmayseemfanciful，butIsuspectthatasimilarparallelismextendstoanalliedyetverydifferentclassoffacts。Itisanoldandalmostuniversalbelief，founded，Ithink，onaconsiderablebodyofevidence，thatslightchangesintheconditionsoflifearebeneficialtoalllivingthings。Weseethisactedonbyfarmersandgardenersintheirfrequentexchangesofseed，tubers，&。c。，fromonesoilorclimatetoanother，andbackagain。Duringtheconvalescenceofanimals，weplainlyseethatgreatbenefitisderivedfromalmostanychangeinthehabitsoflife。

Again，bothwithplantsandanimals，thereisabundantevidence，thatacrossbetweenverydistinctindividualsofthesamespecies，thatisbetweenmembersofdifferentstrainsorsub-breeds，givesvigourandfertilitytotheoffspring。Ibelieve，indeed，fromthefactsalludedtoinourfourthchapter，thatacertainamountofcrossingisindispensableevenwithhermaphrodites。

andthatcloseinterbreedingcontinuedduringseveralgenerationsbetweenthenearestrelations，especiallyifthesebekeptunderthesameconditionsoflife，alwaysinducesweaknessandsterilityintheprogeny。

Henceitseemsthat，ontheonehand，slightchangesintheconditionsoflifebenefitallorganicbeings，andontheotherhand，thatslightcrosses，thatiscrossesbetweenthemalesandfemalesofthesamespecieswhichhavevariedandbecomeslightlydifferent，givevigourandfertilitytotheoffspring。Butwehaveseenthatgreaterchanges，orchangesofaparticularnature，oftenrenderorganicbeingsinsomedegreesterile。

andthatgreatercrosses，thatiscrossesbetweenmalesandfemaleswhichhavebecomewidelyorspecificallydifferent，producehybridswhicharegenerallysterileinsomedegree。Icannotpersuademyselfthatthisparallelismisanaccidentoranillusion。Bothseriesoffactsseemtobeconnectedtogetherbysomecommonbutunknownbond，whichisessentiallyrelatedtotheprincipleoflife。

FertilityofVarietieswhencrossed，andoftheirMongreloff-spring。Itmaybeurged，asamostforcibleargument，thattheremustbesomeessentialdistinctionbetweenspeciesandvarieties，andthattheremustbesomeerrorinalltheforegoingremarks，inasmuchasvarieties，howevermuchtheymaydifferfromeachotherinexternalappearance，crosswithperfectfacility，andyieldperfectlyfertileoffspring。Ifullyadmitthatthisisalmostinvariablythecase。Butifwelooktovarietiesproducedundernature，weareimmediatelyinvolvedinhopelessdifficulties。foriftwohithertoreputedvarietiesbefoundinanydegreesteriletogether，theyareatoncerankedbymostnaturalistsasspecies。Forinstance，theblueandredpimpernel，theprimroseandcowslip，whichareconsideredbymanyofourbestbotanistsasvarieties，aresaidbyGä。rtnernottobequitefertilewhencrossed，andheconsequentlyranksthemasundoubtedspecies。Ifwethusargueinacircle，thefertilityofallvarietiesproducedundernaturewillassuredlyhavetobegranted。

Ifweturntovarieties，produced，orsupposedtohavebeenproduced，underdomestication，wearestillinvolvedindoubt。Forwhenitisstated，forinstance，thattheGermanSpitzdogunitesmoreeasilythanotherdogswithfoxes，orthatcertainSouthAmericanindigenousdomesticdogsdonotreadilycrosswithEuropeandogs，theexplanationwhichwilloccurtoeveryone，andprobablythetrueone，isthatthesedogshavedescendedfromseveralaboriginallydistinctspecies。Neverthelesstheperfectfertilityofsomanydomesticvarieties，differingwidelyfromeachotherinappearance，forinstanceofthepigeonorofthecabbage，isaremarkablefact。moreespeciallywhenwereflecthowmanyspeciesthereare，which，thoughresemblingeachothermostclosely，areutterlysterilewhenintercrossed。Severalconsiderations，however，renderthefertilityofdomesticvarietieslessremarkablethanatfirstappears。Itcan，inthefirstplace，beclearlyshownthatmereexternaldissimilaritybetweentwospeciesdoesnotdeterminetheirgreaterorlesserdegreeofsterilitywhencrossed。andwemayapplythesameruletodomesticvarieties。Inthesecondplace，someeminentnaturalistsbelievethatalongcourseofdomesticationtendstoeliminatesterilityinthesuccessivegenerationsofhybrids，whichwereatfirstonlyslightlysterile。andifthisbeso，wesurelyoughtnottoexpecttofindsterilitybothappearinganddisappearingundernearlythesameconditionsoflife。Lastly，andthisseemstomebyfarthemostimportantconsideration，newracesofanimalsandplantsareproducedunderdomesticationbyman’smethodicalandunconsciouspowerofselection，forhisownuseandpleasure：heneitherwishestoselect，norcouldselect，slightdifferencesinthereproductivesystem，orotherconstitutionaldifferencecorrelatedwiththereproductivesystem。Hesupplieshisseveralvarietieswiththesamefood。treatstheminnearlythesamemanner，anddoesnotwishtoaltertheirgeneralhabitsoflife。Natureactsuniformlyandslowlyduringvastperiodsoftimeonthewholeorganization，inanywaywhichmaybeforeachcreature’sowngood。andthusshemay，eitherdirectly，ormoreprobablyindirectly，throughcorrelation，modifythereproductivesystemintheseveraldescendantsfromanyonespecies。Seeingthisdifferenceintheprocessofselection，ascarriedonbymanandnature，weneednotbesurprisedatsomedifferenceintheresult。

