TRANSLATEDBYALEXANDERTEIXEIRADEMATTOS,F。Z。S。
TRANSLATOR\'SNOTE
InthisvolumeIhavecollectedalltheessaysonWildBeesscatteredthroughthe\"Souvenirsentomologiques,\"withtheexceptionofthoseontheChalicodomae,orMason—beesproper,whichformthecontentsofaseparatevolumeentitled\"TheMason—bees。\"
ThefirsttwoessaysontheHalicti（Chapters12and13）havealreadyappearedinanabbreviatedformin\"TheLifeandLoveoftheInsect,\"
translatedbymyselfandpublishedbyMessrs。A。&C。Black（inAmericabytheMacmillanCo。）in1911。Withthegreatestcourtesyandkindness,Messrs。Blackhavegivenmetheirpermissiontoincludethesetwochaptersinthepresentvolume;theydidsowithoutfeeorconsiderationofanykind,merelyonmyrepresentationthatitwouldbeagreatpityifthisuniformeditionofFabre\'sWorksshouldberenderedincompletebecausecertainessaysformedpartofvolumesofextractspreviouslypublishedinthiscountry。Theirgenerosityisalmostunparalleledinmyexperience;andIwishtothankthempubliclyforitinthenameoftheauthor,oftheFrenchpublishersandoftheEnglishandAmericanpublishers,aswellasinmyown。
Oftheremainingchapters,oneortwohaveappearedinthe\"EnglishReview\"orothermagazines;butmostofthemnowseethelightinEnglishforthefirsttime。
Ihaveoncemore,asinthecaseof\"TheMason—bees,\"tothankMissFrancesRodwellforthehelpwhichshehasgivenmeintheworkoftranslationandresearch;andIamalsogratefulformuchkindassistancereceivedfromthestaffoftheNaturalHistoryMuseumandfromMr。GeoffreyMeade—Waldoinparticular。
ALEXANDERTEIXEIRADEMATTOS。
Chelsea,1915。
CONTENTS。
TRANSLATOR\'SNOTE。
CHAPTER1。BRAMBLE—DWELLERS。
CHAPTER2。THEOSMIAE。
CHAPTER3。THEDISTRIBUTIONOFTHESEXES。
CHAPTER4。THEMOTHERDECIDESTHESEXOFTHEEGG。
CHAPTER5。PERMUTATIONSOFSEX。
CHAPTER6。INSTINCTANDDISCERNMENT。
CHAPTER7。ECONOMYOFENERGY。
CHAPTER8。THELEAF—CUTTERS。
CHAPTER9。THECOTTON—BEES。
CHAPTER10。THERESIN—BEES。
CHAPTER11。THEPOISONOFTHEBEE。
CHAPTER12。THEHALICTI:APARASITE。
CHAPTER13。THEHALICTI:THEPORTRESS。
CHAPTER14。THEHALICTI:PARTHENOGENESIS。
INDEX。
CHAPTER1。BRAMBLE—DWELLERS。
Thepeasant,ashetrimshishedge,whoseriotoustanglethreatenstoencroachupontheroad,cutsthetrailingstemsofthebrambleafootortwofromthegroundandleavestheroot—stock,whichsoondriesup。Thesebramble—stumps,shelteredandprotectedbythethornybrushwood,areingreatdemandamongahostofHymenopterawhohavefamiliestosettle。Thestump,whendry,offerstoanyonethatknowshowtouseitahygienicdwelling,wherethereisnofearofdampfromthesap;itssoftandabundantpithlendsitselftoeasywork;
andthetopoffersaweakspotwhichmakesitpossiblefortheinsecttoreachtheveinofleastresistanceatonce,withoutcuttingawaythroughthehardligneouswall。Tomany,therefore,oftheBeeandWasptribe,whetherhoney—gatherersorhunters,oneofthesedrystalksisavaluablediscoverywhenitsdiametermatchesthesizeofitswould—beinhabitants;anditisalsoaninterestingsubjectofstudytotheentomologistwho,inthewinter,pruning—shearsinhand,cangatherinthehedgerowsafaggotrichinsmallindustrialwonders。Visitingthebramble—busheshaslongbeenoneofmyfavouritepastimesduringtheenforcedleisureofthewintertime;anditisseldombutsomenewdiscovery,someunexpectedfact,makesuptomeformytornfingers。
Mylist,whichisstillfarfrombeingcomplete,alreadynumbersnearlythirtyspeciesofbramble—dwellersintheneighbourhoodofmyhouse;otherobservers,moreassiduousthanI,exploringanotherregionandonecoveringawiderrange,havecountedasmanyasfifty。
IgiveatfootaninventoryofthespecieswhichIhavenoted。
（Bramble—dwellinginsectsintheneighbourhoodofSerignan（Vaucluse）:
