Sexualandasexualreproductionarethusseennottodifferessentially;andwehavealreadyshownthatasexualreproduction，thepowerofregrowthanddevelopmentareallpartsofoneandthesamegreatlaw。
REGROWTHOFAMPUTATEDPARTS。
Thissubjectdeservesalittlefurtherdiscussion。Amultitudeoftheloweranimalsandsomevertebratespossessthiswonderfulpower。Forinstance，Spallanzanicutoffthelegsandtailofthesamesalamandersixtimessuccessively，andBonnet（27/18。Spallanzani’AnEssayonAnimalReproduction’
translatedbyDr。Maty1769page79。Bonnet’Oeuvresd’Hist。Nat。’tome5part14to。edition1781pages343，350。）didsoeighttimes;andoneachoccasionthelimbswerereproducedontheexactlineofamputation，withnopartdeficientorinexcess。Analliedanimal，theaxolotl，hadalimbbittenoff，whichwasreproducedinanabnormalcondition，butwhenthiswasamputateditwasreplacedbyaperfectlimb。（27/19。VulpianasquotedbyProf。Faivre’LaVariabilitedesEspeces’1868page112。）Thenewlimbsinthesecasesbudforth，andaredevelopedinthesamemannerasduringtheregulardevelopmentofayounganimal。Forinstance，withtheAmblystomalurida，threetoesarefirstdeveloped，thenthefourth，andonthehind—feetthefifth，andsoitiswithareproducedlimb。（27/20。Dr。P。Hoy’TheAmericanNaturalist’September1871page579。）
Thepowerofregrowthisgenerallymuchgreaterduringtheyouthofananimalorduringtheearlierstagesofitsdevelopmentthanduringmaturity。ThelarvaeortadpolesoftheBatrachiansarecapableofreproducinglostmembers，butnotsotheadults。（27/21。Dr。GuntherinOwen’AnatomyofVertebrates’
volume11866page567。Spallanzanihasmadesimilarobservations。）Matureinsectshavenopowerofregrowth，exceptinginoneorder，whilstthelarvaeofmanykindshavethispower。Animalslowinthescaleareable，asageneralrule，toreproducelostpartsfarmoreeasilythanthosewhicharemorehighlyorganised。Themyriapodsofferagoodillustrationofthisrule;buttherearesomestrangeexceptionstoit——thusNemerteans，thoughlowlyorganised，aresaidtoexhibitlittlepowerofregrowth。Withthehighervertebrata，suchasbirdsandmammals，thepowerisextremelylimited。（27/22。AthrushwasexhibitedbeforetheBritishAssociationatHullin1853whichhadlostitstarsus，andthismember，itwasasserted，hadbeenthricereproduced;havingbeenlost，Ipresume，eachtimebydisease。SirJ。PagetinformsmethathefeelssomedoubtaboutthefactsrecordedbySirJ。Simpson（’MonthlyJournalofMedicalScience’Edinburgh1848newseriesvolume2page890）oftheregrowthoflimbsinthewombinthecaseofman。）
Inthecaseofthoseanimalswhichmaybebisectedorchoppedintopieces，andofwhicheveryfragmentwillreproducethewhole，thepowerofregrowthmustbediffusedthroughoutthewholebody。NeverthelessthereseemstobemuchtruthintheviewmaintainedbyProf。Lessona（27/23。’AttidellaSoc。Ital。
diSc。Nat。’volume111869page493。），thatthiscapacityisgenerallyalocalisedandspecialone，servingtoreplacepartswhichareeminentlyliabletobelostineachparticularanimal。Themoststrikingcaseinfavourofthisview，isthattheterrestrialsalamander，accordingtoLessona，cannotreproducelostparts，whilstanotherspeciesofthesamegenus，theaquaticsalamander，hasextraordinarypowersofregrowth，aswehavejustseen;andthisanimaliseminentlyliabletohaveitslimbs，tail，eyesandjawsbittenoffbyothertritons。（27/24。Lessonastatesthatthisissointhepaperjustreferredto。Seealso’TheAmericanNaturalist’September1871page579。）Evenwiththeaquaticsalamanderthecapacityistoacertainextentlocalised，forwhenM。Philipeaux（27/25。’ComptesRendus’October1，1866andJune1867。）
extirpatedtheentireforelimbtogetherwiththescapula，thepowerofregrowthwascompletelylost。Itisalsoaremarkablefact，standinginoppositiontoaverygeneralrule，thattheyoungoftheaquaticsalamanderdonotpossessthepowerofrepairingtheirlimbsinanequaldegreewiththeadults（27/26。Bonnet’OeuvresHist。Nat。’volume5page294，asquotedbyProf。Rollestoninhisremarkableaddresstothe36thannualmeetingoftheBritishMedicalAssociation。）butIdonotknowthattheyaremoreactive，orcanotherwisebetterescapethelossoftheirlimbs，thantheadults。Thewalking—stickinsect，Diapheromerafemorata，likeotherinsectsofthesameorder，canreproduceitslegsinthematurestate，andthesefromtheirgreatlengthmustbeliabletobelost:butthecapacityislocalised（asinthecaseofthesalamander），forDr。Scudderfound（27/27。’Proc。BostonSoc。ofNat。Hist。’volume121868—69page1。），thatifthelimbwasremovedwithinthetrochanto—femoralarticulation，itwasneverrenewed。Whenacrabisseizedbyoneofitslegs，thisisthrownoffatthebasaljoint，beingafterwardsreplacedbyanewleg;anditisgenerallyadmittedthatthisisaspecialprovisionforthesafetyoftheanimal。Lastly，withgasteropodmolluscs，whicharewellknowntohavethepowerofreproducingtheirheads，Lessonashowsthattheyareveryliabletohavetheirheadsbittenoffbyfishes;therestofthebodybeingprotectedbytheshell。Evenwithplantsweseesomethingofthesamekind，fornon—deciduousleavesandyoungstemshavenopowerofregrowth，thesepartsbeingeasilyreplacedbygrowthfromnewbuds;whilstthebarkandsubjacenttissuesofthetrunksoftreeshavegreatpowerofregrowth，probablyonaccountoftheirincreaseindiameter，andoftheirliabilitytoinjuryfrombeinggnawedbyanimals。
GRAFT—HYBRIDS。
Itiswellknownfrominnumerabletrialsmadeinallpartsoftheworld，thatbudsmaybeinsertedintoastock，andthattheplantsthusraisedarenotaffectedinagreaterdegreethancanbeaccountedforbychangednutrition。