Ihaveasyetspokenasifthevarietiesofthesamespecieswereinvariablyfertilewhenintercrossed。Butitseemstomeimpossibletoresisttheevidenceoftheexistenceofacertainamountofsterilityinthefewfollowingcases，whichIwillbrieflyabstract。Theevidenceisatleastasgoodasthatfromwhichwebelieveinthesterilityofamultitudeofspecies。

Theevidenceis，also，derivedfromhostilewitnesses，whoinallothercasesconsiderfertilityandsterilityassafecriterionsofspecificdistinction。

Gä。rtnerkeptduringseveralyearsadwarfkindofmaizewithyellowseeds，andatallvarietywithredseeds，growingneareachotherinhisgarden。andalthoughtheseplantshaveseparatedsexes，theynevernaturallycrossed。Hethenfertilizedthirteenflowersoftheonewiththepollenoftheother。butonlyasingleheadproducedanyseed，andthisoneheadproducedonlyfivegrains。Manipulationinthiscasecouldnothavebeeninjurious，astheplantshaveseparatedsexes。Noone，Ibelieve，hassuspectedthatthesevarietiesofmaizearedistinctspecies。anditisimportanttonoticethatthehybridplantsthusraisedwerethemselvesperfectlyfertile。sothatevenGä。rtnerdidnotventuretoconsiderthetwovarietiesasspecificallydistinct。

GiroudeBuzareinguescrossedthreevarietiesofgourd，whichlikethemaizehasseparatedsexes，andheassertsthattheirmutualfertilizationisbysomuchthelesseasyastheirdifferencesaregreater。Howfartheseexperimentsmaybetrusted，Iknownot。buttheformsexperimentisedon，arerankedbySagaret，whomainlyfoundshisclassificationbythetestofinfertility，asvarieties。

Thefollowingcaseisfarmoreremarkable，andseemsatfirstquiteincredible。butitistheresultofanastonishingnumberofexperimentsmadeduringmanyyearsonninespeciesofVerbascum，bysogoodanobserverandsohostileawitness，asGä。rtner：namely，thatyellowandwhitevarietiesofthesamespeciesofVerbascumwhenintercrossedproducelessseed，thandoeithercolouredvarietieswhenfertilizedwithpollenfromtheirowncolouredflowers。Moreover，heassertsthatwhenyellowandwhitevarietiesofonespeciesarecrossedwithyellowandwhitevarietiesofadistinctspecies，moreseedisproducedbythecrossesbetweenthesamecolouredflowers，thanbetweenthosewhicharedifferentlycoloured。

YetthesevarietiesofVerbascumpresentnootherdifferencebesidesthemerecolouroftheflower。andonevarietycansometimesberaisedfromtheseedoftheother。

FromobservationswhichIhavemadeoncertainvarietiesofhollyhock，Iaminclinedtosuspectthattheypresentanalogousfacts。

Kö。lreuter，whoseaccuracyhasbeenconfirmedbyeverysubsequentobserver，hasprovedtheremarkablefact，thatonevarietyofthecommontobaccoismorefertile，whencrossedwithawidelydistinctspecies，thanaretheothervarieties。Heexperimentisedonfiveforms，whicharecommonlyreputedtobevarieties，andwhichhetestedbytheseveresttrial，namely，byreciprocalcrosses，andhefoundtheirmongreloffspringperfectlyfertile。

Butoneofthesefivevarieties，whenusedeitherasfatherormother，andcrossedwiththeNicotianaglutinosa，alwaysyieldedhybridsnotsosterileasthosewhichwereproducedfromthefourothervarietieswhencrossedwithN。glutinosa。Hencethereproductivesystemofthisonevarietymusthavebeeninsomemannerandinsomedegreemodified。

Fromthesefacts。fromthegreatdifficultyofascertainingtheinfertilityofvarietiesinastateofnature，forasupposedvarietyifinfertileinanydegreewouldgenerallyberankedasspecies。frommanselectingonlyexternalcharactersintheproductionofthemostdistinctdomesticvarieties，andfromnotwishingorbeingabletoproducereconditeandfunctionaldifferencesinthereproductivesystem。fromtheseseveralconsiderationsandfacts，Idonotthinkthattheverygeneralfertilityofvarietiescanbeprovedtobeofuniversaloccurrence，ortoformafundamentaldistinctionbetweenvarietiesandspecies。ThegeneralfertilityofvarietiesdoesnotseemtomesufficienttooverthrowtheviewwhichIhavetakenwithrespecttotheverygeneral，butnotinvariable，sterilityoffirstcrossesandofhybrids，namely，thatitisnotaspecialendowment，butisincidentalonslowlyacquiredmodifications，moreespeciallyinthereproductivesystemsoftheformswhicharecrossed。

HybridsandMongrelscompared，independentlyoftheirfertility。Independentlyofthequestionoffertility，theoffspringofspecieswhencrossedandofvarietieswhencrossedmaybecomparedinseveralotherrespects。Gä。rtner，whosestrongwishwastodrawamarkedlineofdistinctionbetweenspeciesandvarieties，couldfindveryfewand，asitseemstome，quiteunimportantdifferencesbetweentheso-calledhybridoffspringofspecies，andtheso-calledmongreloffspringofvarieties。

And，ontheotherhand，theyagreemostcloselyinverymanyimportantrespects。

Ishallherediscussthissubjectwithextremebrevity。Themostimportantdistinctionis，thatinthefirstgenerationmongrelsaremorevariablethanhybrids。butGä。rtneradmitsthathybridsfromspecieswhichhavelongbeencultivatedareoftenvariableinthefirstgeneration。andI

havemyselfseenstrikinginstancesofthisfact。Gä。rtnerfurtheradmitsthathybridsbetweenverycloselyalliedspeciesaremorevariablethanthosefromverydistinctspecies。andthisshowsthatthedifferenceinthedegreeofvariabilitygraduatesaway。Whenmongrelsandthemorefertilehybridsarepropagatedforseveralgenerationsanextremeamountofvariabilityintheiroffspringisnotorious。butsomefewcasesbothofhybridsandmongrelslongretaininguniformityofcharactercouldbegiven。Thevariability，however，inthesuccessivegenerationsofmongrelsis，perhaps，greaterthaninhybrids。