1。MELLIFEROUSHYMENOPTERA。
Osmiatridentata,DUF。andPER。
Osmiadetrita,PEREZ。
Anthidiumscapulare,LATR。
Heriadesrubicola,PEREZ。
Prosopisconfusa,SCHENCK。
Ceratinachalcites,GERM。
Ceratinaalbilabris,FAB。
Ceratinacallosa,FAB。
Ceratinacoerulea,VILLERS。
2。HUNTINGHYMENOPTERA。
Soleniusvagus,FAB。（provisions,Diptera）。
Soleniuslapidarius,LEP。（provisions,Spiders？）。
Cemonusunicolor,PANZ。（provisions,Plant—lice）。
Psenatratus（provisions,BlackPlant—lice）。
Tripoxylonfigulus,LIN。（provisions,Spiders）。
APompilus,unknown（provisions,Spiders）。
Odynerusdelphinalis,GIRAUD。
3。PARASITICALHYMENOPTERA。
ALeucopsis,unknown（parasiteofAnthidiumscapulare）。
AsmallScoliid,unknown（parasiteofSoleniusvagus）。
Omalusauratus（parasiteofvariousbramble—dwellers）。
Cryptusbimaculatus,GRAV。（parasiteofOsmiadetrita）。
Cryptusgyrator,DUF。（parasiteofTripoxylonfigulus）。
Ephialtesdivinator,ROSSI（parasiteofCemonusunicolor）。
Ephialtesmediator,GRAV。（parasiteofPsenatratus）。
Foenuspyrenaicus,GUERIN。
Euritomarubicola,J。GIRAUD（parasiteofOsmiadetrita）。
4。COLEOPTERA。
Zonitismutica,FAB。（parasiteofOsmiatridentata）。
Mostoftheseinsectshavebeensubmittedtoalearnedexpert,ProfessorJeanPerez,ofBordeaux。Itakethisopportunityofrenewingmythanksforhiskindnessinidentifyingthemforme。——
Author\'sNote。）
Theyincludemembersofverydiversecorporations。Some,moreindustriousandequippedwithbettertools,removethepithfromthedrystemandthusobtainaverticalcylindricalgallery,thelengthofwhichmaybenearlyacubit。Thissheathisnextdivided,bypartitions,intomoreorlessnumerousstoreys,eachofwhichformsthecellofalarva。Others,lesswell—endowedwithstrengthandimplements,availthemselvesoftheoldgalleriesofotherinsects,galleriesthathavebeenabandonedafterservingasahomefortheirbuilder\'sfamily。Theironlyworkistomakesomeslightrepairsintheruinedtenement,toclearthechannelofitslumber,suchastheremainsofcocoonsandthelitterofshatteredceilings,andlastlytobuildnewpartitions,eitherwithaplastermadeofclayorwithaconcreteformedofpith—scrapingscementedwithadropofsaliva。
Youcantelltheseborroweddwellingsbytheunequalsizeofthestoreys。Whentheworkerhasherselfboredthechannel,sheeconomizesherspace:sheknowshowcostlyitis。Thecells,inthatcase,areallalike,thepropersizeforthetenant,neithertoolargenortoosmall。Inthisbox,whichhascostweeksoflabour,theinsecthastohousethelargestpossiblenumberoflarvae,whileallottingthenecessaryamountofroomtoeach。Methodinthesuperpositionofthefloorsandeconomyofspaceareheretheabsoluterule。
Butthereisevidenceofwastewhentheinsectmakesuseofabramblehollowedbyanother。ThisisthecasewithTripoxylonfigulus。Toobtainthestore—roomswhereintodepositherscantystockofSpiders,shedividesherborrowedcylinderintoveryunequalcells,bymeansofslenderclaypartitions。Someareacentimetre（。39
inch。——Translator\'sNote。）deep,thepropersizefortheinsect;
othersareasmuchastwoinches。Thesespaciousrooms,outofallproportiontotheoccupier,revealtherecklessextravaganceofacasualproprietresswhosetitle—deedshavecosthernothing。
But,whethertheybetheoriginalbuildersorlabourerstouchinguptheworkofothers,theyallalikehavetheirparasites,whoconstitutethethirdclassofbramble—dwellers。Thesehaveneithergalleriestoexcavatenorvictualstoprovide;theylaytheiregginastrangecell;andtheirgrubfeedseitherontheprovisionsofthelawfulowner\'slarvaoronthatlarvaitself。