Nordotheseedlingsraisedfromsuchinsertedbudspartakeofthecharacterofthestock，thoughtheyaremoreliabletovarythanareseedlingsfromthesamevarietygrowingonitsownroots。Abud，also，maysportintoanewandstrongly—markedvarietywithoutanyotherbudonthesameplantbeingintheleastdegreeaffected。Wemaythereforeinfer，inaccordancewiththecommonview，thateachbudisadistinctindividual，andthatitsformativeelementsdonotspreadbeyondthepartssubsequentlydevelopedfromit。Nevertheless，wehaveseenintheabstractongraft—hybridisationintheeleventhchapterthatbudscertainlyincludeformativematter，whichcanoccasionallycombinewiththatincludedinthetissuesofadistinctvarietyorspecies;aplantintermediatebetweenthetwoparent—formsbeingthusproduced。Inthecaseofthepotatowehaveseenthatthetubersproducedfromabudofonekindinsertedintoanotherareintermediateincolour，size，shapeandstateofsurface;thatthestems，foliage，andevencertainconstitutionalpeculiarities，suchasprecocity，arelikewiseintermediate。Withthesewell—
establishedcases，theevidencethatgraft—hybridshavealsobeenproducedwiththelaburnum，orange，vine，rose，etc。，seemssufficient。Butwedonotknowunderwhatconditionsthisrareformofreproductionispossible。Fromtheseseveralcaseswelearntheimportantfactthatformativeelementscapableofblendingwiththoseofadistinctindividual（andthisisthechiefcharacteristicofsexualgeneration），arenotconfinedtothereproductiveorgans，butarepresentinthebudsandcellulartissueofplants;andthisisafactofthehighestphysiologicalimportance。
DIRECTACTIONOFTHEMALEELEMENTONTHEFEMALE。
Intheeleventhchapter，abundantproofsweregiventhatforeignpollenoccasionallyaffectsinadirectmannerthemother—plant。Thus，whenGallesiofertilisedanorange—flowerwithpollenfromthelemon，thefruitborestripesofperfectlycharacterisedlemon—peel。Withpeas，severalobservershaveseenthecolouroftheseed—coatsandevenofthepoddirectlyaffectedbythepollenofadistinctvariety。Soithasbeenwiththefruitoftheapple，whichconsistsofthemodifiedcalyxandupperpartoftheflower—stalk。Inordinarycasesthesepartsarewhollyformedbythemother—plant。Wehereseethattheformativeelementsincludedwithinthemaleelementorpollenofonevarietycanaffectandhybridise，notthepartwhichtheyareproperlyadaptedtoaffect，namely，theovules，butthepartially—developedtissuesofadistinctvarietyorspecies。Wearethusbroughthalf—waytowardsagraft—
hybrid，inwhichtheformativeelementsincludedwithinthetissuesofoneindividualcombinewiththoseincludedinthetissuesofadistinctvarietyorspecies，thusgivingrisetoanewandintermediateform，independentlyofthemaleorfemalesexualorgans。
Withanimalswhichdonotbreeduntilnearlymature，andofwhichallthepartsarethenfullydeveloped，itishardlypossiblethatthemaleelementshoulddirectlyaffectthefemale。Butwehavetheanalogousandperfectlywell—ascertainedcaseofthemaleelementaffecting（aswiththequaggaandLordMorton’smare）thefemaleorherova，insuchamannerthatwhensheisimpregnatedbyanothermaleheroffspringareaffectedandhybridisedbythefirstmale。Theexplanationwouldbesimpleifthespermatozoacouldkeepalivewithinthebodyofthefemaleduringthelongintervalwhichhassometimeselapsedbetweenthetwoactsofimpregnation;butnoonewillsupposethatthisispossiblewiththehigheranimals。
DEVELOPMENT。
Thefertilisedgermreachesmaturitybyavastnumberofchanges:theseareeitherslightandslowlyeffected，aswhenthechildgrowsintotheman，oraregreatandsudden，aswiththemetamorphosesofmostinsects。Betweentheseextremeswehaveeverygradation，evenwithinthesameclass;thus，asSirJ。
Lubbockhasshown（27/28。’Transact。Linn。Soc。’volume241863page62。）
thereisanEphemerousinsectwhichmoultsabovetwentytimes，undergoingeachtimeaslightbutdecidedchangeofstructure;andthesechanges，ashefurtherremarks，probablyrevealtousthenormalstagesofdevelopment，whichareconcealedandhurriedthroughorsuppressedinmostotherinsects。Inordinarymetamorphoses，thepartsandorgansappeartobecomechangedintothecorrespondingpartsinthenextstageofdevelopment;butthereisanotherformofdevelopment，whichhasbeencalledbyProfessorOwenmetagenesis。Inthiscase\"thenewpartsarenotmouldedupontheinnersurfaceoftheoldones。Theplasticforcehaschangeditscourseofoperation。Theoutercase，andallthatgaveformandcharactertotheprecedentindividual，perishandarecastoff;theyarenotchangedintothecorrespondingpartsofthenewindividual。Theseareduetoanewanddistinctdevelopmentalprocess，\"etc。
（27/29。’Parthenogenesis’1849pages25，26。Prof。Huxleyhassomeexcellentremarks（’MedicalTimes’1856page637）onthissubjectinreferencetothedevelopmentofstar—fishes，andshowshowcuriouslymetamorphosisgraduatesintogemmationorzoid—formation，whichisinfactthesameasmetagenesis。）
Metamorphosis，however，graduatessoinsensibly，intometagenesis，thatthetwoprocessescannotbedistinctlyseparated。Forinstance，inthelastchangewhichCirripedesundergo，thealimentarycanalandsomeotherorgansaremouldedonpre—existingparts;buttheeyesoftheoldandtheyounganimalaredevelopedinentirelydifferentpartsofthebody;thetipsofthematurelimbsareformedwithinthelarvallimbs，andmaybesaidtobemetamorphosedfromthem;buttheirbasalportionsandthewholethoraxaredevelopedinaplaneatrightanglestothelarvallimbsandthorax;andthismaybecalledmetagenesis。ThemetageneticprocessiscarriedtoanextremepointinthedevelopmentofsomeEchinoderms，fortheanimalinthesecondstageofdevelopmentisformedalmostlikeabudwithintheanimalofthefirststage，thelatterbeingthencastofflikeanoldvestment，yetsometimesmaintainingforashortperiodanindependentvitality。（27/30。Prof。J。ReayGreeneinGunther’s’RecordofZoolog。Lit。’1865page625。）If，insteadofasingleindividual，severalweretobethusdevelopedmetageneticallywithinapre—
