Thesesameflowersalsovariedmuchinthenumberofthepetals，andoccasionallyinthenumberofthestamensandpistils;sothattheyweresemi—monstrousinstructure，yettheyproducedplentyoffruit。Mr。
ThompsonremarksthatinthePastimegooseberry"extrabractsareoftenattachedtothesidesofthefruit。"（10/124。’CatalogueofFruitsofHort。
Soc。Garden’3rdedition1842。）
Themostinterestingpointinthehistoryofthegooseberryisthesteadyincreaseinthesizeofthefruit。Manchesteristhemetropolisofthefanciers，andprizesfromfiveshillingstofiveortenpoundsareyearlygivenfortheheaviestfruit。The’GooseberryGrowersRegister’ispublishedannually;theearliestknowncopyisdated1786，butitiscertainthatmeetingsfortheadjudicationofprizeswereheldsomeyearspreviously。（10/125。Mr。ClarksonofManchesterontheCultureoftheGooseberryinLoudon’s’Gardener’sMagazine’volume41828page482。）The’Register’for1845givesanaccountof171GooseberryShows，heldindifferentplacesduringthatyear;andthisfactshowsonhowlargeascaletheculturehasbeencarriedon。Thefruitofthewildgooseberryissaid（10/126。Downing’FruitsofAmerica’page213。）toweighaboutaquarterofanounceor5dwts。，thatis，120grains;abouttheyear1786gooseberrieswereexhibitedweighing10dwts。，sothattheweightwasthendoubled;in181726dwts。17grs。wasattained;therewasnoadvancetill1825，when31
dwts。16grs。wasreached;in1830"Teazer"weighed32dwts。13grs。;in1841"Wonderful"weighed32dwts。16grs。;in1844"London"weighed35
dwts。12grs。，andinthefollowingyear36dwts。16grs。;andin1852inStaffordshire，thefruitofthesamevarietyreachedtheastonishingweightof37dwts。7grs。（10/127。’Gardener’sChronicle’1844page811whereatableisgiven;and1845page819。Fortheextremeweightsgainedsee’JournalofHorticulture’July26，1864page61。）or896grs。;thatis，betweensevenoreighttimestheweightofthewildfruit。Ifindthatasmallapple，61/2inchesincircumference，hasexactlythissameweight。
The"London"gooseberry（whichin1852hadaltogethergained333prizes）
has，uptothepresentyearof1875，neverreachedagreaterweightthanthatattainedin1852。Perhapsthefruitofthegooseberryhasnowreachedthegreatestpossibleweight，unlessinthecourseoftimesomenewanddistinctvarietyshallarise。
Thisgradual，andonthewholesteadyincreaseofweightfromthelatterpartofthelastcenturytotheyear1852，isprobablyinlargepartduetoimprovedmethodsofcultivation，forextremecareisnowtaken;thebranchesandrootsaretrained，compostsaremade，thesoilismulched，andonlyafewberriesareleftoneachbush（10/128。Mr。SaulofLancasterinLoudon’s’Gardener’sMag。’volume31828page421;andvolume101834page42。）buttheincreasenodoubtisinmainpartduetothecontinuedselectionofseedlingswhichhavebeenfoundtobemoreandmorecapableofyieldingsuchextraordinaryfruit。Assuredlythe"Highwayman"in1817couldnothaveproducedfruitlikethatofthe"RoaringLion"in1825;norcouldthe"RoaringLion，"thoughitwasgrownbymanypersonsinmanyplaces，gainthesupremetriumphachievedin1852bythe"London"Gooseberry。
WALNUT（Juglansregia）。
Thistreeandthecommonnutbelongtoawidelydifferentorderfromtheforegoingfruits，andarethereforeherenoticed。ThewalnutgrowswildontheCaucasusandintheHimalaya，whereDr。Hooker（10/129。’HimalayanJournals’1854volume2page334。Moorcroft’Travels’volume2page146
describesfourvarietiescultivatedinKashmir。）foundthefruitoffullsize，but"ashardasahickory—nut。"Ithasbeenfoundfossil，asM。deSaportainformsme，inthetertiaryformation，ofFrance。
InEnglandthewalnutpresentsconsiderabledifferences，intheshapeandsizeofthefruit，inthethicknessofthehusk，andinthethinnessoftheshell;thislatterqualityhasgivenrisetoavarietycalledthethin—
shelled，whichisvaluable，butsuffersfromtheattacksoftitmice。
（10/130。’Gardener’sChronicle’1850page723。）Thedegreetowhichthekernelfillstheshellvariesmuch。InFrancethereisavarietycalledtheGrapeorcluster—walnut，inwhichthenutsgrowin"bunchesoften，fifteen，oreventwentytogether。"Thereisanothervarietywhichbearsonthesametreedifferentlyshapedleaves，liketheheterophylloushornbeam;
thistreeisalsoremarkablefromhavingpendulousbranches，andbearingelongated，large，thin—shellednuts。（10/131。PapertranslatedinLoudon’s’Gardener’sMag。’1829volume5page202。）M。Cardanhasminutelydescribed（10/132。Quotedin’Gardener’sChronicle’1849page101。）somesingularphysiologicalpeculiaritiesintheJune—leafingvariety，whichproducesitsleavesandflowersfourorfiveweekslaterthanthecommonvarieties;andalthoughinAugustitisapparentlyinexactlythesamestateofforwardnessastheotherkinds，itretainsitsleavesandfruitmuchlaterintheautumn。Theseconstitutionalpeculiaritiesarestrictlyinherited。
Lastly，walnut—trees，whichareproperlymonoicous，sometimesentirelyfailtoproducemaleflowers。（10/133。’Gardener’sChronicle’1847pages541and558。）
NUTS（Corylusavellana）。
Mostbotanistsrankallthevarietiesunderthesamespecies，thecommonwildnut。（10/134。Thefollowingdetailsaretakenfromthe’CatalogueofFruits，1842inGardenofHort。Soc。’page103;andfromLoudon’s’Encyclop。ofGardening’page943。）Thehusk，orinvolucre，differsgreatly，beingextremelyshortinBarr’sSpanish，andextremelylonginfilberts，inwhichitiscontractedsoastopreventthenutfallingout。
Thiskindofhuskalsoprotectsthenutfrombirds，fortitmice（Parus）
