Bythetimethatthebaseoftheeleventhcellisreached,thetransversalpartitionisoncemorealmostperpendiculartotheaxis。
Herewhathappenedatthebottomisrepeated。Thereisnolongitudinalpartition;andthespaciouscell,coveringthewholediameterofthecylinder,receivesafemale。Theedificeendswithtwotransversalpartitionsandonelongitudinalpartition,whichmarkout,onthesamelevel,chamberstwelveandthirteen,bothofwhichcontainmales。
Thereisnothingmorecuriousthanthismixingofthetwosexes,whenweknowwithwhatprecisiontheOsmiaseparatestheminalinearseries,wherethenarrowwidthofthecylinderdemandsthatthecellsshallbesetsingly,oneabovetheother。Here,theBeeismakinguseofatubewhosediameterisnotsuitedtoherwork;sheisconstructingacomplexanddifficultedifice,whichperhapswouldnotpossessthenecessarysolidityiftheceilingsweretoobroad。TheOsmiathereforesupportstheseceilingswithlongitudinalpartitions;
andtheunequalchambersresultingfromtheintroductionofthesepartitionsreceivefemalesatonetimeandmalesatanother,accordingtotheircapacity。
CHAPTER5。PERMUTATIONSOFSEX。
Thesexoftheeggisoptional。Thechoicerestswiththemother,whoisguidedbyconsiderationsofspaceand,accordingtotheaccommodationatherdisposal,whichisfrequentlyfortuitousandincapableofmodification,placesafemaleinthiscellandamaleinthat,sothatbothmayhaveadwellingofasizesuitedtotheirunequaldevelopment。ThisistheunimpeachableevidenceofthenumerousandvariedfactswhichIhavesetforth。Peopleunfamiliarwithinsectanatomy——thepublicforwhomIwrite——wouldprobablygivethefollowingexplanationofthismarvellousprerogativeoftheBee:
themotherhasatherdisposalacertainnumberofeggs,someofwhichareirrevocablyfemaleandtheothersirrevocablymale:sheisabletopickoutofeithergrouptheonewhichshewantsattheactualmoment;andherchoiceisdecidedbytheholdingcapacityofthecellthathastobestocked。Everythingwouldthenbelimitedtoajudiciousselectionfromtheheapofeggs。
Shouldthisideaoccurtohim,thereadermusthastentorejectit。
Nothingcouldbemorefalse,asthemerestreferencetoanatomywillshow。ThefemalereproductiveapparatusoftheHymenopteraconsistsgenerallyofsixovariantubes,somethinglikeglove—fingers,dividedintobunchesofthreeandendinginacommoncanal,theoviduct,whichcarriestheeggsoutside。Eachoftheseglove—fingersisfairlywideatthebase,buttaperssharplytowardsthetip,whichisclosed。Itcontains,arrangedinarow,oneaftertheother,likebeadsonastring,acertainnumberofeggs,fiveorsixforinstance,ofwhichtheloweronesaremoreorlessdeveloped,themiddleoneshalf—waytowardsmaturity,andtheupperonesveryrudimentary。Everystageofevolutionishererepresented,distributedregularlyfrombottomtotop,fromthevergeofmaturitytothevagueoutlinesoftheembryo。Thesheathclaspsitsstringofovulessocloselythatanyinversionoftheorderisimpossible。
Besides,aninversionwouldresultinagrossabsurdity:thereplacingofaripereggbyanotherinanearlierstageofdevelopment。
Therefore,ineachovariantube,ineachglove—finger,theemergenceoftheeggsoccursaccordingtotheordergoverningtheirarrangementinthecommonsheath;andanyothersequenceisabsolutelyimpossible。Moreover,atthenestingperiod,thesixovariansheaths,onebyoneandeachinitsturn,haveattheirbaseaneggwhichinaveryshorttimeswellsenormously。Somehoursorevenadaybeforethelaying,thateggbyitselfrepresentsorevenexceedsinbulkthewholeoftheovigenousapparatus。Thisistheeggwhichisonthepointofbeinglaid。Itisabouttodescendintotheoviduct,initsproperorder,atitspropertime;andthemotherhasnopowertomakeanothertakeitsplace。Itisthisegg,necessarilythiseggandnoother,thatwillpresentlybelaidupontheprovisions,whetherthesebeamessofhoneyoraliveprey;italoneisripe,italoneisattheentrancetotheoviduct;noneoftheothers,sincetheyarefartherbackintherowandnotattherightstageofdevelopment,canbesubstitutedatthiscrisis。Itsbirthisinevitable。