existingform，theprocesswouldbecalledoneofalternategeneration。Theyoungthusdevelopedmayeithercloselyresembletheencasingparent—form，aswiththelarvaeofCecidomyia，ormaydiffertoanastonishingdegree，aswithmanyparasiticwormsandjelly—fishes;butthisdoesnotmakeanyessentialdifferenceintheprocess，anymorethanthegreatnessorabruptnessofthechangeinthemetamorphosesofinsects。
Thewholequestionofdevelopmentisofgreatimportanceforourpresentsubject。Whenanorgan，theeye，forinstance，ismetageneticallyformedinapartofthebodywhereduringthepreviousstageofdevelopmentnoeyeexisted，wemustlookatitasanewandindependentgrowth。Theabsoluteindependenceofnewandoldstructures，althoughcorrespondinginstructureandfunction，isstillmoreobviouswhenseveralindividualsareformedwithinapreviousform，asinthecasesofalternategeneration。Thesameimportantprincipleprobablycomeslargelyintoplayeveninthecaseofapparentlycontinuousgrowth，asweshallseewhenweconsidertheinheritanceofmodificationsatcorrespondingages。
Weareledtothesameconclusion，namely，theindependenceofpartssuccessivelydeveloped，byanotherandquitedistinctgroupoffacts。Itiswellknownthatmanyanimalsbelongingtothesameorder，andthereforenotdifferingwidelyfromeachother，passthroughanextremelydifferentcourseofdevelopment。Thuscertainbeetles，notinanywayremarkablydifferentfromothersofthesameorder，undergowhathasbeencalledahyper—metamorphosis——
thatis，theypassthroughanearlystagewhollydifferentfromtheordinarygrub—likelarva。Inthesamesub—orderofcrabs，namely，theMacroura，asFritzMullerremarks，therivercray—fishishatchedunderthesameformwhichiteverafterwardsretains;theyounglobsterhasdividedlegs，likeaMysis;
thePalaemonappearsundertheformofaZoea，andPeneusundertheNauplius—
form;andhowwonderfullytheselarvalformsdifferfromoneanother，isknowntoeverynaturalist。（27/31。FritzMuller’FurDarwin’1864s。65，71。Thehighestauthorityoncrustaceans，Prof。Milne—Edwards，insists（’Annal。desSci。Nat。’2ndseriesZoolog。tome3page322）onthedifferenceinthemetamorphosisofclosely—alliedgenera。）Someothercrustaceans，asthesameauthorobserves，startfromthesamepointandarriveatnearlythesameend，butinthemiddleoftheirdevelopmentarewidelydifferentfromoneanother。
StillmorestrikingcasescouldbegivenwithrespecttotheEchinodermata。
WiththeMedusaeorjelly—fishesProfessorAllmanobserves，\"TheclassificationoftheHydroidawouldbeacomparativelysimpletaskif，ashasbeenerroneouslyasserted，generically—identicalmedusoidsalwaysarosefromgenerically—identicalpolypoids;and，ontheotherhand，thatgenerically—
identicalpolypoidsalwaysgaveorigintogenerically—identicalmedusoids。\"Soagain，Dr。StrethillWrightremarks，\"Inthelife—historyoftheHydroidaeanyphase，planuloid，polypoid，ormedusoid，maybeabsent。\"（27/32。Prof。Allman’AnnalsandMag。ofNat。Hist。’3rdseriesvolume131864page348;Dr。S。
Wrightibidvolume81861page127。Seealsopage358foranalogousstatementsbySars。）
Accordingtothebeliefnowgenerallyacceptedbyourbestnaturalists，allthemembersofthesameorderorclass，forinstance，theMedusaeortheMacrourouscrustaceans，aredescendedfromacommonprogenitor。Duringtheirdescenttheyhavedivergedmuchinstructure，buthaveretainedmuchincommon;andthishasoccurred，thoughtheyhavepassedthroughandstillpassthroughmarvellouslydifferentmetamorphoses。Thisfactwellillustrateshowindependenteachstructureisfromthatwhichprecedesandthatwhichfollowsitinthecourseofdevelopment。
THEFUNCTIONALINDEPENDENCEOFTHEELEMENTSORUNITSOFTHEBODY。
Physiologistsagreethatthewholeorganismconsistsofamultitudeofelementalparts，whicharetoagreatextentindependentofoneanother。Eachorgan，saysClaudeBernard（27/33。’TissusVivants’1866page22。），hasitsproperlife，itsautonomy;itcandevelopandreproduceitselfindependentlyoftheadjoiningtissues。AgreatGermanauthority，Virchow（27/34。’CellularPathology’translatedbyDr。Chance1860pages14，18，83，460。），assertsstillmoreemphaticallythateachsystemconsistsofan\"enormousmassofminutecentresofaction……Everyelementhasitsownspecialaction，andeventhoughitderiveitsstimulustoactivityfromotherparts，yetaloneeffectstheactualperformanceofduties……Everysingleepithelialandmuscularfibre—
cellleadsasortofparasiticalexistenceinrelationtotherestofthebody……Everysinglebone—corpusclereallypossessesconditionsofnutritionpeculiartoitself。\"Eachelement，asSirJ。Pagetremarks，livesitsappointedtimeandthendies，andisreplacedafterbeingcastofforabsorbed。（27/35。Paget’SurgicalPathology’volume11853pages12—14。）I
presumethatnophysiologistdoubtsthat，forinstance，eachbone—corpuscleofthefingerdiffersfromthecorrespondingcorpuscleinthecorrespondingjointofthetoe;andtherecanhardlybeadoubtthateventhoseonthecorrespondingsidesofthebodydiffer，thoughalmostidenticalinnature。
Thisnearapproachtoidentityiscuriouslyshowninmanydiseasesinwhichthesameexactpointsontherightandleftsidesofthebodyaresimilarlyaffected;thusSirJ。Paget（27/36。Ibidpage19。）givesadrawingofadiseasedpelvis，inwhichthebonehasgrownintoamostcomplicatedpattern，but\"thereisnotonespotorlineononesidewhichisnotrepresented，asexactlyasitwouldbeinamirror，ontheother。\"