Attheheadofthispopulation,asregardsboththefinishandthemagnitudeofthestructure,standstheThree—prongedOsmia（Osmiatridentata,DUF。andPER。）,towhomthischaptershallbespeciallydevoted。Hergallery,whichhasthediameterofaleadpencil,sometimesdescendstoadepthoftwentyinches。Itisatfirstalmostexactlycylindrical;but,inthecourseofthevictualling,changesoccurwhichmodifyitslightlyatgeometricallydetermineddistances。
Theworkofboringpossessesnogreatinterest。InthemonthofJuly,weseetheinsect,perchedonabramble—stump,attackthepithanddigitselfawell。Whenthisisdeepenough,theOsmiagoesdown,tearsoffafewparticlesofpithandcomesupagaintoflingherloadoutside。ThismonotonouslabourcontinuesuntiltheBeedeemsthegallerylongenough,oruntil,asoftenhappens,shefindsherselfstoppedbyanimpassableknot。
Nextcomestherationofhoney,thelayingoftheeggandthepartitioning,thelastadelicateoperationtowhichtheinsectproceedsbydegreesfromthebasetothetop。Atthebottomofthegallery,apileofhoneyisplacedandanegglaiduponthepile;
thenapartitionisbuilttoseparatethiscellfromthenext,foreachlarvamusthaveitsspecialchamber,aboutacentimetreandahalf（。58inch。——Translator\'sNote。）long,havingnocommunicationwiththechambersadjoining。Thematerialsemployedforthispartitionarebramble—sawdust,gluedintoapastewiththeinsects\'
saliva。Whencearethesematerialsobtained？DoestheOsmiagooutside,togatheronthegroundtherubbishwhichsheflungoutwhenboringthecylinder？Onthecontrary,sheisfrugalofhertimeandhasbetterthingstodothantopickupthescatteredparticlesfromthesoil。Thechannel,asIsaid,isatfirstuniforminsize,almostcylindrical;itssidesstillretainathincoatingofpith,formingthereserveswhichtheOsmia,asaprovidentbuilder,haseconomizedwherewithtoconstructthepartitions。Soshescrapesawaywithhermandibles,keepingwithinacertainradius,aradiusthatcorrespondswiththedimensionsofthecellwhichsheisgoingtobuildnext;
moreover,sheconductsherworkinsuchawayastohollowoutmoreinthemiddleandleavethetwoendscontracted。Inthismanner,thecylindricalchannelofthestartissucceeded,intheworkedportion,byanovoidcavityflattenedatbothends,aspaceresemblingalittlebarrel。Thisspacewillformthesecondcell。
Asfortherubbish,itisutilizedonthespotforthelidorcoverthatservesasaceilingforonecellandafloorforthenext。Ourownmaster—builderscouldnotcontrivemoresuccessfullytomakethebestuseoftheirlabourers\'time。Onthefloorthusobtained,asecondrationofhoneyisplaced;andaneggislaidonthesurfaceofthepaste。Lastly,attheupperendofthelittlebarrel,apartitionisbuiltwiththescrapingsobtainedinthecourseofthefinalworkonthethirdcell,whichitselfisshapedlikeaflattenedovoid。Andsotheworkgoeson,celluponcell,eachsupplyingthematerialsforthepartitionseparatingitfromtheonebelow。Onreachingtheendofthecylinder,theOsmiaclosesupthecasewithathicklayerofthesamemortar。Thenthatbramble—stumpisdonewith;
theBeewillnotreturntoit。Ifherovariesarenotyetexhausted,otherdrystemswillbeexploitedinthesamefashion。
Thenumberofcellsvariesgreatly,accordingtothequalitiesofthestalk。Ifthebramble—stumpbelong,regularandsmooth,wemaycountasmanyasfifteen:that,atleast,isthehighestfigurewhichmyobservationshavesupplied。Toobtainagoodideaoftheinternaldistribution,wemustsplitthestalklengthwise,inthewinter,whentheprovisionshavelongbeenconsumedandwhenthelarvaearewrappedintheircocoons。Wethenseethat,atregularintervals,thecasebecomesslightlynarrower;andineachofthenecksthusformedacirculardiskisfixed,apartitiononeortwomillimetresthick。