havebeenobserved（10/135。’Gardener’sChron。’1860page956。）topassoverfilberts，andattackcobsandcommonnutsgrowinginthesameorchard。
Inthepurple—filbertthehuskispurple，andinthefrizzled—filbertitiscuriouslylaciniated;inthered—filbertthepellicleofthekernelisred。
Theshellisthickinsomevarieties，butisthininCosford’s—nut，andinonevarietyisofabluishcolour。Thenutitselfdiffersmuchinsizeandshape，beingovateandcompressedinfilberts，nearlyroundandofgreatsizeincobsandSpanishnuts，oblongandlongitudinallystriatedinCosford’s，andobtuselyfour—sidedintheDowntonSquarenut。
CUCURBITACEOUSPLANTS。
Theseplantshavebeenforalongperiodtheopprobriumofbotanists;
numerousvarietieshavebeenrankedasspecies，and，whathappensmorerarely，formswhichnowmustbeconsideredasspecieshavebeenclassedasvarieties。Owingtotheadmirableexperimentalresearchesofadistinguishedbotanist，M。Naudin（10/136。’AnnalesdesSc。Nat。Bot。’4thseriesvolume61856page5。），afloodoflighthasrecentlybeenthrownonthisgroupofplants。M。Naudin，duringmanyyears，observedandexperimentedonabove1200livingspecimens，collectedfromallquartersoftheworld。SixspeciesarenowrecognisedinthegenusCucurbita;butthreealonehavebeencultivatedandconcernus，namely，C。maximaandpepo，whichincludeallpumpkins，gourds，squashes，andthevegetablemarrow，andC。moschata。Thesethreespeciesarenotknowninawildstate;butAsaGray（10/137。’AmericanJourn。ofScience’2ndseriesvolume241857page442。）givesgoodreasonforbelievingthatsomepumpkinsarenativesofN。
America。
Thesethreespeciesarecloselyallied，andhavethesamegeneralhabit，buttheirinnumerablevarietiescanalwaysbedistinguished，accordingtoNaudin，bycertainalmostfixedcharacters;andwhatisstillmoreimportant，whencrossedtheyyieldnoseed，oronlysterileseed;whilstthevarietiesspontaneouslyintercrosswiththeutmostfreedom。Naudininsistsstrongly（page15），that，thoughthesethreespecieshavevariedgreatlyinmanycharacters，yetithasbeeninsocloselyananalogousmannerthatthevarietiescanhearrangedinalmostparallelseries，aswehaveseenwiththeformsofwheat，withthetwomainracesofthepeach，andinothercases。Thoughsomeofthevarietiesareinconstantincharacter，yetothers，whengrownseparatelyunderuniformconditionsoflife，are，asNaudinrepeatedly（pages6，16，35）urges，"doueesd’unestabilitepresquecomparableacelledesespeceslesmieuxcaracterisees。"
Onevariety，l’Orangin（pages43，63），hassuchprepotencyintransmittingitscharacter，thatwhencrossedwithothervarietiesavastmajorityoftheseedlingscometrue。Naudin，referring（page47）toC。pepo，saysthatitsraces"nedifferentdesespecesveritablesqu’encequ’ellespeuvents’allierlesunesauxautresparvoied’hybridite，sansqueleurdescendanceperdelafacultedeseperpetuer。"Ifweweretotrusttoexternaldifferencesalone，andgiveupthetestofsterility，amultitudeofspecieswouldhavetobeformedoutofthevarietiesofthesethreespeciesofCucurbita。Manynaturalistsatthepresentdaylayfartoolittlestress，inmyopinion，onthetestofsterility;yetitisnotimprobablethatdistinctspeciesofplantsafteralongcourseofcultivationandvariationmayhavetheirmutualsterilityeliminated，aswehaveeveryreasontobelievehasoccurredwithdomesticatedanimals。Nor，inthecaseofplantsundercultivation，shouldwebejustifiedinassumingthatvarietiesneveracquireaslightdegreeofmutualsterility，asweshallmorefullyseeinafuturechapterwhencertainfactsaregivenonthehighauthorityofGartnerandKolreuter。（10/138。Gartner’Bastarderzeugung’1849s。87ands。169withrespecttoMaize;onVerbascumibids。92and181;alsohis’KenntnissderBefruchtung’s。137。
WithrespecttoNicotianaseeKolreuter’ZweiteForts。’1764s。53;thoughthisisasomewhatdifferentcase。）
TheformsofC。pepoareclassedbyNaudinundersevensections，eachincludingsubordinatevarieties。Heconsidersthisplantasprobablythemostvariableintheworld。Thefruitofonevariety（pages33，46）exceedsinvaluethatofanotherbymorethantwothousandfold！Whenthefruitisofverylargesize，thenumberproducedisfew（page45）;whenofsmallsize，manyareproduced。Nolessastonishing（page33）isthevariationintheshapeofthefruit，thetypicalformapparentlyisegg—like，butthisbecomeseitherdrawnoutintoacylinder，orshortenedintoaflatdisc。Wehavealsoanalmostinfinitediversityinthecolourandstateofsurfaceofthefruit，inthehardnessbothoftheshellandoftheflesh，andinthetasteoftheflesh，whichiseitherextremelysweet，farinaceous，orslightlybitter。Theseedsalsodifferinaslightdegreeinshape，andwonderfullyinsize（page34），namely，fromsixorseventomorethantwenty—fivemillimetresinlength。
Inthevarietieswhichgrowuprightordonotrunandclimb，thetendrils，thoughuseless（page31），areeitherpresentorarerepresentedbyvarioussemi—monstrousorgans，orarequiteabsent。Thetendrilsareevenabsentinsomerunningvarietiesinwhichthestemsaremuchelongated。Itisasingularfactthat（page31）inallthevarietieswithdwarfedstems，theleavescloselyresembleeachotherinshape。]