Manyfactssupportthisviewoftheindependentlifeofeachminuteelementofthebody。Virchowinsiststhatasinglebone—corpuscleorasinglecellintheskinmaybecomediseased。Thespurofacock，afterbeinginsertedintotheearofanox，livedforeightyears，andacquiredaweightof396grammes（nearlyfourteenounces），andtheastonishinglengthoftwenty—fourcentimetres，oraboutnineinches;sothattheheadoftheoxappearedtobearthreehorns。（27/37。SeeProf。Mantegazza’sinterestingwork’DegliinnestiAnimali’etc。Milano1865page51tab。3。）Thetailofapighasbeengraftedintothemiddleofitsback，andreacquiredsensibility。Dr。Ollier（27/38。
’DelaProductionArtificielledesOs’page8。）insertedapieceofperiosteumfromtheboneofayoungdogundertheskinofarabbit，andtruebonewasdeveloped。Amultitudeofsimilarfactscouldbegiven。Thefrequentpresenceofhairsandofperfectlydevelopedteeth，eventeethoftheseconddentition，inovariantumours（27/39。IsidoreGeoffroySaint—Hilaire’Hist。desAnomalies’tome2pages549，560，562;Virchowibidpage484。LawsonTait’ThePathologyofDiseasesoftheOvaries’1874pages61，62。），arefactsleadingtothesameconclusion。Mr。LawsonTaitreferstoatumourinwhich\"over300
teethwerefound，resemblinginmanyrespectsmilk—teeth;\"andtoanothertumour，\"fullofhairwhichhadgrownandbeenshedfromonelittlespotofskinnotbiggerthanthetipofmylittlefinger。Theamountofhairinthesac，haditgrownfromasimilarlysizedareaofthescalp，wouldhavetakenalmostalifetimetogrowandbeshed。\"
Whethereachoftheinnumerableautonomouselementsofthebodyisacellorthemodifiedproductofacell，isamoredoubtfulquestion，evenifsowideadefinitionbegiventotheterm，astoincludecell—likebodieswithoutwallsandwithoutnuclei。（27/40。Forthemostrecentclassificationofcells，seeErnstHackel’GenerelleMorpholog。’b。21866s。275。）Thedoctrineofomniscellulaecellulaisadmittedforplants，andwidelyprevailswithrespecttoanimals。（27/41。Dr。W。Turner’ThePresentAspectofCellularPathology’
’EdinburghMedicalJournal’April1863。）ThusVirchow，thegreatsupporterofthecellulartheory，whilstallowingthatdifficultiesexist，maintainsthateveryatomoftissueisderivedfromcells，andthesefrompre—existingcells，andtheseprimarilyfromtheegg，whichheregardsasagreatcell。Thatcells，stillretainingthesamenature，increasebyself—divisionorproliferation，isadmittedbyeveryone。Butwhenanorganismundergoesgreatchangesofstructureduringdevelopment，thecells，whichateachstagearesupposedtobedirectlyderivedfrompreviouslyexistingcells，mustlikewisebegreatlychangedinnature;thischangeisattributedbythesupportersofthecellulardoctrinetosomeinherentpowerwhichthecellspossess，andnottoanyexternalagency。Othersmaintainthatcellsandtissuesofallkindsmaybeformed，independentlyofpre—existingcells，fromplasticlymphorblastema。Whicheverviewmaybecorrect，everyoneadmitsthatthebodyconsistsofamultitudeoforganicunits，allofwhichpossesstheirownproperattributes，andaretoacertainextentindependentofallothers。
Henceitwillbeconvenienttouseindifferentlythetermscellsororganicunits，orsimplyunits。
VARIABILITYANDINHERITANCE。
Wehaveseeninthetwenty—secondchapterthatvariabilityisnotaprincipleco—ordinatewithlifeorreproduction，butresultsfromspecialcauses，generallyfromchangedconditionsactingduringsuccessivegenerations。Thefluctuatingvariabilitythusinducedisapparentlydueinparttothesexualsystembeingeasilyaffected，sothatitisoftenrenderedimpotent;andwhennotsoseriouslyaffected，itoftenfailsinitsproperfunctionoftransmittingtrulythecharactersoftheparentstotheoffspring。Butvariabilityisnotnecessarilyconnectedwiththesexualsystem，asweseeinthecasesofbud—variation。Althoughweareseldomabletotracethenatureoftheconnection，manydeviationsofstructurenodoubtresultfromchangedconditionsactingdirectlyontheorganisation，independentlyofthereproductivesystem。Insomeinstanceswemayfeelsureofthis，whenall，ornearlyalltheindividualswhichhavebeensimilarlyexposedaresimilarlyanddefinitelyaffected，ofwhichseveralinstanceshavebeengiven。Butitisbynomeansclearwhytheoffspringshouldbeaffectedbytheexposureoftheparentstonewconditions，andwhyitisnecessaryinmostcasesthatseveralgenerationsshouldhavebeenthusexposed。
How，again，canweexplaintheinheritedeffectsoftheuseordisuseofparticularorgans？Thedomesticatedduckflieslessandwalksmorethanthewildduck，anditslimb—boneshavebecomediminishedandincreasedinacorrespondingmannerincomparisonwiththoseofthewildduck。Ahorseistrainedtocertainpaces，andthecoltinheritssimilarconsensualmovements。
Thedomesticatedrabbitbecomestamefromcloseconfinement;thedog，intelligentfromassociatingwithman;theretrieveristaughttofetchandcarry;andthesementalendowmentsandbodilypowersareallinherited。
Nothinginthewholecircuitofphysiologyismorewonderful。Howcantheuseordisuseofaparticularlimborofthebrainaffectasmallaggregateofreproductivecells，seatedinadistantpartofthebody，insuchamannerthatthebeingdevelopedfromthesecellsinheritsthecharactersofeitheroneorbothparents？Evenanimperfectanswertothisquestionwouldbesatisfactory。
Inthechaptersdevotedtoinheritanceitwasshownthatamultitudeofnewlyacquiredcharacters，whetherinjuriousorbeneficial，whetherofthelowestorhighestvitalimportance，areoftenfaithfullytransmitted——frequentlyevenwhenoneparentalonepossessessomenewpeculiarity;andwemayonthewholeconcludethatinheritanceistherule，andnon—inheritancetheanomaly。Insomeinstancesacharacterisnotinherited，fromtheconditionsoflifebeingdirectlyopposedtoitsdevelopment;inmanyinstances，fromtheconditionsincessantlyinducingfreshvariability，aswithgraftedfruit—treesandhighly—cultivatedflowers。Intheremainingcasesthefailuremaybeattributedtoreversion，bywhichthechildresemblesitsgrandparentsormoreremoteprogenitors，insteadofitsparents。