（。039to。079inch。——Translator\'sNote。）Theroomsseparatedbythesepartitionsformsomanylittlebarrelsorkegs,eachcompactlyfilledwithareddish,transparentcocoon,throughwhichthelarvashows,bentintoafish—hook。Thewholesuggestsastringofrough,ovalamberbeads,touchingattheiramputatedends。
Inthisstringofcocoons,whichistheoldest,whichtheyoungest？
Theoldestisobviouslythebottomone,theonewhosecellwasthefirstbuilt;theyoungestistheoneatthetopoftherow,theoneinthecelllastbuilt。Theoldestofthelarvaestartsthepile,downatthebottomofthegallery;thelatestarrivalendsitatthetop;andthoseinbetweenfollowupononeanother,accordingtoage,frombasetoapex。
LetusnextobservethatthereisnoroomintheshaftfortwoOsmiaeatatimeonthesamelevel,foreachcocoonfillsupthestorey,thekegthatbelongstoit,withoutleavinganyvacantspace;letusalsoremarkthat,whentheyattainthestageofperfection,theOsmiaemustallemergefromtheshaftbytheonlyorificewhichthebramble—
stemboasts,theorificeatthetop。Thereisherebutoneobstacle,easytoovercome:aplugofgluedpith,ofwhichtheinsect\'smandiblesmakeshortwork。Downbelow,thestalkoffersnoreadyoutlet;besides,itisprolongedundergroundindefinitelybytheroots。Everywhereelseistheligneousfence,generallytoohardandthicktobreakthrough。ItisinevitablethereforethatalltheOsmiae,whenthetimecomestoquittheirdwelling,shouldgooutbythetop;and,asthenarrownessoftheshaftbarsthepassageoftheprecedinginsectaslongasthenextinsect,theoneaboveit,remainsinposition,theremovalmustbeginatthetop,extendfromcelltocellandendatthebottom。Consequently,theorderofexitistheconversetotheorderofbirth:theyoungerOsmiaeleavethenestfirst,theireldersleaveitlast。
Theoldest,thatistosay,thebottomone,wasthefirsttofinishhersupplyofhoneyandtospinhercocoon。Takingprecedenceofallhersistersinthewholeseriesofheractions,shewasthefirsttobursthersilkenbagandtodestroytheceilingthatclosesherroom:
atleast,thatiswhatthelogicofthesituationtakesforgranted。
Inheranxietytogetout,howwillshesetaboutherrelease？Thewayisblockedbythenearestcocoons,asyetintact。Toclearherselfapassagethroughthestringofthosecocoonswouldmeantoexterminatetheremainderofthebrood;thedeliveranceofonewouldmeanthedestructionofalltherest。Insectsarenotoriouslyobstinateintheiractionsandunscrupulousintheirmethods。IftheBeeatthebottomoftheshaftwantstoleaveherlodging,willshesparethosewhobarherroad？
Thedifficultyisgreat,obviously;itseemsinsuperable。Thereuponwebecomesuspicious:webegintowonderiftheemergencefromthecocoon,thatistosay,thehatching,reallytakesplaceintheorderofprimogeniture。Mightitnotbe——byaverysingularexception,itistrue,butonewhichisnecessaryinsuchcircumstances——thattheyoungestoftheOsmiaeburstshercocoonfirstandtheoldestlast;
inshort,thatthehatchingproceedsfromonechambertothenextintheinversedirectiontothatwhichtheageoftheoccupantswouldleadustopresume？Inthatcase,thewholedifficultywouldberemoved:eachOsmia,assherenthersilkenprison,wouldfindaclearroadinfrontofher,theOsmiaenearertheoutlethavinggoneoutbeforeher。Butisthisreallyhowthingshappen？Ourtheoriesveryoftendonotagreewiththeinsect\'spractice;evenwhereourreasoningseemsmostlogical,weshouldbemoreprudenttoseewhathappensbeforeventuringonanypositivestatements。LeonDufourwasnotsoprudentwhenhe,thefirstinthefield,tookthislittleprobleminhand。HedescribestousthehabitsofanOdynerus（Odynerusrubicola,DUF。）whopilesupclaycellsintheshaftofadrybramble—stalk;and,fullofenthusiasmforhisindustriousWasp,hegoesontosay:
\'Pictureastringofeightcementshells,placedendtoendandcloselywedgedinsideawoodensheath。Thelowestwasundeniablymadefirstandconsequentlycontainsthefirst—laidegg,which,accordingtorules,shouldgivebirthtothefirstwingedinsect。Howdoyouimaginethatthelarvainthatfirstshellwasbiddentowaiveitsrightofprimogenitureandonlytocompleteitsmetamorphosisafterallitsjuniors？Whataretheconditionsbroughtintoplaytoproducearesultapparentlysocontrarytothelawsofnature？Humbleyourselfinthepresenceoftherealityandconfessyourignorance,ratherthanattempttohideyourembarrassmentundervainexplanations！
\'Ifthefirstegglaidbythebusymotherweredestinedtobethefirst—bornoftheOdyneri,thatone,inordertoseethelightimmediatelyafterachievingwings,wouldhavehadtheoptioneitherofbreakingthroughthedoublewallsofhisprisonorofperforating,frombottomtotop,thesevenshellsaheadofhim,inordertoemergethroughthetruncateendofthebramble—stem。Nownature,whilerefusinganywayofescapelaterally,wasalsoboundtovetoanydirectinvasion,thebrutalgimlet—workwhichwouldinevitablyhavesacrificedsevenmembersofonefamilyforthesafetyofanonlyson。
Natureisasingeniousindesignassheisfertileinresource,andshemusthaveforeseenandforestalledeverydifficulty。Shedecidedthatthelast—builtcradleshouldyieldthefirst—bornchild;thatthisoneshouldcleartheroadforhisnextoldestbrother,thesecondforthethirdandsoon。AndthisistheorderinwhichthebirthofourOdynerioftheBramblesactuallytakesplace。\'
Yes,myreveredmaster,Iwilladmitwithouthesitationthatthebramble—dwellersleavetheirsheathintheconverseordertothatoftheirages:theyoungestfirst,theoldestlast;ifnotinvariably,atleastveryoften。Butdoesthehatching,bywhichImeantheemergencefromthecocoon,takeplaceinthesameorder？Doestheevolutionoftheelderwaituponthatoftheyounger,sothateachmaygivethosewhowouldbarhispassagetimetoeffecttheirdeliveranceandtoleavetheroadclear？Iverymuchfearthatlogichascarriedyourdeductionsbeyondtheboundsofreality。Rationallyspeaking,mydearsir,nothingcouldbemoreaccuratethanyourinferences;andyetwemustforgothetheoryofthestrangeinversionwhichyousuggest。NoneoftheBramble—beeswithwhomIhaveexperimentedbehavesafterthatfashion。IknownothingpersonalaboutOdynerusrubicola,whoappearstobeastrangerinmydistrict;
but,asthemethodofleavingmustbealmostthesamewhenthehabitationisexactlysimilar,itisenough,Ithink,toexperimentwithsomeofthebramble—dwellersinordertolearnthehistoryoftherest。
Mystudieswill,bypreference,bearupontheThree—prongedOsmia,wholendsherselfmorereadilytolaboratoryexperiments,bothbecausesheisstrongerandbecausethesamestalkwillcontainagoodlynumberofhercells。Thefirstfacttobeascertainedistheorderofhatching。Itakeaglasstube,closedatoneend,openattheotherandofadiametersimilartothatoftheOsmia\'stunnel。InthisIplace,oneabovetheother,exactlyintheirnaturalorder,thetencocoons,orthereabouts,whichIextractfromastumpofbramble。Theoperationisperformedinwinter。Thelarvae,atthattime,havelongbeenenvelopedintheirsilkencase。Toseparatethecocoonsfromoneanother,Iemployartificialpartitionsconsistingoflittlerounddisksofsorghum,orIndianmillet,abouthalfacentimetrethick。（Aboutone—fifthofaninch。——Translator\'sNote。）
Thisisawhitepith,divestedofitsfibrouswrapperandeasyfortheOsmia\'smandiblestoattack。Mydiaphragmsaremuchthickerthanthenaturalpartitions;thisisanadvantage,asweshallsee。Inanycase,Icouldnotwellusethinnerones,forthesedisksmustbeabletowithstandthepressureoftherammerwhichplacestheminpositioninthetube。Ontheotherhand,theexperimentshowedmethattheOsmiamakesshortworkofthematerialwhenitisacaseofdrillingaholethroughit。