Thisgreatervariabilityofmongrelsthanofhybridsdoesnotseemtomeatallsurprising。Fortheparentsofmongrelsarevarieties，andmostlydomesticvarieties(veryfewexperimentshavingbeentriedonnaturalvarieties)，andthisimpliesinmostcasesthattherehasbeenrecentvariability。

andthereforewemightexpectthatsuchvariabilitywouldoftencontinueandbesuper-addedtothatarisingfromthemereactofcrossing。Theslightdegreeofvariabilityinhybridsfromthefirstcrossorinthefirstgeneration，incontrastwiththeirextremevariabilityinthesucceedinggenerations，isacuriousfactanddeservesattention。ForitbearsonandcorroboratestheviewwhichIhavetakenonthecauseofordinaryvariability。namely，thatitisduetothereproductivesystembeingeminentlysensitivetoanychangeintheconditionsoflife，beingthusoftenrenderedeitherimpotentoratleastincapableofitsproperfunctionofproducingoffspringidenticalwiththeparent-form。Nowhybridsinthefirstgenerationaredescendedfromspecies(excludingthoselongcultivated)whichhavenothadtheirreproductivesystemsinanywayaffected，andtheyarenotvariable。

buthybridsthemselveshavetheirreproductivesystemsseriouslyaffected，andtheirdescendantsarehighlyvariable。

Buttoreturntoourcomparisonofmongrelsandhybrids：Gä。rtnerstatesthatmongrelsaremoreliablethanhybridstoreverttoeitherparent-form。

butthis，ifitbetrue，iscertainlyonlyadifferenceindegree。Gä。rtnerfurtherinsiststhatwhenanytwospecies，althoughmostcloselyalliedtoeachother，arecrossedwithathirdspecies，thehybridsarewidelydifferentfromeachother。whereasiftwoverydistinctvarietiesofonespeciesarecrossedwithanotherspecies，thehybridsdonotdiffermuch。

Butthisconclusion，asfarasIcanmakeout，isfoundedonasingleexperiment。

andseemsdirectlyopposedtotheresultsofseveralexperimentsmadebyKö。lreuter。

Thesealonearetheunimportantdifferences，whichGä。rtnerisabletopointout，betweenhybridandmongrelplants。Ontheotherhand，theresemblanceinmongrelsandinhybridstotheirrespectiveparents，moreespeciallyinhybridsproducedfromnearlyrelatedspecies，followsaccordingtoGä。rtnerthesamelaws。Whentwospeciesarecrossed，onehassometimesaprepotentpowerofimpressingitslikenessonthehybrid。andsoIbelieveittobewithvarietiesofplants。Withanimalsonevarietycertainlyoftenhasthisprepotentpoweroveranothervariety。Hybridplantsproducedfromareciprocalcross，generallyresembleeachotherclosely。andsoitiswithmongrelsfromareciprocalcross。Bothhybridsandmongrelscanbereducedtoeitherpureparent-form，byrepeatedcrossesinsuccessivegenerationswitheitherparent。

Theseseveralremarksareapparentlyapplicabletoanimals。butthesubjectishereexcessivelycomplicated，partlyowingtotheexistenceofsecondarysexualcharacters。butmoreespeciallyowingtoprepotencyintransmittinglikenessrunningmorestronglyinonesexthanintheother，bothwhenonespeciesiscrossedwithanother，andwhenonevarietyiscrossedwithanothervariety。Forinstance，Ithinkthoseauthorsareright，whomaintainthattheasshasaprepotentpoweroverthehorse，sothatboththemuleandthehinnymoreresembletheassthanthehorse。butthattheprepotencyrunsmorestronglyinthemale-assthaninthefemale，sothatthemule，whichistheoffspringofthemale-assandmare，ismorelikeanass，thanisthehinny，whichistheoffspringofthefemale-assandstallion。

Muchstresshasbeenlaidbysomeauthorsonthesupposedfact，thatmongrelanimalsaloneareborncloselylikeoneoftheirparents。butitcanbeshownthatthisdoessometimesoccurwithhybrids。yetIgrantmuchlessfrequentlywithhybridsthanwithmongrels。LookingtothecaseswhichIhavecollectedofcross-bredanimalscloselyresemblingoneparent，theresemblancesseemchieflyconfinedtocharactersalmostmonstrousintheirnature，andwhichhavesuddenlyappearedsuchasalbinism，melanism，deficiencyoftailorhorns，oradditionalfingersandtoes。anddonotrelatetocharacterswhichhavebeenslowlyacquiredbyselection。Consequently，suddenreversionstotheperfectcharacterofeitherparentwouldbemorelikelytooccurwithmongrels，whicharedescendedfromvarietiesoftensuddenlyproducedandsemi-monstrousincharacter，thanwithhybrids，whicharedescendedfromspeciesslowlyandnaturallyproduced。OnthewholeIentirelyagreewithDrProsperLucas，who，afterarranginganenormousbodyoffactswithrespecttoanimals，comestotheconclusion，thatthelawsofresemblanceofthechildtoitsparentsarethesame，whetherthetwoparentsdiffermuchorlittlefromeachother，namelyintheunionofindividualsofthesamevariety，orofdifferentvarieties，orofdistinctspecies。

Layingasidethequestionoffertilityandsterility，inallotherrespectsthereseemstobeageneralandclosesimilarityintheoffspringofcrossedspecies，andofcrossedvarieties。Ifwelookatspeciesashavingbeenspeciallycreated，andatvarietiesashavingbeenproducedbysecondarylaws，thissimilaritywouldbeanastonishingfact。Butitharmonizesperfectlywiththeviewthatthereisnoessentialdistinctionbetweenspeciesandvarieties。