Inheritanceisgovernedbyvariouslaws。Characterswhichfirstappearatanyparticularagetendtoreappearatacorrespondingage。Theyoftenbecomeassociatedwithcertainseasonsoftheyear，andreappearintheoffspringatacorrespondingseason。Iftheyappearratherlateinlifeinonesex，theytendtoreappearexclusivelyinthesamesexatthesameperiodoflife。
Theprincipleofreversion，recentlyalludedto，isoneofthemostwonderfuloftheattributesofInheritance。Itprovestousthatthetransmissionofacharacteranditsdevelopment，whichordinarilygotogetherandthusescapediscrimination，aredistinctpowers;andthesepowersinsomecasesareevenantagonistic，foreachactsalternatelyinsuccessivegenerations。Reversionisnotarareevent，dependingonsomeunusualorfavourablecombinationofcircumstances，butoccurssoregularlywithcrossedanimalsandplants，andsofrequentlywithuncrossedbreeds，thatitisevidentlyanessentialpartoftheprincipleofinheritance。Weknowthatchangedconditionshavethepowerofevokinglong—lostcharacters，asinthecaseofanimalsbecomingferal。Theactofcrossinginitselfpossessesthispowerinahighdegree。Whatcanbemorewonderfulthanthatcharacters，whichhavedisappearedduringscores，orhundreds，oreventhousandsofgenerations，shouldsuddenlyreappearperfectlydeveloped，asinthecaseofpigeonsandfowls，bothwhenpurelybredandespeciallywhencrossed;oraswiththezebrinestripesondun—colouredhorses，andothersuchcases？Manymonstrositiescomeunderthissamehead，aswhenrudimentaryorgansareredeveloped，orwhenanorganwhichwemustbelievewaspossessedbyanearlyprogenitorofthespecies，butofwhichnotevenarudimentisleft，suddenlyreappears，aswiththefifthstameninsomeScrophulariaceae。Wehavealreadyseenthatreversionactsinbud—
reproduction;andweknowthatitoccasionallyactsduringthegrowthofthesameindividualanimal，especially，butnotexclusively，ifofcrossedparentage，——asintherarecasesdescribedoffowls，pigeons，cattle，andrabbits，whichhaverevertedtothecoloursofoneoftheirparentsorancestorsastheyadvancedinyears。
Weareledtobelieve，asformerlyexplained，thateverycharacterwhichoccasionallyreappearsispresentinalatentformineachgeneration，innearlythesamemannerasinmaleandfemaleanimalsthesecondarycharactersoftheoppositesexlielatentandreadytobeevolvedwhenthereproductiveorgansareinjured。Thiscomparisonofthesecondarysexualcharacterswhichlielatentinbothsexes，withotherlatentcharacters，isthemoreappropriatefromthecaserecordedofaHen，whichassumedsomeofthemasculinecharacters，notofherownrace，butofanearlyprogenitor;shethusexhibitedatthesametimetheredevelopmentoflatentcharactersofbothkinds。Ineverylivingcreaturewemayfeelassuredthatahostoflong—lostcharactersliereadytobeevolvedunderproperconditions。Howcanwemakeintelligibleandconnectwithotherfacts，thiswonderfulandcommoncapacityofreversion，——thispowerofcallingbacktolifelong—lostcharacters？
PARTII。
Ihavenowenumeratedthechieffactswhicheveryonewoulddesiretoseeconnectedbysomeintelligiblebond。Thiscanbedone，ifwemakethefollowingassumptions，andmuchmaybeadvancedinfavourofthechiefone。
Thesecondaryassumptionscanlikewisebesupportedbyvariousphysiologicalconsiderations。Itisuniversallyadmittedthatthecellsorunitsofthebodyincreasebyself—divisionorproliferation，retainingthesamenature，andthattheyultimatelybecomeconvertedintothevarioustissuesandsubstancesofthebody。ButbesidesthismeansofincreaseIassumethattheunitsthrowoffminutegranuleswhicharedispersedthroughoutthewholesystem;thatthese，whensuppliedwithpropernutriment，multiplybyself—division，andareultimatelydevelopedintounitslikethosefromwhichtheywereoriginallyderived。Thesegranulesmaybecalledgemmules。Theyarecollectedfromallpartsofthesystemtoconstitutethesexualelements，andtheirdevelopmentinthenextgenerationformsanewbeing;buttheyarelikewisecapableoftransmissioninadormantstatetofuturegenerationsandmaythenbedeveloped。Theirdevelopmentdependsontheirunionwithotherpartiallydevelopedornascentcellswhichprecedethemintheregularcourseofgrowth。
WhyIusethetermunion，willbeseenwhenwediscussthedirectactionofpollenonthetissuesofthemother—plant。Gemmulesaresupposedtobethrownoffbyeveryunit，notonlyduringtheadultstate，butduringeachstageofdevelopmentofeveryorganism;butnotnecessarilyduringthecontinuedexistenceofthesameunit。Lastly，Iassumethatthegemmulesintheirdormantstatehaveamutualaffinityforeachother，leadingtotheiraggregationintobudsorintothesexualelements。Hence，itisnotthereproductiveorgansorbudswhichgenerateneworganisms，buttheunitsofwhicheachindividualiscomposed。TheseassumptionsconstitutetheprovisionalhypothesiswhichIhavecalledPangenesis。Viewsinmanyrespectssimilarhavebeenpropoundedbyvariousauthors。（27/42。Mr。G。H。Lewes（’FortnightlyReview’November1，1868page506）remarksonthenumberofwriterswhohaveadvancednearlysimilarviews。MorethantwothousandyearsagoAristotlecombatedaviewofthiskind，which，asIhearfromDr。W。Ogle，washeldbyHippocratesandothers。Ray，inhis’WisdomofGod’（2ndedition1692page68），saysthat\"everypartofthebodyseemstoclubandcontributetotheseed。\"The\"organicmolecules\"ofBuffon（’Hist。Nat。Gen。’editionof1749tome2pages54，62，329，333，420，425）appearatfirstsighttobethesameasthegemmulesofmyhypothesis，buttheyareessentiallydifferent。