Tokeepoutthelight,whichwoulddisturbmyinsectsdestinedtospendtheirlarvallifeincompletedarkness,Icoverthetubewithathickpapersheath,easytoremoveandreplacewhenthetimecomesforobservation。Lastly,thetubesthuspreparedandcontainingeitherOsmiaeorotherbramble—dwellersarehungvertically,withtheopeningatthetop,inasnugcornerofmystudy。Eachoftheseappliancesfulfilsthenaturalconditionsprettysatisfactorily:thecocoonsfromthesamebramble—stickarestackedinthesameorderwhichtheyoccupiedinthenativeshaft,theoldestatthebottomofthetubeandtheyoungestclosetotheorifice;theyareisolatedbymeansofpartitions;theyareplacedvertically,headupwards;
moreover,mydevicehastheadvantageofsubstitutingfortheopaquewallofthebrambleatransparentwallwhichwillenablemetofollowthehatchingdaybyday,atanymomentwhichIthinkopportune。
ThemaleOsmiasplitshiscocoonattheendofJuneandthefemaleatthebeginningofJuly。Whenthistimecomes,wemustredoubleourwatchandinspectthetubesseveraltimesadayifwewouldobtainexactstatisticsofthebirths。Well,duringthesixyearsthatI
havestudiedthisquestion,Ihaveseenandseenagain,adnauseam;
andIaminapositiontodeclarethatthereisnoordergoverningthesequenceofhatchings,absolutelynone。Thefirstcocoontoburstmaybetheoneatthebottomofthetube,theoneatthetop,theoneinthemiddleorinanyotherpart,indifferently。Thesecondtobesplitmayadjointhefirstoritmayberemovedfromitbyanumberofspaces,eitheraboveorbelow。Sometimesseveralhatchingsoccuronthesameday,withinthesamehour,somefartherbackintherowofcells,somefartherforward;andthiswithoutanyapparentreasonforthesimultaneity。Inshort,thehatchingsfollowupononeanother,Iwillnotsayhaphazard——foreachofthemhasitsappointedplaceintime,determinedbyimpenetrablecauses——butatanyratecontrarytoourcalculations,basedonthisortheotherconsideration。
Hadwenotbeendeceivedbyourtooshallowlogic,wemighthaveforeseenthisresult。Theeggsarelaidintheirrespectivecellsatintervalsofafewdays,ofafewhours。Howcanthisslightdifferenceinageaffectthetotalevolution,whichlastsayear？
Mathematicalaccuracyhasnothingtodowiththecase。Eachgerm,eachgrubhasitsindividualenergy,determinedweknownothowandvaryingineachgermorgrub。Thisexcessofvitalitybelongstotheeggbeforeitleavestheovary。Mightitnot,atthemomentofhatching,bethecausewhythisorthatlarvatakesprecedenceofitseldersoritsjuniors,chronologybeingaltogetherasecondaryconsideration？Whenthehensitsuponhereggs,istheoldestalwaysthefirsttohatch？Inthesameway,theoldestlarva,lodgedinthebottomstorey,neednotnecessarilyreachtheperfectstatefirst。
Asecondargument,hadwereflectedmoredeeplyonthematter,wouldhaveshakenourfaithinanystrictmathematicalsequence。Thesamebroodformingthestringofcocoonsinabramble—stemcontainsbothmalesandfemales;andthetwosexesaredividedintheseriesindiscriminately。NowitistheruleamongtheBeesforthemalestoissuefromthecocoonalittleearlierthanthefemales。InthecaseoftheThree—prongedOsmia,themalehasaboutaweek\'sstart。
Consequently,inapopulousgallery,thereisalwaysacertainnumberofmales,whoarehatchedsevenoreightdaysbeforethefemalesandwhoaredistributedhereandthereovertheseries。Thiswouldbeenoughtomakeanyregularhatching—sequenceimpossibleineitherdirection。