Thosenaturalistswhobelieveintheimmutabilityofspeciesoftenmaintainthat，eveninthemostvariableforms，thecharacterswhichtheyconsiderofspecificvalueareunchangeable。Togiveanexamplefromaconscientiouswriter（10/139。Godron’Del’Espece’tome2page64。）who，relyingonthelaboursofM。Naudin，andreferringtothespeciesofCucurbita，says，"aumilieudetouteslesvariationsdufruit，lestiges，lesfeuilles，lescalices，lescorolles，lesetaminesrestentinvariablesdanschacuned’elles。"YetM。Naudin，indescribingCucurbitapepo（page30），says，"Ici，d’ailleurs，cenesontpasseulementlesfruitsquivarient，c’estaussilefeuillageettoutleportdelaplante。Neanmoins，jecroisqu’onladistingueratoujoursfacilementdesdeuxautresespeces，sil’onveutnepasperdredevuelescaracteresdifferentielsquejem’efforcedefaireressortir。Cescaracteressontquelquefoispeumarques:ilarrivememequeplusieursd’entreeuxs’effacentpresqueentierement，maisiienrestetoujoursquelques—unsquiremettentl’observateursurlavoie。"Nowletitbenotedwhatadifference，withregardtotheimmutabilityoftheso—
calledspecificcharactersthisparagraphproducesonthemind，fromthatabovequotedfromM。Godron。
Iwilladdanotherremark:naturalistscontinuallyassertthatnoimportantorganvaries;butinsayingthistheyunconsciouslyargueinaviciouscircle;forifanorgan，letitbewhatitmay，ishighlyvariable，itisregardedasunimportant，andunderasystematicpointofviewthisisquitecorrect。Butaslongasconstancyisthustakenasthecriterionofimportance，itwillindeedbelongbeforeanimportantorgancanbeshowntobeinconstant。Theenlargedformofthestigmas，andtheirsessilepositiononthesummitoftheovary，mustbeconsideredasimportantcharacters，andwereusedbyGasparinitoseparatecertainpumpkinsasaDISTINCTGENUS;butNaudinsays（page20），thesepartshavenoconstancy，andintheflowersoftheTurbanvarietiesofC。maximatheysometimesresumetheirordinarystructure。Again，inC。maxima，thecarpels（page19）
whichformtheturbanprojectevenasmuchastwo—thirdsoftheirlengthoutofthereceptacle，andthislatterpartisthusreducedtoasortofplatform;butthisremarkablestructureoccursonlyincertainvarieties，andgraduatesintothecommonforminwhichthecarpelsarealmostentirelyenvelopedwithinthereceptacle。InC。moschatatheovarium（page50）
variesgreatlyinshape，beingoval，nearlyspherical，orcylindrical，moreorlessswollenintheupperpart，orconstrictedroundthemiddle，andeitherstraightorcurved。WhentheovariumisshortandovaltheinteriorstructuredoesnotdifferfromthatofC。maximaandpepo，butwhenitiselongatedthecarpelsoccupyonlytheterminalandswollenportion。Imayaddthatinonevarietyofthecucumber（Cucumissativus）thefruitregularlycontainsfivecarpelsinsteadofthree。（10/140。Naudin’Annal。
desSc。Nat。’4thseriesBot。tome111859page28。）Ipresumethatitwillnotbedisputedthatweherehaveinstancesofgreatvariabilityinorgansofthehighestphysiologicalimportance，andwithmostplantsofthehighestclassificatoryimportance。
[Sageret（10/141。’MemoiresurlesCucurbitacees’1826pages6，24。）andNaudinfoundthatthecucumber（C。sativus）couldnotbecrossedwithanyotherspeciesofthegenus;thereforenodoubtitisspecificallydistinctfromthemelon。Thiswillappeartomostpersonsasuperfluousstatement;
yetwehearfromNaudin（10/142。’FloredesSerres’October1861quotedin’Gardener’sChronicle’1861page1135。IhaveoftenconsultedandtakensomefactsfromM。Naudin’sMemoironCucumisin’Annal。desSc。Nat。’4thseriesBot。tome111859page5。）thatthereisaraceofmelons，inwhichthefruitissolikethatofthecucumber，"bothexternallyandinternally，thatitishardlypossibletodistinguishtheonefromtheotherexceptbytheleaves。"Thevarietiesofthemelonseemtobeendless，forNaudinaftersixyears’studyhadnotcometotheendofthem:hedividesthemintotensections，includingnumeroussub—varietieswhichallintercrosswithperfectease。（10/143。SeealsoSageret’Memoire’page7。）OftheformsconsideredbyNaudintobevarieties，botanistshavemadethirtydistinctspecies！"andtheyhadnottheslightestacquaintancewiththemultitudeofnewformswhichhaveappearedsincetheirtime。"Noristhecreationofsomanyspeciesatallsurprisingwhenweconsiderhowstrictlytheircharactersaretransmittedbyseed，andhowwonderfullytheydifferinappearance:"Miraestquidemfoliorumethabitusdiversitas，sedmultomagisfructuum，"saysNaudin。Thefruitisthevaluablepart，andthis，inaccordancewiththecommonrule，isthemostmodifiedpart。Somemelonsareonlyaslargeassmallplums，othersweighasmuchassixty—sixpounds。Onevarietyhasascarletfruit！Anotherisnotmorethananinchindiameter，butsometimesmorethanayardinlength，"twistingaboutinalldirectionslikeaserpent。"Itisasingularfactthatinthislattervarietymanypartsoftheplant，namely，thestems，thefootstalksofthefemaleflowers，themiddlelobeoftheleaves，andespeciallytheovarium，aswellasthematurefruit，allshowastrongtendencytobecomeelongated。
Severalvarietiesofthemelonareinterestingfromassumingthecharacteristicfeaturesofdistinctspeciesandevenofdistinctthoughalliedgenera:thustheserpent—melonhassomeresemblancetothefruitofTrichosanthesanguina;wehaveseenthatothervarietiescloselyresemblecucumbers;someEgyptianvarietieshavetheirseedsattachedtoaportionofthepulp，andthisischaracteristicofcertainwildforms。Lastly，avarietyofmelonfromAlgiersisremarkablefromannouncingitsmaturityby"aspontaneousandalmostsuddendislocation，"whendeepcrackssuddenlyappear，andthefruitfallstopieces;andthisoccurswiththewildC。
momordica。Finally，M。Naudinwellremarksthatthis"extraordinaryproductionofracesandvarietiesbyasinglespeciesandtheirpermanencewhennotinterferedwithbycrossing，arephenomenawellcalculatedtocausereflection。"
USEFULANDORNAMENTALTREES。