Bonnet（’Oeuvresd’Hist。Nat。’tome5part117814toeditionpage334）speaksofthelimbshavinggermsadaptedforthereparationofallpossiblelosses;
butwhetherthesegermsaresupposedtobethesamewiththosewithinbudsandthesexualorgansisnotclear。Prof。Owensays（’AnatomyofVertebrates’
volume31868page813）thathefailstoseeanyfundamentaldifferencebetweentheviewswhichhepropoundedinhis’Parthenogenesis’（1849pages5—
8），andwhichhenowconsidersaserroneous，andmyhypothesisofpangenesis:
butareviewer（’JournalofAnat。andPhys。’May1869page441）showshowdifferenttheyreallyare。Iformerlythoughtthatthe\"physiologicalunits\"
ofHerbertSpencer（’PrinciplesofBiology’volume1chapters4and81863—64）
werethesameasmygemmules，butInowknowthatthisisnotthecase。
Lastly，itappearsfromareviewofthepresentworkbyProf。Mantegazza（’NuovaAntologia，Maggio’1868），thathe（inhis’ElementidiIgiene’Ediz。3
page540）clearlyforesawthedoctrineofpangenesis。）
Beforeproceedingtoshow，firstly，howfartheseassumptionsareinthemselvesprobable，andsecondly，howfartheyconnectandexplainthevariousgroupsoffactswithwhichweareconcerned，itmaybeusefultogiveanillustration，assimpleaspossible，ofthehypothesis。IfoneoftheProtozoabeformed，asitappearsunderthemicroscope，ofasmallmassofhomogeneousgelatinousmatter，aminuteparticleorgemmulethrownofffromanypartandnourishedunderfavourablecircumstanceswouldreproducethewhole;butiftheupperandlowersurfacesweretodifferintexturefromeachotherandfromthecentralportion，thenallthreepartswouldhavetothrowoffgemmules，whichwhenaggregatedbymutualaffinitywouldformeitherbudsorthesexualelements，andwouldultimatelybedevelopedintoasimilarorganism。Preciselythesameviewmaybeextendedtooneofthehigheranimals;althoughinthiscasemanythousandgemmulesmustbethrownofffromthevariouspartsofthebodyateachstageofdevelopment;thesegemmulesbeingdevelopedinunionwithpre—existingnascentcellsindueorderofsuccession。
Physiologistsmaintain，aswehaveseen，thateachunitofthebody，thoughtoalargeextentdependentonothers，islikewisetoacertainextentindependentorautonomous，andhasthepowerofincreasingbyself—division。I
goonestepfurther，andassumethateachunitcastsofffreegemmuleswhicharedispersedthroughoutthesystem，andarecapableunderproperconditionsofbeingdevelopedintosimilarunits。Norcanthisassumptionbeconsideredasgratuitousandimprobable。Itismanifestthatthesexualelementsandbudsincludeformativematterofsomekind，capableofdevelopment;andwenowknowfromtheproductionofgraft—hybridsthatsimilarmatterisdispersedthroughoutthetissuesofplants，andcancombinewiththatofanotheranddistinctplant，givingrisetoanewbeing，intermediateincharacter。Weknowalsothatthemaleelementcanactdirectlyonthepartiallydevelopedtissuesofthemother—plant，andonthefutureprogenyoffemaleanimals。Theformativematterwhichisthusdispersedthroughoutthetissuesofplants，andwhichiscapableofbeingdevelopedintoeachunitorpart，mustbegeneratedtherebysomemeans;andmychiefassumptionisthatthismatterconsistsofminuteparticlesorgemmulescastofffromeachunitorcell。（27/43。Mr。
Lownehasobserved（’JournalofQueckettMicroscopicalClub’September23，1870）certainremarkablechangesinthetissuesofthelarvaofafly，whichmakeshimbelieve\"itpossiblethatorgansandorganismsaresometimesdevelopedbytheaggregationofexcessivelyminutegemmules，suchasthosewhichMr。Darwin’shypothesisdemands。\"）
ButIhavefurthertoassumethatthegemmulesintheirundevelopedstatearecapableoflargelymultiplyingthemselvesbyself—division，likeindependentorganisms。Delpinoinsiststhatto\"admitofmultiplicationbyfissiparityincorpuscles，analogoustoseedsorbuds……isrepugnanttoallanalogy。\"Butthisseemsastrangeobjection，asThuret（27/44。’AnnalesdesSc。Nat。’3rdseriesBot。tome141850page244。）hasseenthezoosporeofanalgadivideitself，andeachhalfgerminated。Haeckeldividedthesegmentedovumofasiphonophoraintomanypieces，andtheseweredeveloped。Nordoestheextrememinutenessofthegemmules，whichcanhardlydiffermuchinnaturefromthelowestandsimplestorganisms，renderitimprobablethattheyshouldgrowandmultiply。Agreatauthority，Dr。Beale（27/45。’DiseaseGerms’page20。），says\"thatminuteyeastcellsarecapableofthrowingoffbudsorgemmules，muchlessthanthe1/100000ofaninchindiameter;\"andthesehethinksare\"capableofsubdivisionpracticallyadinfinitum。\"
Aparticleofsmall—poxmatter，sominuteastobebornebythewind，mustmultiplyitselfmanythousandfoldinapersonthusinoculated;andsowiththecontagiousmatterofscarletfever。（27/46。SeesomeveryinterestingpapersonthissubjectbyDr。Bealein’MedicalTimesandGazette’September9，1865
pages273，330。）Ithasrecentlybeenascertained（27/47。ThirdReportoftheR。Comm。ontheCattlePlagueasquotedin’Gardener’sChronicle’1866page446。）thataminuteportionofthemucousdischargefromananimalaffectedwithrinderpest，ifplacedinthebloodofahealthyox，increasessofastthatinashortspaceoftime\"thewholemassofblood，weighingmanypounds，isinfected，andeverysmallparticleofthatbloodcontainsenoughpoisontogive，withinlessthanforty—eighthours，thediseasetoanotheranimal。\"