Thesesurmisesaccordwiththefacts:thechronologicalsequenceofthecellstellsusnothingaboutthechronologicalsequenceofthehatchings,whichtakeplacewithoutanydefiniteorder。Thereis,therefore,nosurrenderofrightsofprimogeniture,asLeonDufourthought:eachinsect,regardlessoftheothers,burstsitscocoonwhenitstimecomes;andthistimeisdeterminedbycauseswhichescapeournoticeandwhich,nodoubt,dependuponthepotentialitiesoftheeggitself。Itisthecasewiththeotherbramble—dwellerswhichIhavesubjectedtothesametest（Osmiadetrita,Anthidiumscapulare,Soleniusvagus,etc。）;anditmustalsobethecasewithOdynerusrubicola:sothemoststrikinganalogiesinformus。
ThereforethesingularexceptionwhichmadesuchanimpressiononDufour\'smindisasheerlogicalillusion。
Anerrorremovedistantamounttoatruthgained;andyet,ifitweretoendhere,theresultofmyexperimentwouldpossessbutslightvalue。Afterdestruction,letusturntoconstruction;andperhapsweshallfindthewherewithaltocompensateusforanillusionlost。Letusbeginbywatchingtheexit。
ThefirstOsmiatoleavehercocoon,nomatterwhatplacesheoccupiesintheseries,forthwithattackstheceilingseparatingherfromthefloorabove。Shecutsafairlycleanholeinit,shapedlikeatruncatecone,havingitslargerbaseonthesidewheretheBeeisanditssmallerbaseopposite。Thisconformationoftheexit—doorisacharacteristicofthework。Whentheinsecttriestoattackthediaphragm,itfirstdigsmoreorlessatrandom;then,astheboringprogresses,theactionisconcentrateduponanareawhichnarrowsuntilitpresentsnomorethanjustthenecessarypassage。Noristhecone—shapedaperturespecialtotheOsmia:Ihaveseenitmadebytheotherbramble—dwellersthroughmythickdisksofsorghum—pith。Undernaturalconditions,thepartitions,which,forthatmatter,areverythin,aredestroyedabsolutely,forthecontractionofthecellatthetopleavesbarelythewidthwhichtheinsectneeds。Thetruncate,cone—shapedbreachhasoftenbeenofgreatusetome。Itswidebasemadeitpossibleforme,withoutbeingpresentatthework,tojudgewhichofthetwoneighbouringOsmiaehadpiercedthepartition;ittoldmethedirectionofanocturnalmigrationwhichIhadbeenunabletowitness。
Thefirst—hatchedOsmia,wherevershemaybe,hasmadeaholeinherceiling。Sheisnowinthepresenceofthenextcocoon,withherheadattheopeningofthehole。Infrontofhersister\'scradle,sheusuallystops,consumedwithshyness;shedrawsbackintohercell,floundersamongtheshredsofthecocoonandthewreckageoftheruinedceiling;shewaitsaday,twodays,threedays,moreifnecessary。Shouldimpatiencegaintheupperhand,shetriestoslipbetweenthewallofthetunnelandthecocoonthatblockstheway。
Sheevenundertakesthelaboriousworkofgnawingatthewall,soastowidentheinterval,ifpossible。Wefindtheseattempts,intheshaftofabramble,atplaceswherethepithisremoveddowntotheverywood,wherethewooditselfisgnawedtosomedepth。Ineedhardlysaythat,althoughtheselateralinroadsareperceptibleaftertheevent,theyescapetheeyeatthemomentwhentheyarebeingmade。
Ifwewouldwitnessthem,wemustslightlymodifytheglassapparatus。Ilinetheinsideofthetubewithathickpieceofwhity—
brownpacking—paper,butonlyoveronehalfofthecircumference;theotherhalfisleftbare,sothatImaywatchtheOsmia\'sattempts。
Well,thecaptiveinsectfiercelyattacksthislining,whichtoitseyesrepresentsthepithylayerofitsusualabode;ittearsitawaybytinyparticlesandstrivestocutitselfaroadbetweenthecocoonandtheglasswall。Themales,whoarealittlesmaller,haveabetterchanceofsuccessthanthefemales。Flatteningthemselves,makingthemselvesthin,slightlyspoilingtheshapeofthecocoon,which,however,thankstoitselasticity,soonrecoversitsfirstcondition,theyslipthroughthenarrowpassageandreachthenextcell。Thefemales,wheninahurrytogetout,doasmuch,iftheyfindthetubeatallamenabletotheprocess。Butnosooneristhefirstpartitionpassedthanasecondpresentsitself。