Treesdeserveapassingnoticeonaccountofthenumerousvarietieswhichtheypresent，differingintheirprecocity，intheirmannerofgrowth，theirfoliage，andbark。Thusofthecommonash（Fraxinusexcelsior）thecatalogueofMessrs。LawsonofEdinburghincludestwenty—onevarieties，someofwhichdiffermuchintheirbark;thereisayellow，astreakedreddish—white，apurple，awart—barkedandafungous—barkedvariety。
（10/144。Loudon’s’ArboretumetFruticetum’volume2page1217。）Ofholliesnolessthaneighty—fourvarietiesaregrownalongsideeachotherinMr。
Paul’snursery。（10/145。’Gardener’sChronicle’1866page1096。）Inthecaseoftrees，alltherecordedvarieties，asfarasIcanfindout，havebeensuddenlyproducedbyonesingleactofvariation。Thelengthoftimerequiredtoraisemanygenerations，andthelittlevaluesetonthefancifulvarieties，explainshowitisthatsuccessivemodificationshavenotbeenaccumulatedbyselection;hence，also，itfollowsthatwedonotheremeetwithsub—varietiessubordinatetovarieties，andtheseagainsubordinatetohighergroups。OntheContinent，however，wheretheforestsaremorecarefullyattendedtothaninEngland，Alph。DeCandolle（10/146。
’Geograph。Bot。’page1096。）saysthatthereisnotaforesterwhodoesnotsearchforseedsfromthatvarietywhichheesteemsthemostvaluable。
Ourusefultreeshaveseldombeenexposedtoanygreatchangeofconditions;theyhavenotbeenrichlymanured，andtheEnglishkindsgrowundertheirproperclimate。Yetinexaminingextensivebedsofseedlingsinnursery—gardensconsiderabledifferencesmaybegenerallyobservedinthem;
andwhilsttouringinEnglandIhavebeensurprisedattheamountofdifferenceintheappearanceofthesamespeciesinourhedgerowsandwoods。Butasplantsvarysomuchinatrulywildstate，itwouldbedifficultforevenaskilfulbotanisttopronouncewhether，asIbelievetobethecase，hedgerowtreesvarymorethanthosegrowinginaprimevalforest。Treeswhenplantedbymaninwoodsorhedgesdonotgrowwheretheywouldnaturallybeabletoholdtheirplaceagainstahostofcompetitors，andarethereforeexposedtoconditionsnotstrictlynatural:eventhisslightchangewouldprobablysufficetocauseseedlingsraisedfromsuchtreestobevariable。Whetherornotourhalf—wildEnglishtrees，asageneralrule，aremorevariablethantreesgrowingintheirnativeforests，therecanhardlybeadoubtthattheyhaveyieldedagreaternumberofstrongly—markedandsingularvariationsofstructure。
Inmannerofgrowth，wehaveweepingorpendulousvarietiesofthewillow，ash，elm，oak，andyew，andothertrees;andthisweepinghabitissometimesinherited，thoughinasingularlycapriciousmanner。IntheLombardypoplar，andincertainfastigiateorpyramidalvarietiesofthorns，junipers，oaks，etc。，wehaveanoppositekindofgrowth。TheHessianoak（10/147。’Gardener’sChronicle’1842page36。），whichisfamousfromitsfastigiatehabitandsize，bearshardlyanyresemblanceingeneralappearancetoacommonoak;"itsacornsarenotsuretoproduceplantsofthesamehabit;some，however，turnoutthesameastheparent—tree。"
AnotherfastigiateoakissaidtohavebeenfoundwildinthePyrenees，andthisisasurprisingcircumstance;itgenerallycomessotruebyseed，thatDeCandolleconsidereditasspecificallydistinct。（10/148。Loudon’s’ArboretumetFruticetum’volume3page1731。）ThefastigiateJuniper（J。
suecica）likewisetransmitsitscharacterbyseed。（10/149。Ibidvolume4
page2489。）Dr。FalconerinformsmethatintheBotanicGardensatCalcuttathegreatheatcausedapple—treestobecomefastigiate;andwethusseethesameresultfollowingfromtheeffectsofclimateandfromsomeunknowncause。（10/150。Godron’Del’Espece’tome2page91describesfourvarietiesofRobiniaremarkablefromtheirmannerofgrowth。）
Infoliagewehavevariegatedleaveswhichareofteninherited;darkpurpleorredleaves，asinthehazel，barberry，andbeech，thecolourinthesetwolattertreesbeingsometimesstronglyandsometimesweaklyinherited（10/151。’JournalofaHorticulturalTour，byCaledonianHort。Soc。’1823
page107。Alph。DeCandolle’Geograph。Bot。’page1083。Verlot’SurLaProductiondesVarietes’1865page55fortheBarberry。）;deeply—cutleaves;andleavescoveredwithprickles，asinthevarietyofthehollywellcalledferox，whichissaidtoreproduceitselfbyseed。（10/152。
Loudon’s’ArboretumetFruticetum’volume2page508。）Infact，nearlyallthepeculiarvarietiesevinceatendency，moreorlessstronglymarked，toreproducethemselvesbyseed。（10/153。Verlot’DesVarietes’1865page92。）
Thisistoacertainextentthecase，accordingtoBosc（10/154。Loudon’s’ArboretumetFruticetum’volume3page1376。），withthreevarietiesoftheelm，namely，thebroad—leafed，lime—leafed，andtwistedelm，inwhichlatterthefibresofthewoodaretwisted。Evenwiththeheterophylloushornbeam（Carpinusbetulus），whichbearsoneachtwigleavesoftwoshapes，"severalplantsraisedfromseedallretained"thesamepeculiarity。"
（10/155。’Gardener’sChronicle’1841page687。）Iwilladdonlyoneotherremarkablecaseofvariationinfoliage，namely，theoccurrenceoftwosub—
varietiesoftheashwithsimpleinsteadofpinnatedleaves，andwhichgenerallytransmittheircharacterbyseed。（10/156。Godron’Del’Espece’
tome2page89。InLoudon’s’Gardener’sMag。’volume121836page371，avariegatedbushyashisdescribedandfigured，ashavingsimpleleaves;itoriginatedinIreland。）Theoccurrence，intreesbelongingtowidelydifferentorders，ofweepingandfastigiatevarieties，andoftreesbearingdeeplycut，variegated，andpurpleleaves，showsthatthesedeviationsofstructuremustresultfromsomeverygeneralphysiologicallaws。