Whatwillityield,amaleorafemale？Nolodginghasbeenprepared,nofoodcollectedforit;andyetbothfoodandlodginghavetobeinkeepingwiththesexthatwillproceedfromit。Andhereisamuchmorepuzzlingcondition:thesexofthategg,whoseadventispredestined,hastocorrespondwiththespacewhichthemotherhappenstohavefoundforacell。Thereisthereforenoroomforhesitation,strangethoughthestatementmayappear:theegg,asitdescendsfromitsovariantube,hasnodeterminedsex。Itisperhapsduringthefewhoursofitsrapiddevelopmentatthebaseofitsovariansheath,itisperhapsonitspassagethroughtheoviductthatitreceives,atthemother'spleasure,thefinalimpressthatwillproduce,tomatchthecradlewhichithastofill,eitherafemaleoramale。
Thereuponthefollowingquestionpresentsitself。Letusadmitthat,whenthenormalconditionsremain,alayingwouldhaveyieldedmfemalesandnmales。Then,ifmyconclusionsarecorrect,itmustbeinthemother'spower,whentheconditionsaredifferent,totakefromthemgroupandincreasethengrouptothesameextent;itmustbepossibleforherlayingtoberepresentedasm—1,m—2,m—3,etc。
femalesandbyn1,n2,n3,etc。males,thesumofmnremainingconstant,butoneofthesexesbeingpartlypermutedintotheother。
Theultimateconclusionevencannotbedisregarded:wemustadmitasetofeggsrepresentedbym—m,orzero,femalesandofnmmales,oneofthesexesbeingcompletelyreplacedbytheother。Conversely,itmustbepossibleforthefeminineseriestobeaugmentedfromthemasculineseriestotheextentofabsorbingitentirely。ItwastosolvethisquestionandsomeothersconnectedwithitthatI
undertook,forthesecondtime,toreartheThree—hornedOsmiainmystudy。
Theproblemonthisoccasionisamoredelicateone;butIamalsobetter—equipped。Myapparatusconsistsoftwosmall,closedpacking—
cases,withthefrontsideofeachpiercedwithfortyholes,inwhichIcaninsertmyglasstubesandkeeptheminahorizontalposition。I
thusobtainfortheBeesthedarknessandmysterywhichsuittheirworkandformyselfthepowerofwithdrawingfrommyhive,atanytime,anytubethatIwish,withtheOsmiainside,soastocarryittothelightandfollow,ifneedbewiththeaidofthelens,theoperationsofthebusyworker。Myinvestigations,howeverfrequentandminute,innowayhinderthepeaceableBee,whoremainsabsorbedinhermaternalduties。
Imarkaplentifulnumberofmyguestswithavarietyofdotsonthethorax,whichenablesmetofollowanyoneOsmiafromthebeginningtotheendofherlaying。Thetubesandtheirrespectiveholesarenumbered;alist,alwayslyingopenonmydesk,enablesmetonotefromdaytoday,sometimesfromhourtohour,whathappensineachtubeandparticularlytheactionsoftheOsmiaewhosebacksbeardistinguishingmarks。Assoonasonetubeisfilled,Ireplaceitbyanother。Moreover,IhavescatteredinfrontofeitherhiveafewhandfulsofemptySnail—shells,speciallychosenfortheobjectwhichIhaveinview。ReasonswhichIwillexplainlaterledmetoprefertheshellsofHelixcaespitum。Eachoftheshells,asandwhenstocked,receivedthedateofthelayingandthealphabeticalsigncorrespondingwiththeOsmiatowhomitbelonged。Inthisway,I
spentfiveorsixweeksincontinualobservation。Tosucceedinanenquiry,thefirstandforemostconditionispatience。ThisconditionIfulfilled;anditwasrewardedwiththesuccesswhichIwasjustifiedinexpecting。
Thetubesemployedareoftwokinds。Thefirst,whicharecylindricalandofthesamewidththroughout,willbeofuseforconfirmingthefactsobservedinthefirstyearofmyexperimentsinindoorrearing。
Theothers,themajority,consistoftwocylinderswhichareofverydifferentdiameters,setendtoend。Thefrontcylinder,theonewhichprojectsalittlewayoutsidethehiveandformstheentrance—