Thisispiercedinitsturn。Inthesamewaywillthethirdbepiercedandothersafterthat,iftheinsectcanmanagethem,aslongasitsstrengthholdsout。Tooweakfortheserepeatedborings,themalesdonotgofarthroughmythickplugs。Iftheycontrivetocutthroughthefirst,itisasmuchastheycando;and,evenso,theyarefarfromalwayssucceeding。But,intheconditionspresentedbythenativestalk,theyhaveonlyfeebletissuestoovercome;andthen,slipping,asIhavesaid,betweenthecocoonandthewall,whichisslightlywornowingtothecircumstancesdescribed,theyareabletopassthroughtheremainingoccupiedchambersandtoreachtheoutsidefirst,whatevertheiroriginalplaceinthestackofcells。Itisjustpossiblethattheirearlyeclosionforcesthismethodofexituponthem,amethodwhich,thoughoftenattempted,doesnotalwayssucceed。Thefemales,furnishedwithstrongertools,makegreaterprogressinmytubes。Iseesomewhopiercethreeorfourpartitions,oneaftertheother,andaresomanystagesaheadbeforethosewhomtheyhaveleftbehindareevenhatched。Whiletheyareengagedinthislongandtoilsomeoperation,others,nearertotheorifice,haveclearedapassagewhereofthosefromadistancewillavailthemselves。Inthisway,itmayhappenthat,whenthewidthofthetubepermits,anOsmiainabackrowwillneverthelessbeoneofthefirsttoemerge。
Inthebramble—stem,whichisofexactlythesamediameterasthecocoon,thisescapebythesideofthecolumnappearshardlypracticable,excepttoafewmales;andeventhesehavetofindawallwhichhassomuchpiththatbyremovingittheycaneffectapassage。Letusthenimagineatubesonarrowastopreventanyexitsaveinthenaturalsequenceofthecells。Whatwillhappen？Averysimplething。Thenewly—hatchedOsmia,afterperforatinghispartition,findshimselffacedwithanunbrokencocoonthatobstructstheroad。Hemakesafewattemptsuponthesidesand,realizinghisimpotence,retiresintohiscell,wherehewaitsfordaysanddays,untilhisneighbourburstshercocooninherturn。Hispatienceisinexhaustible。However,itisnotputtoanoverlongtest,forwithinaweek,moreorless,thewholestringoffemalesishatched。
WhentwoneighbouringOsmiaearereleasedatthesametime,mutualvisitsarepaidthroughtheaperturebetweenthetworooms:theoneabovegoesdowntothefloorbelow;theonebelowgoesuptothefloorabove;sometimesbothofthemareinthesamecelltogether。
Mightnotthisintercoursetendtocheerthemandencouragethemtopatience？Meanwhile,slowly,doorsareopeninghereandtherethroughtheseparatingwalls;theroadisclearedbysections;andamomentarriveswhentheleaderofthefilewalksout。Theothersfollow,ifready;buttherearealwayslaggardswhokeeptherear—rankswaitinguntiltheyaregone。
Tosumup,first,thehatchingofthelarvaetakesplacewithoutanyorder;secondly,theexodusproceedsregularlyfromsummittobase,butonlyinconsequenceoftheinsect\'sinabilitytomoveforwardsolongastheuppercellsarenotvacated。Wehaveherenotanexceptionalevolution,intheinverseratiotoage,butthesimpleimpossibilityofemergingotherwise。Shouldachanceoccurofgoingoutbeforeitsturn,theinsectdoesnotfailtoseizeit,aswecanseebythelateralmovementswhichsendtheimpatientonesafewranksaheadandevenreleasethemorefavouredaltogether。TheonlyremarkablethingthatIperceiveisthescrupulousrespectshowntotheasyetunopenedneighbouringcocoon。Howevereagertocomeout,theOsmiaismostcarefulnottotouchitwithhismandibles:itistaboo。Hewilldemolishthepartition,hewillgnawtheside—wallfiercely,eventhoughtherebenothingleftbutwood,hewillreduceeverythingaroundhimtodust;buttouchacocoonthatobstructshisway？Never！Hewillnotmakehimselfanoutletbybreakinguphissisters\'cradles。