Theretentionoffreeandundevelopedgemmulesinthesamebodyfromearlyyouthtooldagewillappearimprobable，butweshouldrememberhowlongseedsliedormantintheearthandbudsinthebarkofatree。Theirtransmissionfromgenerationtogenerationwillappearstillmoreimprobable;buthereagainweshouldrememberthatmanyrudimentaryanduselessorganshavebeentransmittedduringanindefinitenumberofgenerations。Weshallpresentlyseehowwellthelong—continuedtransmissionofundevelopedgemmulesexplainsmanyfacts。
Aseachunit，orgroupofsimilarunits，throughoutthebody，castsoffitsgemmules，andasallarecontainedwithinthesmallestovule，andwithineachspermatozoonorpollen—grain，andassomeanimalsandplantsproduceanastonishingnumberofpollen—grainsandovules（27/48。Mr。F。Bucklandfound6，867，840eggsinacod—fish（’LandandWater’1868page62）。AnAscarisproducesabout64，000，000eggs（Carpenter’s’Comp。Phys。’1854page590）。Mr。
J。Scott，oftheRoyalBotanicGardenofEdinburgh，calculated，inthesamemannerasIhavedoneforsomeBritishOrchids（’FertilisationofOrchids’
page344），thenumberofseedsinacapsuleofanAcroperaandfoundthenumbertobe371，250。Nowthisplantproducesseveralflowersonaraceme，andmanyracemesduringaseason。Inanalliedgenus，Gongora，Mr。Scotthasseentwentycapsulesproducedonasingleraceme;tensuchracemesontheAcroperawouldyieldaboveseventy—fourmillionsofseed。），thenumberandminutenessofthegemmulesmustbesomethinginconceivable。Butconsideringhowminutethemoleculesare，andhowmanygototheformationofthesmallestgranuleofanyordinarysubstance，thisdifficultywithrespecttothegemmulesisnotinsuperable。FromthedataarrivedatbySirW。Thomson，mysonGeorgefindsthatacubeof1/10000ofaninchofglassorwatermustconsistofbetween16
millionmillions，and131thousandmillionmillionmolecules。Nodoubtthemoleculesofwhichanorganismisformedarelarger，frombeingmorecomplex，thanthoseofaninorganicsubstance，andprobablymanymoleculesgototheformationofagemmule;butwhenwebearinmindthatacubeof1/10000ofaninchismuchsmallerthananypollen—grain，ovuleorbud，wecanseewhatavastnumberofgemmulesoneofthesebodiesmightcontain。
Thegemmulesderivedfromeachpartororganmustbethoroughlydispersedthroughoutthewholesystem。Weknow，forinstance，thatevenaminutefragmentofaleafofaBegoniawillreproducethewholeplant;andthatifafresh—waterwormischoppedintosmallpieces，eachwillreproducethewholeanimal。Consideringalsotheminutenessofthegemmulesandthepermeabilityofallorganictissues，thethoroughdispersionofthegemmulesisnotsurprising。Thatmattermaybereadilytransferredwithouttheaidofvesselsfromparttopartofthebody，wehaveagoodinstanceinacaserecordedbySirJ。Pagetofalady，whosehairlostitscolourateachsuccessiveattackofneuralgiaandrecovereditagaininthecourseofafewdays。Withplants，however，andprobablywithcompoundanimals，suchascorals，thegemmulesdonotordinarilyspreadfrombudtobud，butareconfinedtothepartsdevelopedfromeachseparatebud;andofthisfactnoexplanationcanbegiven。
Theassumedelectiveaffinityofeachgemmuleforthatparticularcellwhichprecedesitindueorderofdevelopmentissupportedbymanyanalogies。Inallordinarycasesofsexualreproduction，themaleandfemaleelementscertainlyhaveamutualaffinityforeachother:thus，itisbelievedthatabouttenthousandspeciesofCompositaeexist，andtherecanbenodoubtthatifthepollenofallthesespeciescouldbesimultaneouslyorsuccessivelyplacedonthestigmaofanyonespecies，thisonewouldelectwithunerringcertaintyitsownpollen。Thiselectivecapacityisallthemorewonderful，asitmusthavebeenacquiredsincethemanyspeciesofthisgreatgroupofplantsbranchedofffromacommonprogenitor。Onanyviewofthenatureofsexualreproduction，theformativematterofeachpartcontainedwithintheovulesandthemaleelementactoneachotherbysomelawofspecialaffinity，sothatcorrespondingpartsaffectoneanother;thus，acalfproducedfromashort—hornedcowbyalong—hornedbullhasitshornsaffectedbytheunionofthetwoforms，andtheoffspringfromtwobirdswithdifferentlycolouredtailshavetheirtailsaffected。
Thevarioustissuesofthebodyplainlyshow，asmanyphysiologistshaveinsisted（27/49。Paget’LecturesonPathology’page27;Virchow’CellularPathology’translatedbyDr。Chancepages123，126，294。ClaudeBernard’DesTissusVivants’pages177，210，337;Muller’Physiology’Englishtranslationpage290。），anaffinityforspecialorganicsubstances，whethernaturalorforeigntothebody。Weseethisinthecellsofthekidneysattractingureafromtheblood;incurareaffectingcertainnerves;Lyttavesicatoriathekidneys;andthepoisonousmatterofvariousdiseases，assmall—pox，scarlet—
fever，hooping—cough，glanders，andhydrophobia，affectingcertaindefinitepartsofthebody。Ithasalsobeenassumedthatthedevelopmentofeachgemmuledependsonitsunionwithanothercellorunitwhichhasjustcommenceditsdevelopment，andwhichprecedesitindueorderofgrowth。Thattheformativematterwithinthepollenofplants，whichbyourhypothesisconsistsofgemmules，canunitewithandmodifythepartiallydevelopedcellsofthemother—plant，wehaveclearlyseeninthesectiondevotedtothissubject。Asthetissuesofplantsareformed，asfarasisknown，onlybytheproliferationofpre—existingcells，wemustconcludethatthegemmulesderivedfromtheforeignpollendonotbecomedevelopedintonewandseparatecells，butpenetrateandmodifythenascentcellsofthemother—plant。Thisprocessmaybecomparedwithwhattakesplaceintheactofordinaryfertilisation，duringwhichthecontentsofthepollen—tubespenetratetheclosedembryonicsacwithintheovule，anddeterminethedevelopmentoftheembryo。Accordingtothisview，thecellsofthemother—plantmayalmostliterallybesaidtobefertilisedbythegemmulesderivedfromtheforeignpollen。Inthiscaseandinallothersthepropergemmulesmustcombineindueorderwithpre—existingnascentcells，owingtotheirelectiveaffinities。A