Differencesingeneralappearanceandfoliage，notmorestronglymarkedthanthoseaboveindicated，haveledgoodobserverstorankasdistinctspeciescertainformswhicharenowknowntobemerevarieties。Thus，aplane—treelongcultivatedinEnglandwasconsideredbyalmosteveryoneasaNorthAmericanspecies:butisnowascertainedbyoldrecords，asIaminformedbyDr。Hooker，tobeavariety。So，again，theThujapendulaorfiliformiswasrankedbysuchgoodobserversasLambert，Wallich，andothers，asatruespecies;butitisnowknownthattheoriginalplants，fiveinnumber，suddenlyappearedinabedofseedlings，raisedatMr。
Loddige’snursery，fromT。orientalis;andDr。HookerhasadducedexcellentevidencethatatTurinseedsofT。pendulahavereproducedtheparentform，T。orientalis。（10/157。’Gardener’sChronicle’1863page575。）
Everyonemusthavenoticedhowcertainindividualtreesregularlyputforthandshedtheirleavesearlierorlaterthanothersofthesamespecies。Thereisafamoushorse—chestnutintheTuilerieswhichisnamedfromleafingsomuchearlierthantheothers。ThereisalsoanoaknearEdinburghwhichretainsitsleavestoaverylateperiod。Thesedifferenceshavebeenattributedbysomeauthorstothenatureofthesoilinwhichthetreesgrow;butArchbishopWhatelygraftedanearlythornonalateone，andviceversa，andbothgraftskepttotheirproperperiods，whichdifferedbyaboutafortnight，asiftheystillgrewontheirownstocks。
（10/158。QuotedfromRoyalIrishAcademyin’Gardener’sChronicle’1841
page767。）ThereisaCornishvarietyoftheelmwhichisalmostanevergreen，andissotenderthattheshootsareoftenkilledbythefrost;
andthevarietiesoftheTurkishoak（Q。cerris）maybearrangedasdeciduous，sub—evergreen，andevergreen。（10/159。Loudon’ArboretumetFruticetum:’forElmseevolume3page1376;forOakpage1846。）
SCOTCHFIR（Pinussylvestris）。
Ialludetothistreeasitbearsonthequestionofthegreatervariabilityofourhedgerowtreescomparedwiththoseunderstrictlynaturalconditions。Awell—informedwriter（10/160。’Gardener’sChronicle’
1849？，page822。）statesthattheScotchfirpresentsfewvarietiesinitsnativeScotchforests;butthatit"variesmuchinfigureandfoliage，andinthesize，shape，andcolourofitscones，whenseveralgenerationshavebeenproducedawayfromitsnativelocality。"Thereislittledoubtthatthehighlandandlowlandvarietiesdifferinthevalueoftheirtimber，andthattheycanbepropagatedtrulybyseed;thusjustifyingLoudon’sremark，that"avarietyisoftenofasmuchimportanceasaspecies，andsometimesfarmoreso。"（10/161。’ArboretumetFruticetum’volume4page2150。）Imaymentiononeratherimportantpointinwhichthistreeoccasionallyvaries;
intheclassificationoftheConiferae，sectionsarefoundedonwhethertwo，three，orfiveleavesareincludedinthesamesheath;theScotchfirhasproperlyonlytwoleavesthusenclosed，butspecimenshavebeenobservedwithgroupsofthreeleavesinasheath。（10/162。’Gardener’sChronicle’1852page693。）Besidesthesedifferencesinthesemi—cultivatedScotchfir，thereareinseveralpartsofEuropenaturalorgeographicalraces，whichhavebeenrankedbysomeauthorsasdistinctspecies。（10/163。
See’BeitragezurKenntnissEuropaischerPinus—artenvonDr。Christ:Flora，1864。’HeshowsthatintheOber—EngadinP。sylvestrisandmontanaareconnectedbyintermediatelinks。）Loudon（10/164。’ArboretumetFruticetum’
volume4pages2159and2189。）considersP。pumilio，withitsseveralsub—
varieties，asmughus，nana，etc。，whichdiffermuchwhenplantedindifferentsoils，andonlycome"tolerablytruefromseed，"asalpinevarietiesoftheScotchfir;ifthiswereprovedtobethecase，itwouldbeaninterestingfactasshowingthatdwarfingfromlongexposuretoasevereclimateistoacertainextentinherited。
THEHAWTHORN（Crataegusoxyacantha）。
TheHawthornhasvariedmuch。Besidesendlessslightervariationsintheformoftheleaves，andinthesize，hardness，fleshiness，andshapeoftheberries，Loudon（10/165。Ibidvolume2page830;Loudon’s’Gardener’sMag。’
volume61830page714。）enumeratestwenty—ninewell—markedvarieties。
Besidesthosecultivatedfortheirprettyflowers，thereareotherswithgolden—yellow，black，andwhitishberries;otherswithwoollyberries，andotherswithre—curvedthorns。Loudontrulyremarksthatthechiefreasonwhythehawthornhasyieldedmorevarietiesthanmostothertrees，isthatnurserymenselectanyremarkablevarietyoutoftheimmensebedsofseedlingswhichareannuallyraisedformakinghedges。Theflowersofthehawthornusuallyincludefromonetothreepistils;butintwovarieties，namedmonogynaandsibirica，thereisonlyasinglepistil;andd’AssostatesthatthecommonthorninSpainisconstantlyinthisstate。（10/166。
Loudon’s’ArboretumetFruticetum’volume2page834。）Thereisalsoavarietywhichisapetalous，orhasitspetalsreducedtomererudiments。
ThefamousGlastonburythornflowersandleafstowardstheendofDecember，atwhichtimeitbearsberriesproducedfromanearliercropofflowers。
（10/167。Loudon’s’Gardener’sMag。’volume91833page123。）Itisworthnoticethatseveralvarietiesofthehawthorn，aswellasofthelimeandjuniper，areverydistinctintheirfoliageandhabitwhilstyoung，butinthecourseofthirtyorfortyyearsbecomeextremelylikeeachother（10/168。Ibidvolume111835page503。）thusremindingusofthewell—knownfactthatthedeodar，thecedarofLebanon，andthatoftheAtlas，aredistinguishedwiththegreatesteasewhilstyoung，butwithdifficultywhenold。
FLOWERS。