SummaryofChapter。Firstcrossesbetweenformssufficientlydistincttoberankedasspecies，andtheirhybrids，areverygenerally，butnotuniversally，sterile。Thesterilityisofalldegrees，andisoftensoslightthatthetwomostcarefulexperimentalistswhohaveeverlived，havecometodiametricallyoppositeconclusionsinrankingformsbythistest。Thesterilityisinnatelyvariableinindividualsofthesamespecies，andiseminentlysusceptibleoffavourableandunfavourableconditions。

Thedegreeofsterilitydoesnotstrictlyfollowsystematicaffinity，butisgovernedbyseveralcuriousandcomplexlaws。Itisgenerallydifferent，andsometimeswidelydifferent，inreciprocalcrossesbetweenthesametwospecies。Itisnotalwaysequalindegreeinafirstcrossandinthehybridproducedfromthiscross。

Inthesamemannerasingraftingtrees，thecapacityofonespeciesorvarietytotakeonanother，isincidentalongenerallyunknowndifferencesintheirvegetativesystems，soincrossing，thegreaterorlessfacilityofonespeciestounitewithanother，isincidentalonunknowndifferencesintheirreproductivesystems。Thereisnomorereasontothinkthatspecieshavebeenspeciallyendowedwithvariousdegreesofsterilitytopreventthemcrossingandblendinginnature，thantothinkthattreeshavebeenspeciallyendowedwithvariousandsomewhatanalogousdegreesofdifficultyinbeinggraftedtogetherinordertopreventthembecominginarchedinourforests。

Thesterilityoffirstcrossesbetweenpurespecies，whichhavetheirreproductivesystemsperfect，seemstodependonseveralcircumstances。

insomecaseslargelyontheearlydeathoftheembryo。Thesterilityofhybrids，whichhavetheirreproductivesystemsimperfect，andwhichhavehadthissystemandtheirwholeorganisationdisturbedbybeingcompoundedoftwodistinctspecies，seemscloselyalliedtothatsterilitywhichsofrequentlyaffectspurespecies，whentheirnaturalconditionsoflifehavebeendisturbed。Thisviewissupportedbyaparallelismofanotherkind。namely，thatthecrossingofformsonlyslightlydifferentisfavourabletothevigourandfertilityoftheiroffspring。andthatslightchangesintheconditionsoflifeareapparentlyfavourabletothevigourandfertilityofallorganicbeings。Itisnotsurprisingthatthedegreeofdifficultyinunitingtwospecies，andthedegreeofsterilityoftheirhybrid-offspringshouldgenerallycorrespond，thoughduetodistinctcauses。forbothdependontheamountofdifferenceofsomekindbetweenthespecieswhicharecrossed。Norisitsurprisingthatthefacilityofeffectingafirstcross，thefertilityofthehybridsproduced，andthecapacityofbeinggraftedtogetherthoughthislattercapacityevidentlydependsonwidelydifferentcircumstancesshouldallrun，toacertainextent，parallelwiththesystematicaffinityoftheformswhicharesubjectedtoexperiment。forsystematicaffinityattemptstoexpressallkindsofresemblancebetweenallspecies。

Firstcrossesbetweenformsknowntobevarieties，orsufficientlyaliketobeconsideredasvarieties，andtheirmongreloffspring，areverygenerally，butnotquiteuniversally，fertile。Noristhisnearlygeneralandperfectfertilitysurprising，whenwerememberhowliablewearetoargueinacirclewithrespecttovarietiesinastateofnature。andwhenwerememberthatthegreaternumberofvarietieshavebeenproducedunderdomesticationbytheselectionofmereexternaldifferences，andnotofdifferencesinthereproductivesystem。Inallotherrespects，excludingfertility，thereisaclosegeneralresemblancebetweenhybridsandmongrels。Finally，then，thefactsbrieflygiveninthischapterdonotseemtomeopposedto，butevenrathertosupporttheview，thatthereisnofundamentaldistinctionbetweenspeciesandvarieties。PreviousChapterNextChapterTheOriginofSpecies-Chapter9TheOriginofSpeciesbyCharlesDarwinPreviousChapterNextChapterChapter9-OntheImperfectionoftheGeologicalRecordINthesixthchapterIenumeratedthechiefobjectionswhichmightbejustlyurgedagainsttheviewsmaintainedinthisvolume。Mostofthemhavenowbeendiscussed。One，namelythedistinctnessofspecificforms，andtheirnotbeingblendedtogetherbyinnumerabletransitionallinks，isaveryobviousdifficulty。Iassignedreasonswhysuchlinksdonotcommonlyoccuratthepresentday，underthecircumstancesapparentlymostfavourablefortheirpresence，namelyonanextensiveandcontinuousareawithgraduatedphysicalconditions。Iendeavouredtoshow，thatthelifeofeachspeciesdependsinamoreimportantmanneronthepresenceofotheralreadydefinedorganicforms，thanonclimate。and，therefore，thatthereallygoverningconditionsoflifedonotgraduateawayquiteinsensiblylikeheatormoisture。Iendeavoured，also，toshowthatintermediatevarieties，fromexistinginlessernumbersthantheformswhichtheyconnect，willgenerallybebeatenoutandexterminatedduringthecourseoffurthermodificationandimprovement。Themaincause，however，ofinnumerableintermediatelinksnotnowoccurringeverywherethroughoutnaturedependsontheveryprocessofnaturalselection，throughwhichnewvarietiescontinuallytaketheplacesofandexterminatetheirparent-forms。Butjustinproportionasthisprocessofexterminationhasactedonanenormousscale，somustthenumberofintermediatevarieties，whichhaveformerlyexistedontheearth，betrulyenormous。Whythenisnoteverygeologicalformationandeverystratumfullofsuchintermediatelinks？Geologyassuredlydoesnotrevealanysuchfinelygraduatedorganicchain。andthis，perhaps，isthemostobviousandgravestobjectionwhichcanbeurgedagainstmytheory。Theexplanationlies，asIbelieve，intheextremeimperfectionofthegeologicalrecord。