slightdifferenceinnaturebetweenthegemmulesandthenascentcellswouldbefarfrominterferingwiththeirmutualunionanddevelopment，forwewellknowinthecaseofordinaryreproductionthatsuchslightdifferentiationinthesexualelementsfavoursinamarkedmannertheirunionandsubsequentdevelopment，aswellasthevigouroftheoffspringthusproduced。
Thusfarwehavebeenablebytheaidofourhypothesistothrowsomeobscurelightontheproblemswhichhavecomebeforeus;butitmustbeconfessedthatmanypointsremainaltogetherdoubtful。Thusitisuselesstospeculateatwhatperiodofdevelopmenteachunitofthebodycastsoffitsgemmules，asthewholesubjectofthedevelopmentofthevarioustissuesisasyetfarfromclear。Wedonotknowwhetherthegemmulesaremerelycollectedbysomeunknownmeansatcertainseasonswithinthereproductiveorgans，orwhetherafterbeingthuscollectedtheyrapidlymultiplythere，astheflowofbloodtotheseorgansateachbreedingseasonseemstorenderprobable。Nordoweknowwhythegemmulescollecttoformbudsincertaindefiniteplaces，leadingtothesymmetricalgrowthoftreesandcorals。Wehavenomeansofdecidingwhethertheordinarywearandtearofthetissuesismadegoodbymeansofgemmules，ormerelybytheproliferationofpre—existingcells。Ifthegemmulesarethusconsumed，asseemsprobablefromtheintimateconnectionbetweentherepairofwaste，regrowth，anddevelopment，andmoreespeciallyfromtheperiodicalchangeswhichmanymaleanimalsundergoincolourandstructure，thensomelightwouldbethrownonthephenomenaofoldage，withitslessenedpowerofreproductionandoftherepairofinjuries，andontheobscuresubjectoflongevity。Thefactofcastratedanimals，whichdonotcastoffinnumerablegemmulesintheactofreproduction，notbeinglonger—livedthanperfectmales，seemsopposedtothebeliefthatgemmulesareconsumedintheordinaryrepairofwastedtissues;unlessindeedthegemmulesafterbeingcollectedinsmallnumberswithinthereproductiveorgansaretherelargelymultiplied。（27/50。Prof。RayLankesterhasdiscussedseveralofthepointsherereferredtoasbearingonpangenesis，inhisinterestingessay，’OnComparativeLongevityinManandtheLowerAnimals’1870pages33，77，etc。）
Thatthesamecellsorunitsmayliveforalongperiodandcontinuemultiplyingwithoutbeingmodifiedbytheirunionwithfreegemmulesofanykind，isprobablefromsuchcasesasthatofthespurofacockwhichgrewtoanenormoussizewhengraftedintotheearofanox。Howfarunitsaremodifiedduringtheirnormalgrowthbyabsorbingpeculiarnutrimentfromthesurroundingtissues，independentlyoftheirunionwithgemmulesofadistinctnature，isanotherdoubtfulpoint。（27/51。Dr。Rossreferstothissubjectinhis’GraftTheoryofDisease’1872page53。）Weshallappreciatethisdifficultybycallingtomindwhatcomplexyetsymmetricalgrowthsthecellsofplantsyieldwheninoculatedbythepoisonofagall—insect。Withanimalsvariouspolypoidexcrescencesandtumoursaregenerallyadmitted（27/52。
Virchow’CellularPathology’translatedbyDr。Chance1860pages60，162，245，441，454。）tobethedirectproduct，throughproliferation，ofnormalcellswhichhavebecomeabnormal。Intheregulargrowthandrepairofbones，thetissuesundergo，asVirchowremarks（27/53。Ibidpages412—426。），awholeseriesofpermutationsandsubstitutions。\"Thecartilagecellsmaybeconvertedbyadirecttransformationintomarrow—cells，andcontinueassuch;
ortheymayfirstbeconvertedintoosseousandthenintomedullarytissue;orlastly，theymayfirstbeconvertedintomarrowandthenintobone。Sovariablearethepermutationsofthesetissues，inthemselvessonearlyallied，andyetintheirexternalappearancesocompletelydistinct。\"Butasthesetissuesthuschangetheirnatureatanyage，withoutanyobviouschangeintheirnutrition，wemustsupposeinaccordancewithourhypothesisthatgemmulesderivedfromonekindoftissuecombinewiththecellsofanotherkind，andcausethesuccessivemodifications。
Wehavegoodreasontobelievethatseveralgemmulesarerequisiteforthedevelopmentofoneandthesameunitorcell;forwecannototherwiseunderstandtheinsufficiencyofasingleorevenoftwoorthreepollen—grainsorspermatozoa。Butwearefarfromknowingwhetherthegemmulesofalltheunitsarefreeandseparatefromoneanother，orwhethersomearefromthefirstunitedintosmallaggregates。Afeather，forinstance，isacomplexstructure，and，aseachseparatepartisliabletoinheritedvariations，I
concludethateachfeathergeneratesalargenumberofgemmules;butitispossiblethatthesemaybeaggregatedintoacompoundgemmule。Thesameremarkappliestothepetalsofflowers，whicharesometimeshighlycomplexstructures，witheachridgeandhollowcontrivedforaspecialpurpose，sothateachpartmusthavebeenseparatelymodified，andthemodificationstransmitted;consequently，separategemmules，accordingtoourhypothesis，musthavebeenthrownofffromeachcellorunit。But，aswesometimesseehalfanantherorasmallportionofafilamentbecomingpetali—form，orpartsormerestripesofthecalyxassumingthecolourandtextureofthecorolla，itisprobablethatwithpetalsthegemmulesofeachcellarenotaggregatedtogetherintoacompoundgemmule，butarefreeandseparate。Eveninsosimpleacaseasthatofaperfectcell，withitsprotoplasmiccontents，nucleus，nucleolus，andwalls，wedonotknowwhetherornotitsdevelopmentdependsonacompoundgemmulederivedfromeachpart。（27/54。SeesomegoodcriticismsonthisheadbyDelpinoandbyMr。G。H。Lewesinthe’FortnightlyReview’
November1，1868page509。）