Ishallnotforseveralreasonstreatthevariabilityofplantswhicharecultivatedfortheirflowersaloneatanygreatlength。Manyofourfavouritekindsintheirpresentstatearethedescendantsoftwoormorespeciescrossedandcommingledtogether，andthiscircumstancealonewouldrenderitdifficulttodetectthedifferenceduetovariation。Forinstance，ourRoses，Petunias，Calceolarias，Fuchsias，Verbenas，Gladioli，Pelargoniums，etc。，certainlyhavehadamultipleorigin。Abotanistwellacquaintedwiththeparent—formswouldprobablydetectsomecuriousstructuraldifferencesintheircrossedandcultivateddescendant;andhewouldcertainlyobservemanynewandremarkableconstitutionalpeculiarities。Iwillgiveafewinstances，allrelatingtothePelargonium，andtakenchieflyfromMr。Beck（10/169。’Gardener’sChronicle’1845page623。）afamouscultivatorofthisplant:somevarietiesrequiremorewaterthanothers;someare"veryimpatientoftheknifeiftoogreedilyusedinmakingcuttings;"some，whenpotted，scarcely"showarootattheoutsideoftheballoftheearth;"onevarietyrequiresacertainamountofconfinementinthepottomakeitthrowupaflower—
stem;somevarietiesbloomwellatthecommencementoftheseason，othersattheclose;onevarietyisknown（10/170。D。Beatonin’CottageGardener’
1860page377。SeealsoMr。BeckonthehabitsofQueenMabin’Gardener’sChronicle’1845page226。），whichwillstand"evenpine—appletopandbottomheat，withoutlookinganymoredrawnthanifithadstoodinacommongreenhouse;andBlancheFleurseemsasifmadeonpurposeforgrowinginwinter，likemanybulbs，andtorestallsummer。"Theseoddconstitutionalpeculiaritieswouldenableaplantinastateofnaturetobecomeadaptedtowidelydifferentcircumstancesandclimates。
Flowerspossesslittleinterestunderourpresentpointofview，becausetheyhavebeenalmostexclusivelyattendedtoandselectedfortheirbeautifulcolour，size，perfectoutline，andmannerofgrowth。Intheseparticularshardlyonelong—cultivatedflowercanbenamedwhichhasnotvariedgreatly。Whatdoesafloristcarefortheshapeandstructureoftheorgansoffructification，unless，indeed，theyaddtothebeautyoftheflower？Whenthisisthecase，flowersbecomemodifiedinimportantpoints;
stamensandpistilsmaybeconvertedintopetals，andadditionalpetalsmaybedeveloped，asinalldoubleflowers。Theprocessofgradualselectionbywhichflowershavebeenrenderedmoreandmoredouble，eachstepintheprocessofconversionbeinginherited，hasbeenrecordedinseveralinstances。Intheso—calleddoubleflowersoftheCompositae，thecorollasofthecentralfloretsaregreatlymodified，andthemodificationsarelikewiseinherited。Inthecolumbine（Aquilegiavulgaris）someofthestamensareconvertedintopetalshavingtheshapeofnectaries，oneneatlyfittingintotheother;butinonevarietytheyareconvertedintosimplepetals。（10/171。Moquin—Tandon’ElementsdeTeratologie’1841page213。）Inthe"hoseinhose"primulae，thecalyxbecomesbrightlycolouredandenlargedsoastoresembleacorolla;andMr。W。Woolerinformsmethatthispeculiarityistransmitted;forhecrossedacommonpolyanthuswithonehavingacolouredcalyx（10/172。Seealso’CottageGardener’1860page133。）andsomeoftheseedlingsinheritedthecolouredcalyxduringatleastsixgenerations。Inthe"hen—and—chicken"daisythemainflowerissurroundedbyabroodofsmallflowersdevelopedfrombudsintheaxilsofthescalesoftheinvolucre。Awonderfulpoppyhasbeendescribed，inwhichthestamensareconvertedintopistils;andsostrictlywasthispeculiarityinheritedthat，outof154seedlings，onealonerevertedtotheordinaryandcommontype。（10/173。QuotedbyAlph。deCandolle’Bibl。
Univ。’November1862page58。）Ofthecock’s—comb（Celosiacristata），whichisanannual，thereareseveralracesinwhichtheflower—stemiswonderfully"fasciated"orcompressed;andonehasbeenexhibited（10/174。
Knight’Transact。Hort。Soc。’volume4page322。）actuallyeighteeninchesinbreadth。PeloricracesofGloxiniaspeciosaandAntirrhinummajuscanbepropagatedbyseed，andtheydifferinawonderfulmannerfromthetypicalformbothinstructureandappearance。
AmuchmoreremarkablemodificationhasbeenrecordedbySirWilliamandDr。Hooker（10/175。’BotanicalMagazine’tab。5160figure4;Dr。Hookerin’Gardener’sChronicle’1860page190;Prof。Harveyin’Gardener’sChronicle’1860page145;Mr。Crockerin’Gardener’sChron。’1861page1092。）inBegoniafrigida。Thisplantproperlyproducesmaleandfemaleflowersonthesamefascicles;andinthefemaleflowerstheperianthissuperior;butaplantatKewproduced，besidestheordinaryflowers，otherswhichgraduatedtowardsaperfecthermaphroditestructure;andintheseflowerstheperianthwasinferior。Toshowtheimportanceofthismodificationunderaclassificatorypointofview，ImayquotewhatProf。
Harveysays，namely，thathadit"occurredinastateofnature，andhadabotanistcollectedaplantwithsuchflowers，hewouldnotonlyhaveplaceditinadistinctgenusfromBegonia，butwouldprobablyhaveconsidereditasthetypeofanewnaturalorder。"Thismodificationcannotinonesensebeconsideredasamonstrosity，foranalogousstructuresnaturallyoccurinotherorders，aswithSaxifragaeandAristolochiaceae。TheinterestofthecaseislargelyaddedtobyMr。C。W。Crocker’sobservationthatseedlingsfromtheNORMALflowersproducedplantswhichbore，inaboutthesameproportionastheparent—plant，hermaphroditeflowershavinginferiorperianths。Thehermaphroditeflowersfertilisedwiththeirownpollenweresterile。