Inthefirstplaceitshouldalwaysbeborneinmindwhatsortofintermediateformsmust，onmytheory，haveformerlyexisted。Ihavefounditdifficult，whenlookingatanytwospecies，toavoidpicturingtomyself，formsdirectlyintermediatebetweenthem。Butthisisawhollyfalseview。weshouldalwayslookforformsintermediatebetweeneachspeciesandacommonbutunknownprogenitor。andtheprogenitorwillgenerallyhavedifferedinsomerespectsfromallitsmodifieddescendants。Togiveasimpleillustration：thefantailandpouterpigeonshavebothdescendedfromtherock-pigeon。ifwepossessedalltheintermediatevarietieswhichhaveeverexisted，weshouldhaveanextremelycloseseriesbetweenbothandtherock-pigeon。butweshouldhavenovarietiesdirectlyintermediatebetweenthefantailandpouter。

none，forinstance，combiningatailsomewhatexpandedwithacropsomewhatenlarged，thecharacteristicfeaturesofthesetwobreeds。Thesetwobreeds，moreover，havebecomesomuchmodified，thatifwehadnohistoricalorindirectevidenceregardingtheirorigin，itwouldnothavebeenpossibletohavedeterminedfromamerecomparisonoftheirstructurewiththatoftherock-pigeon，whethertheyhaddescendedfromthisspeciesorfromsomeotheralliedspecies，suchasC。oenas。

Sowithnaturalspecies，ifwelooktoformsverydistinct，forinstancetothehorseandtapir，wehavenoreasontosupposethatlinkseverexisteddirectlyintermediatebetweenthem，butbetweeneachandanunknowncommonparent。Thecommonparentwillhavehadinitswholeorganisationmuchgeneralresemblancetothetapirandtothehorse。butinsomepointsofstructuremayhavedifferedconsiderablyfromboth，evenperhapsmorethantheydifferfromeachother。Henceinallsuchcases，weshouldbeunabletorecognisetheparent-formofanytwoormorespecies，evenifwecloselycomparedthestructureoftheparentwiththatofitsmodifieddescendants，unlessatthesametimewehadanearlyperfectchainoftheintermediatelinks。

Itisjustpossiblebymytheory，thatoneoftwolivingformsmighthavedescendedfromtheother。forinstance，ahorsefromatapir。andinthiscasedirectintermediatelinkswillhaveexistedbetweenthem。Butsuchacasewouldimplythatoneformhadremainedforaverylongperiodunaltered，whilstitsdescendantshadundergoneavastamountofchange。andtheprincipleofcompetitionbetweenorganismandorganism，betweenchildandparent，willrenderthisaveryrareevent。forinallcasesthenewandimprovedformsoflifewilltendtosupplanttheoldandunimproved。

Bythetheoryofnaturalselectionalllivingspecieshavebeenconnectedwiththeparent-speciesofeachgenus，bydifferencesnotgreaterthanweseebetweenthevarietiesofthesamespeciesatthepresentday。andtheseparent-species，nowgenerallyextinct，haveintheirturnbeensimilarlyconnectedwithmoreancientspecies。andsoonbackwards，alwaysconvergingtothecommonancestorofeachgreatclass。Sothatthenumberofintermediateandtransitionallinks，betweenalllivingandextinctspecies，musthavebeeninconceivablygreat。Butassuredly，ifthistheorybetrue，suchhaveliveduponthisearth。

OnthelapseofTime。Independentlyofournotfindingfossilremainsofsuchinfinitelynumerousconnectinglinks，itmaybeobjected，thattimewillnothavesufficedforsogreatanamountoforganicchange，allchangeshavingbeeneffectedveryslowlythroughnaturalselection。

Itishardlypossibleformeeventorecalltothereader，whomaynotbeapracticalgeologist，thefactsleadingthemindfeeblytocomprehendthelapseoftime。HewhocanreadSirCharlesLyell’sgrandworkonthePrinciplesofGeology，whichthefuturehistorianwillrecogniseashavingproducedarevolutioninnaturalscience，yetdoesnotadmithowincomprehensiblyvasthavebeenthepastperiodsoftime，mayatonceclosethisvolume。

NotthatitsufficestostudythePrinciplesofGeology，ortoreadspecialtreatisesbydifferentobserversonseparateformations，andtomarkhoweachauthorattemptstogiveaninadequateideaofthedurationofeachformationoreveneachstratum。Amanmustforyearsexamineforhimselfgreatpilesofsuperimposedstrata，andwatchtheseaatworkgrindingdownoldrocksandmakingfreshsediment，beforehecanhopetocomprehendanythingofthelapseoftime，themonumentsofwhichweseearoundus。

Itisgoodtowanderalonglinesofsea-coast，whenformedofmoderatelyhardrocks，andmarktheprocessofdegradation。Thetidesinmostcasesreachthecliffsonlyforashorttimetwiceaday，andthewaveseatintothemonlywhentheyarechargedwithsandorpebbles。forthereisreasontobelievethatpurewatercaneffectlittleornothinginwearingawayrock。Atlastthebaseofthecliffisundermined，hugefragmentsfalldown，andtheseremainingfixed，havetobewornaway，atombyatom，untilreducedinsizetheycanberolledaboutbythewaves，andthenaremorequicklygroundintopebbles，sand，ormud。Buthowoftendoweseealongthebasesofretreatingcliffsroundedboulders，allthicklyclothedbymarineproductions，showinghowlittletheyareabradedandhowseldomtheyarerolledabout！Moreover，ifwefollowforafewmilesanylineofrockycliff，whichisundergoingdegradation，wefindthatitisonlyhereandthere，alongashortlengthorroundapromontory，thatthecliffsareatthepresenttimesuffering。Theappearanceofthesurfaceandthevegetationshowthatelsewhereyearshaveelapsedsincethewaterswashedtheirbase。

Hewhomostcloselystudiestheactionoftheseaonourshores，will，Ibelieve，bemostdeeplyimpressedwiththeslownesswithwhichrockycoastsarewornaway。TheobservationsonthisheadbyHughMiller，andbythatexcellentobserverMr。SmithofJordanHill，aremostimpressive。

Withthemindthusimpressed，letanyoneexaminebedsofconglomeratemanythousandfeetinthickness，which，thoughprobablyformedataquickerratethanmanyotherdeposits，yet，frombeingformedofwornandroundedpebbles，eachofwhichbearsthestampoftime，aregoodtoshowhowslowlythemasshasbeenaccumulated。LethimrememberLyell’sprofoundremark，thatthethicknessandextentofsedimentaryformationsaretheresultandmeasureofthedegradationwhichtheearth’scrusthaselsewheresuffered。

Andwhatanamountofdegradationisimpliedbythesedimentarydepositsofmanycountries！ProfessorRamsayhasgivenmethemaximumthickness，inmostcasesfromactualmeasurement，inafewcasesfromestimate，ofeachformationindifferentpartsofGreatBritain。andthisistheresult：-

FeetPalaeozoicstrata(notincludingigneousbeds)57，154

Secondarystrata13，190

Tertiarystrata2，240

makingaltogether72，584feet。thatis，verynearlythirteenandthree-quartersBritishmiles。Someoftheseformations，whicharerepresentedinEnglandbythinbeds，arethousandsoffeetinthicknessontheContinent。Moreover，betweeneachsuccessiveformation，wehave，intheopinionofmostgeologists，enormouslylongblankperiods。SothattheloftypileofsedimentaryrocksinBritain，givesbutaninadequateideaofthetimewhichhaselapsedduringtheiraccumulation。yetwhattimethismusthaveconsumed！GoodobservershaveestimatedthatsedimentisdepositedbythegreatMississippiriverattherateofonly600feetinahundredthousandyears。Thisestimatemaybequiteerroneous。yet，consideringoverwhatwidespacesveryfinesedimentistransportedbythecurrentsofthesea，theprocessofaccumulationinanyoneareamustbeextremelyslow。

Buttheamountofdenudationwhichthestratahaveinmanyplacessuffered，independentlyoftherateofaccumulationofthedegradedmatter，probablyoffersthebestevidenceofthelapseoftime。Irememberhavingbeenmuchstruckwiththeevidenceofdenudation，whenviewingvolcanicislands，whichhavebeenwornbythewavesandparedallroundintoperpendicularcliffsofoneortwothousandfeetinheight。forthegentleslopeofthelava-streams，duetotheirformerlyliquidstate，showedataglancehowfarthehard，rockybedshadonceextendedintotheopenocean。Thesamestoryisstillmoreplainlytoldbyfaults，thosegreatcracksalongwhichthestratahavebeenupheavedononeside，orthrowndownontheother，totheheightordepthofthousandsoffeet。forsincethecrustcracked，thesurfaceofthelandhasbeensocompletelyplaneddownbytheactionofthesea，thatnotraceofthesevastdislocationsisexternallyvisible。

TheCravenfault，forinstance，extendsforupwardsof30miles，andalongthislinetheverticaldisplacementofthestratahasvariedfrom600to3000feet。Prof。RamsayhaspublishedanaccountofadownthrowinAngleseaof2300feet。andheinformsmethathefullybelievesthereisoneinMerionethshireof12，000feet。yetinthesecasesthereisnothingonthesurfacetoshowsuchprodigiousmovements。thepileofrocksontheoneorothersidehavingbeensmoothlysweptaway。Theconsiderationofthesefactsimpressesmymindalmostinthesamemannerasdoesthevainendeavourtograpplewiththeideaofeternity。

Iamtemptedtogiveoneothercase，thewell-knownoneofthedenudationoftheWeald。ThoughitmustbeadmittedthatthedenudationoftheWealdhasbeenameretrifle，incomparisonwiththatwhichhasremovedmassesofourPalaeozoicstrata，inpartstenthousandfeetinthickness，asshowninProf。Ramsay’smasterlymemoironthissubject。YetitisanadmirablelessontostandontheNorthDownsandtolookatthedistantSouthDowns。

for，rememberingthatatnogreatdistancetothewestthenorthernandsouthernescarpmentsmeetandclose，onecansafelypicturetooneselfthegreatdomeofrockswhichmusthavecovereduptheWealdwithinsolimitedaperiodassincethelatterpartoftheChalkformation。ThedistancefromthenortherntothesouthernDownsisabout22miles，andthethicknessoftheseveralformationsisonanaverageabout1100feet，asIaminformedbyProf。Ramsay。Butif，assomegeologistssuppose，arangeofolderrocksunderliestheWeald，ontheflanksofwhichtheoverlyingsedimentarydepositsmighthaveaccumulatedinthinnermassesthanelsewhere，theaboveestimatewouldbeerroneous。butthissourceofdoubtprobablywouldnotgreatlyaffecttheestimateasappliedtothewesternextremityofthedistrict。

If，then，weknewtherateatwhichtheseacommonlywearsawayalineofcliffofanygivenheight，wecouldmeasurethetimerequisitetohavedenudedtheWeald。This，ofcourse，cannotbedone。butwemay，inordertoformsomecrudenotiononthesubject，assumethattheseawouldeatintocliffs500feetinheightattherateofoneinchinacentury。Thiswillatfirstappearmuchtoosmallanallowance。butitisthesameasifweweretoassumeacliffoneyardinheighttobeeatenbackalongawholelineofcoastattherateofoneyardinnearlyeverytwenty-twoyears。Idoubtwhetheranyrock，evenassoftaschalk，wouldyieldatthisrateexceptingonthemostexposedcoasts。thoughnodoubtthedegradationofaloftycliffwouldbemorerapidfromthebreakageofthefallenfragments。