Iffloristshadattendedto，selected，andpropagatedbyseedothermodificationsofstructurebesidesthosewhicharebeautiful，ahostofcuriousvarietieswouldcertainlyhavebeenraised;andtheywouldprobablyhavetransmittedtheircharacterssotrulythatthecultivatorwouldhavefeltaggrieved，asinthecaseofculinaryvegetables，ifhiswholebedhadnotpresentedauniformappearance。Floristshaveattendedinsomeinstancestotheleavesoftheirplant，andhavethusproducedthemostelegantandsymmetricalpatternsofwhite，red，andgreen，which，asinthecaseofthepelargonium，aresometimesstrictlyinherited。（10/176。Alph。
deCandolle’Geograph。Bot。’page1083;’Gardener’sChronicle’1861page433。TheinheritanceofthewhiteandgoldenzonesinPelargoniumlargelydependsonthenatureofthesoil。SeeD。Beatonin’JournalofHorticulture’1861page64。）Anyonewhowillhabituallyexaminehighly—
cultivatedflowersingardensandgreenhouseswillobservenumerousdeviationsinstructure;butmostofthesemustberankedasmeremonstrosities，andareonlysofarinterestingasshowinghowplastictheorganisationbecomesunderhighcultivation。FromthispointofviewsuchworksasProfessorMoquin—Tandon’s’Teratologie’arehighlyinstructive。
ROSES。
Theseflowersofferaninstanceofanumberofformsgenerallyrankedasspecies，namely，R。centifolia，gallica，alba，damascena，spinosissima，bracteata，indica，semperflorens，moschata，etc。，whichhavelargelyvariedandbeenintercrossed。ThegenusRosaisanotoriouslydifficultone，and，thoughsomeoftheaboveformsareadmittedbyallbotaniststobedistinctspecies，othersaredoubtful;thus，withrespecttotheBritishforms，Babingtonmakesseventeen，andBenthamonlyfivespecies。Thehybridsfromsomeofthemostdistinctforms——forinstance，fromR。indica，fertilisedbythepollenofR。centifolia——produceanabundanceofseed;IstatethisontheauthorityofMr。Rivers（10/177。’RoseAmateur’sGuide’T。Rivers1837page21。）fromwhoseworkIhavedrawnmostofthefollowingstatements。Asalmostalltheaboriginalformsbroughtfromdifferentcountrieshavebeencrossedandre—crossed，itisnowonderthatTargioni—
Tozzetti，inspeakingofthecommonrosesoftheItaliangardens，remarksthat"thenativecountryandpreciseformofthewildtypeofmostofthemareinvolvedinmuchuncertainty。"（10/178。’JournalHort。Soc。’volume9
1855page182。）Nevertheless，Mr。RiversinreferringtoR。indica（page68）saysthatthedescendantsofeachgroupmaygenerallyberecognisedbyacloseobserver。Thesameauthoroftenspeaksofrosesashavingbeenalittlehybridised;butitisevidentthatinverymanycasesthedifferencesduetovariationandtohybridisationcannowonlybeconjecturallydistinguished。
Thespecieshavevariedbothbyseedandbybud;suchmodifiedbudsbeingoftencalledbygardenerssports。InthefollowingchapterIshallfullydiscussthislattersubject，andshallshowthatbud—variationscanbepropagatednotonlybygraftingandbudding，butoftenbyseed。Wheneveranewroseappearswithanypeculiarcharacter，howeverproduced，ifityieldsseed，Mr。Rivers（page4）fullyexpectsittobecometheparent—typeofanewfamily。Thetendencytovaryissostronginsomekinds，asintheVillageMaid（Riverspage16），thatwhengrownindifferentsoilsitvariessomuchincolourthatithasbeenthoughttoformseveraldistinctkinds。
Altogetherthenumberofkindsisverygreat:thusM。Desportes，inhisCataloguefor1829，enumerates2562ascultivatedinFrance;butnodoubtalargeproportionofthesearemerelynominal。
Itwouldbeuselesstospecifythemanypointsofdifferencebetweenthevariouskinds，butsomeconstitutionalpeculiaritiesmaybementioned。
SeveralFrenchroses（Riverspage12）willnotsucceedinEngland;andanexcellenthorticulturist（10/179。TheRev。W。F。Radclyffein’JournalofHorticulture’March14，1865page207。）remarks，that"Eveninthesamegardenyouwillfindthatarosethatwilldonothingunderasouthwallwilldowellunderanorthone。ThatisthecasewithPaulJosephhere。Itgrowsstronglyandbloomsbeautifullyclosetoanorthwall。Forthreeyearssevenplantshavedonenothingunderasouthwall。"Manyrosescanbeforced，"manyaretotallyunfitforforcing，amongwhichisGeneralJacqueminot。"（10/180。’Gardener’sChronicle’1831page46。）FromtheeffectsofcrossingandvariationMr。Riversenthusiasticallyanticipates（page87）thatthedaywillcomewhenallourroses，evenmoss—roses，willhaveevergreenfoliage，brilliantandfragrantflowers，andthehabitofbloomingfromJunetillNovember。"Adistantviewthisseems，butperseveranceingardeningwillyetachievewonders，"asassuredlyithasalreadyachievedwonders。
Itmaybeworthwhilebrieflytogivethewell—knownhistoryofoneclassofroses。In1793somewildScotchroses（R。spinosissima）weretransplantedintoagarden（10/181。Mr。Sabinein’Transact。Hort。Soc。’
volume4page285。）;andoneoftheseboreflowersslightlytingedwithred，fromwhichaplantwasraisedwithsemi—monstrousflowers，alsotingedwithred;seedlingsfromthisflowerweresemi—double，andbycontinuedselection，inaboutnineortenyears，eightsub—varietieswereraised。InthecourseoflessthantwentyyearsthesedoubleScotchroseshadsomuchincreasedinnumberandkind，thattwenty—sixwell—markedvarieties，classedineightsections，weredescribedbyMr。Sabine。In1841（10/182。
’AnEncyclop。ofPlants’byJ。C。Loudon1841page443。）itissaidthatthreehundredvarietiescouldbeprocuredinthenursery—gardensnearGlasgow;andthesearedescribedasblush，crimson，purple，red，marbled，two—coloured，white，andyellow，andasdifferingmuchinthesizeandshapeoftheflower。
PANSYORHEARTSEASE（Violatricolor，etc。）。
Thehistoryofthisflowerseemstobeprettywellknown;itwasgrowninEvelyn’sgardenin1687;butthevarietieswerenotattendedtotill1810—
1812，whenLadyMonke，togetherwithMr。Lee，thewell—knownnursery—man，energeticallycommencedtheirculture;andinthecourseofafewyearstwentyvarietiescouldbepurchased。（10/183。Loudon’s’Gardener’sMagazine’volume111835page427;alsoJournalofHorticulture’April14，1863page275。）Ataboutthesameperiod，namelyin1813or1814，LordGambiercollectedsomewildplants，andhisgardener，Mr。Thomson，cultivatedthem，togetherwithsomecommongardenvarieties，andsooneffectedagreatimprovement。Thefirstgreatchangewastheconversionofthedarklinesinthecentreoftheflowerintoadarkeyeorcentre，whichatthatperiodhadneverbeenseen，butisnowconsideredoneofthechiefrequisitesofafirst—rateflower。In1835abookentirelydevotedtothisflowerwaspublished，andfourhundrednamedvarietieswereonsale。Fromthesecircumstancesthisplantseemedtomeworthstudying，moreespeciallyfromthegreatcontrastbetweenthesmall，dull，elongated，irregularflowersofthewildpansy，andthebeautiful，flat，symmetrical，circular，velvet—likeflowers，morethantwoinchesindiameter，magnificentlyandvariouslycoloured，whichareexhibitedatourshows。ButwhenIcametoenquiremoreclosely，Ifoundthat，thoughthevarietiesweresomodern，yetthatmuchconfusionanddoubtprevailedabouttheirparentage。Floristsbelievethatthevarieties（10/184。Loudon’s’Gardener’sMagazine’volume8
page575:volume9page689。）aredescendedfromseveralwildstocks，namely，V。tricolor，lutea，grandiflora，amoena，andaltaica，moreorlessintercrossed。AndwhenIlookedtobotanicalworkstoascertainwhethertheseformsoughttoberankedasspecies，Ifoundequaldoubtandconfusion。Violaaltaicaseemstobeadistinctform，butwhatpartithasplayedintheoriginofourvarietiesIknownot;itissaidtohavebeencrossedwithV。lutea。Violaamoena（10/185。SirJ。E。Smith’EnglishFlora’
volume1page306。H。C。Watson’CybeleBritannica’volume11847page181。）
isnowlookedatbyallbotanistsasanaturalvarietyofV。grandiflora;
andthisandV。sudeticahavebeenprovedtobeidenticalwithV。lutea。
ThelatterandV。tricolor（includingitsadmittedvarietyV。arvensis）arerankedasdistinctspeciesbyBabington，andlikewisebyM。Gay（10/186。
Quotedfrom’AnnalesdesSciences’intheCompaniontothe’Bot。Mag。’
volume11835page159。）whohaspaidparticularattentiontothegenus;
butthespecificdistinctionbetweenV。luteaandtricolorischieflygroundedontheonebeingstrictlyandtheothernotstrictlyperennial，aswellasonsomeotherslightandunimportantdifferencesintheformofthestemandstipules。Benthamunitesthesetwoforms;andahighauthorityonsuchmatters，Mr。H。C。Watson（10/187。’CybeleBritannica’volume1page173。SeealsoDr。Herbertonthechangesofcolourintransplantedspecimens，andonthenaturalvariationsofV。grandiflora，in’Transact。
Hort。Soc。’volume4page19。）saysthat，"whileV。tricolorpassesintoV。
arvensisontheoneside，itapproximatessomuchtowardsV。luteaandV。
Curtisiiontheotherside，thatadistinctionbecomesscarcelymoreeasybetweenthem。"
Hence，afterhavingcarefullycomparednumerousvarieties，Igaveuptheattemptastoodifficultforanyoneexceptaprofessedbotanist。Mostofthevarietiespresentsuchinconstantcharacters，thatwhengrowninpoorsoil，orwhenfloweringoutoftheirproperseason，theyproduceddifferentlycolouredandmuchsmallerflowers。Cultivatorsspeakofthisorthatkindasbeingremarkablyconstantortrue;butbythistheydonotmean，asinothercases，thatthekindtransmitsitscharacterbyseed，butthattheindividualplantdoesnotchangemuchunderculture。Theprincipleofinheritance，however，doesholdgoodtoacertainextentevenwiththefleetingvarietiesoftheHeartsease，fortogaingoodsortsitisindispensabletosowtheseedofgoodsorts。Nevertheless，inalmosteverylargeseed—bedafew，almostwildseedlingsreappearthroughreversion。Oncomparingthechoicestvarietieswiththenearestalliedwildforms，besidesthedifferenceinthesize，outline，andcolouroftheflowers，theleavessometimesdifferinshape，asdoesthecalyxoccasionallyinthelengthandbreadthofthesepals。Thedifferencesintheformofthenectarymoreespeciallydeservenotice;becausecharactersderivedfromthisorganhavebeenmuchusedinthediscriminationofmostofthespeciesofViola。Inalargenumberofflowerscomparedin1842Ifoundthatinthegreaternumberthenectarywasstraight;inotherstheextremitywasalittleturnedupwards，ordownwards，orinwards，soastobecompletelyhooked;inothers，insteadofbeinghooked，itwasfirstturnedrectangularlydownwards，andthenbackwardsandupwards;inothers，theextremitywasconsiderablyenlarged;andlastly，insomethebasalpartwasdepressed，becoming，asusual，laterallycompressedtowardstheextremity。
Inalargenumberofflowers，ontheotherhand，examinedbymein1856
fromanursery—gardeninadifferentpartofEngland，thenectaryhardlyvariedatall。NowM。Gaysaysthatincertaindistricts，especiallyinAuvergne，thenectaryofthewildV。grandifloravariesinthemannerjustdescribed。MustweconcludefromthisthatthecultivatedvarietiesfirstmentionedwerealldescendedfromV。grandiflora，andthatthesecondlot，thoughhavingthesamegeneralappearance，weredescendedfromV。tricolor，ofwhichthenectary，accordingtoM。Gay，issubjecttolittlevariation？
Orisitnotmoreprobablethatboththesewildformswouldbefoundunderotherconditionstovaryinthesamemanneranddegree，thusshowingthattheyoughtnottoberankedasspecificallydistinct？