Ontheotherhand，Idonotbelievethatanylineofcoast，tenortwentymilesinlength，eversuffersdegradationatthesametimealongitswholeindentedlength。andwemustrememberthatalmostallstratacontainharderlayersornodules，whichfromlongresistingattritionformabreakwateratthebase。Hence，underordinarycircumstances，Iconcludethatforacliff500feetinheight，adenudationofoneinchpercenturyforthewholelengthwouldbeanampleallowance。Atthisrate，ontheabovedata，thedenudationoftheWealdmusthaverequired306，662，400years。orsaythreehundredmillionyears。

TheactionoffreshwateronthegentlyinclinedWealdendistrict，whenupraised，couldhardlyhavebeengreat，butitwouldsomewhatreducetheaboveestimate。Ontheotherhand，duringoscillationsoflevel，whichweknowthisareahasundergone，thesurfacemayhaveexistedformillionsofyearsasland，andthushaveescapedtheactionofthesea：whendeeplysubmergedforperhapsequallylongperiods，itwould，likewise，haveescapedtheactionofthecoast-waves。Sothatinallprobabilityafarlongerperiodthan300millionyearshaselapsedsincethelatterpartoftheSecondaryperiod。

Ihavemadethesefewremarksbecauseitishighlyimportantforustogainsomenotion，howeverimperfect，ofthelapseofyears。Duringeachoftheseyears，overthewholeworld，thelandandthewaterhasbeenpeopledbyhostsoflivingforms。Whataninfinitenumberofgenerations，whichthemindcannotgrasp，musthavesucceededeachotherinthelongrollofyears！Nowturntoourrichestgeologicalmuseums，andwhatapaltrydisplaywebehold！

OnthepoornessofourPalaeontologicalcollections。ThatourPalaeontologicalcollectionsareveryimperfect，isadmittedbyeveryone。

TheremarkofthatadmirablePalaeontologist，thelateEdwardForbes，shouldnotbeforgotten，namely，thatnumbersofourfossilspeciesareknownandnamedfromsingleandoftenbrokenspecimens，orfromafewspecimenscollectedonsomeonespot。Onlyasmallportionofthesurfaceoftheearthhasbeengeologicallyexplored，andnopartwithsufficientcare，astheimportantdiscoveriesmadeeveryyearinEuropeprove。Noorganismwhollysoftcanbepreserved。Shellsandboneswilldecayanddisappearwhenleftonthebottomofthesea，wheresedimentisnotaccumulating。

Ibelievewearecontinuallytakingamosterroneousview，whenwetacitlyadmittoourselvesthatsedimentisbeingdepositedovernearlythewholebedofthesea，ataratesufficientlyquicktoembedandpreservefossilremains。Throughoutanenormouslylargeproportionoftheocean，thebrightbluetintofthewaterbespeaksitspurity。Themanycasesonrecordofaformationconformablycovered，afteranenormousintervaloftime，byanotherandlaterformation，withouttheunderlyingbedhavingsufferedintheintervalanywearandtear，seemexplicableonlyontheviewofthebottomoftheseanotrarelylyingforagesinanunalteredcondition。

Theremainswhichdobecomeembedded，ifinsandorgravel，willwhenthebedsareupraisedgenerallybedissolvedbythepercolationofrain-water。

Isuspectthatbutfewoftheverymanyanimalswhichliveonthebeachbetweenhighandlowwatermarkarepreserved。Forinstance，theseveralspeciesoftheChthamalinae(asub-familyofsessilecirripedes)coattherocksallovertheworldininfinitenumbers：theyareallstrictlylittoral，withtheexceptionofasingleMediterraneanspecies，whichinhabitsdeepwaterandhasbeenfoundfossilinSicily，whereasnotoneotherspecieshashithertobeenfoundinanytertiaryformation：yetitisnowknownthatthegenusChthamalusexistedduringthechalkperiod。ThemolluscangenusChitonoffersapartiallyanalogouscase。

WithrespecttotheterrestrialproductionswhichlivedduringtheSecondaryandPalaeozoicperiods，itissuperfluoustostatethatourevidencefromfossilremainsisfragmentaryinanextremedegree。Forinstance，notalandshellisknownbelongingtoeitherofthesevastperiods，withoneexceptiondiscoveredbySirC。LyellinthecarboniferousstrataofNorthAmerica。Inregardtomammiferousremains，asingleglanceatthehistoricaltablepublishedintheSupplementtoLyell’sManual，willbringhomethetruth，howaccidentalandrareistheirpreservation，farbetterthanpagesofdetail。Noristheirraritysurprising，whenwerememberhowlargeaproportionofthebonesoftertiarymammalshavebeendiscoveredeitherincavesorinlacustrinedeposits。andthatnotacaveortruelacustrinebedisknownbelongingtotheageofoursecondaryorpalaeozoicformations。

Buttheimperfectioninthegeologicalrecordmainlyresultsfromanotherandmoreimportantcausethananyoftheforegoing。namely，fromtheseveralformationsbeingseparatedfromeachotherbywideintervalsoftime。Whenweseetheformationstabulatedinwrittenworks，orwhenwefollowtheminnature，itisdifficulttoavoidbelievingthattheyarecloselyconsecutive。

Butweknow，forinstance，fromSirR。Murchison’sgreatworkonRussia，whatwidegapsthereareinthatcountrybetweenthesuperimposedformations。

soitisinNorthAmerica，andinmanyotherpartsoftheworld。Themostskilfulgeologist，ifhisattentionhadbeenexclusivelyconfinedtotheselargeterritories，wouldneverhavesuspectedthatduringtheperiodswhichwereblankandbarreninhisowncountry，greatpilesofsediment，chargedwithnewandpeculiarformsoflife，hadelsewherebeenaccumulated。Andifineachseparateterritory，hardlyanyideacanbeformedofthelengthoftimewhichhaselapsedbetweentheconsecutiveformations，wemayinferthatthiscouldnowherebeascertained。Thefrequentandgreatchangesinthemineralogicalcompositionofconsecutiveformations，generallyimplyinggreatchangesinthegeographyofthesurroundinglands，whencethesedimenthasbeenderived，accordswiththebeliefofvastintervalsoftimehavingelapsedbetweeneachformation。

Butwecan，Ithink，seewhythegeologicalformationsofeachregionarealmostinvariablyintermittent。thatis，havenotfollowedeachotherinclosesequence。ScarcelyanyfactstruckmemorewhenexaminingmanyhundredmilesoftheSouthAmericancoasts，whichhavebeenupraisedseveralhundredfeetwithintherecentperiod，thantheabsenceofanyrecentdepositssufficientlyextensivetolastforevenashortgeologicalperiod。Alongthewholewestcoast，whichisinhabitedbyapeculiarmarinefauna，tertiarybedsaresoscantilydeveloped，thatnorecordofseveralsuccessiveandpeculiarmarinefaunaswillprobablybepreservedtoadistantage。AlittlereflectionwillexplainwhyalongtherisingcoastofthewesternsideofSouthAmerica，noextensiveformationswithrecentortertiaryremainscananywherebefound，thoughthesupplyofsedimentmustforageshavebeengreat，fromtheenormousdegradationofthecoast-rocksandfrommuddystreamsenteringthesea。Theexplanation，nodoubt，is，thatthelittoralandsub-littoraldepositsarecontinuallywornaway，assoonastheyarebroughtupbytheslowandgradualrisingofthelandwithinthegrindingactionofthecoast-waves。

Wemay，Ithink，safelyconcludethatsedimentmustbeaccumulatedinextremelythick，solid，orextensivemasses，inordertowithstandtheincessantactionofthewaves，whenfirstupraisedandduringsubsequentoscillationsoflevel。Suchthickandextensiveaccumulationsofsedimentmaybeformedintwoways。either，inprofounddepthsofthesea，inwhichcase，judgingfromtheresearchesofE。Forbes，wemayconcludethatthebottomwillbeinhabitedbyextremelyfewanimals，andthemasswhenupraisedwillgiveamostimperfectrecordoftheformsoflifewhichthenexisted。

or，sedimentmaybeaccumulatedtoanythicknessandextentoverashallowbottom，ifitcontinueslowlytosubside。Inthislattercase，aslongastherateofsubsidenceandsupplyofsedimentnearlybalanceeachother，theseawillremainshallowandfavourableforlife，andthusafossiliferousformationthickenough，whenupraised，toresistanyamountofdegradation，maybeformed。

Iamconvincedthatallourancientformations，whicharerichinfossils，havethusbeenformedduringsubsidence。Sincepublishingmyviewsonthissubjectin1845，IhavewatchedtheprogressofGeology，andhavebeensurprisedtonotehowauthorafterauthor，intreatingofthisorthatgreatformation，hascometotheconclusionthatitwasaccumulatedduringsubsidence。Imayadd，thattheonlyancienttertiaryformationonthewestcoastofSouthAmerica，whichhasbeenbulkyenoughtoresistsuchdegradationasithasasyetsuffered，butwhichwillhardlylasttoadistantgeologicalage，wascertainlydepositedduringadownwardoscillationoflevel，andthusgainedconsiderablethickness。

Allgeologicalfactstellusplainlythateachareahasundergonenumerousslowoscillationsoflevel，andapparentlytheseoscillationshaveaffectedwidespaces。Consequentlyformationsrichinfossilsandsufficientlythickandextensivetoresistsubsequentdegradation，mayhavebeenformedoverwidespacesduringperiodsofsubsidence，butonlywherethesupplyofsedimentwassufficienttokeeptheseashallowandtoembedandpreservetheremainsbeforetheyhadtimetodecay。Ontheotherhand，aslongasthebedofthesearemainedstationary，thickdepositscouldnothavebeenaccumulatedintheshallowparts，whicharethemostfavourabletolife。

Stilllesscouldthishavehappenedduringthealternateperiodsofelevation。

or，tospeakmoreaccurately，thebedswhichwerethenaccumulatedwillhavebeendestroyedbybeingupraisedandbroughtwithinthelimitsofthecoast-action。

Thusthegeologicalrecordwillalmostnecessarilyberenderedintermittent。

Ifeelmuchconfidenceinthetruthoftheseviews，fortheyareinstrictaccordancewiththegeneralprinciplesinculcatedbySirC。Lyell。andE。Forbesindependentlyarrivedatasimilarconclusion。

Oneremarkishereworthapassingnotice。Duringperiodsofelevationtheareaofthelandandoftheadjoiningshoalpartsoftheseawillbeincreased，andnewstationswilloftenbeformed。allcircumstancesmostfavourable，aspreviouslyexplained，fortheformationofnewvarietiesandspecies。butduringsuchperiodstherewillgenerallybeablankinthegeologicalrecord。Ontheotherhand，duringsubsidence，theinhabitedareaandnumberofinhabitantswilldecrease(exceptingtheproductionsontheshoresofacontinentwhenfirstbrokenupintoanarchipelago)，andconsequentlyduringsubsidence，thoughtherewillbemuchextinction，fewernewvarietiesorspecieswillbeformed。anditisduringtheseveryperiodsofsubsidence，thatourgreatdepositsrichinfossilshavebeenaccumulated。Naturemayalmostbesaidtohaveguardedagainstthefrequentdiscoveryofhertransitionalorlinkingforms。