hole,variesinwidthbetween8and12millimetres。（Between。312to。468inch。——Translator'sNote。）Thesecond,thebackone,containedentirelywithinmypacking—case,isclosedatitsfarendandis5to6millimetres（。195to。234inch。——Translator'sNote。）indiameter。
Eachofthetwopartsofthedouble—galleriedtunnel,onenarrowandonewide,measuresatmostadecimetre（3。9inches。——Translator'sNote。）inlength。Ithoughtitadvisabletohavetheseshorttubes,astheOsmiaisthuscompelledtoselectdifferentlodgings,eachofthembeinginsufficientinitselftoaccommodatethetotallaying。InthiswayIshallobtainagreatervarietyinthedistributionofthesexes。Lastly,atthemouthofeachtube,whichprojectsslightlyoutsidethecase,thereisalittlepapertongue,formingasortofperchonwhichtheOsmiaalightsonherarrivalandgivingeasyaccesstothehouse。Withthesefacilities,theswarmcolonizedfifty—twodouble—galleriedtubes,thirty—sevencylindricaltubes,seventy—eightSnail—shellsandafewoldnestsoftheMason—beeoftheShrubs。FromthisrichmineofmaterialIwilltakewhatIwanttoprovemycase。
Everyseries,evenwhenincomplete,beginswithfemalesandendswithmales。TothisruleIhavenotyetfoundanexception,atleastingalleriesofnormaldiameter。Ineachnewabode,themotherbusiesherselffirstofallwiththemoreimportantsex。Bearingthispointinmind,woulditbepossibleforme,bymanoeuvring,toobtainaninversionofthisorderandmakethelayingbeginwithmales？Ithinkso,fromtheresultsalreadyascertainedandtheirresistibleconclusionstobedrawnfromthem。Thedouble—galleriedtubesareinstalledinordertoputmyconjecturestotheproof。
Thebackgallery,5or6millimetres（。195to。234inch。——
Translator'sNote。）wide,istoonarrowtoserveasalodgingfornormallydevelopedfemales。If,therefore,theOsmia,whoisveryeconomicalofherspace,wishestooccupythem,shewillbeobligedtoestablishmalesthere。Andherlayingmustnecessarilybeginhere,becausethiscorneristherear—mostpartofthetube。Theforemostgalleryiswide,withanentrance—dooronthefrontofthehive。
Here,findingtheconditionstowhichsheisaccustomed,themotherwillgoonwithherlayingintheorderwhichsheprefers。
Letusnowseewhathashappened。Ofthefifty—twodoublegalleriedtubes,aboutathirddidnothavetheirnarrowpassagecolonized。TheOsmiacloseditsaperturecommunicatingwiththelargepassage;andthelatteralonereceivedtheeggs。Thiswasteofspacewasinevitable。ThefemaleOsmiae,thoughnearlyalwayslargerthanthemales,presentmarkeddifferencesamongoneanother:somearebigger,somearesmaller。IhadtoadjustthewidthofthenarrowgalleriestoBeesofaveragedimensions。Itmayhappenthereforethatagalleryistoosmalltoadmitthelarge—sizedmotherstowhomchanceallotsit。WhentheOsmiaisunabletoenterthetube,obviouslyshewillnotcolonizeit。Shethenclosestheentrancetothisspacewhichshecannotuseanddoesherlayingbeyondit,inthewidetube。HadI
triedtoavoidtheseuselessapparatusbychoosingtubesoflargercalibre,Ishouldhaveencounteredanotherdrawback:themedium—sizedmothers,findingthemselvesalmostcomfortable,wouldhavedecidedtolodgefemalesthere。Ihadtobepreparedforit:aseachmotherselectedherhouseatwillandasIwasunabletointerfereinherchoice,anarrowtubewouldbecolonizedornot,accordingastheOsmiawhoowneditwasorwasnotabletomakeherwayinside。
Thereremainsomefortypairsoftubeswithbothgalleriescolonized。
Inthesetherearetwothingstotakeintoconsideration。Thenarrowreartubesof5or51/2millimetres（。195to。214inch。——
Translator'sNote。）——andthesearethemostnumerous——containmalesandmalesonly,butinshortseries,betweenoneandfive。Themotherisheresomuchhamperedinherworkthattheyarerarelyoccupiedfromendtoend;theOsmiaseemsinahurrytoleavethemandtogoandcolonizethefronttube,whoseamplespacewillleaveherthelibertyofmovementnecessaryforheroperations。Theotherreartubes,theminority,whosediameterisabout6millimetres（。234
inch。——Translator'sNote。）,containsometimesonlyfemalesandsometimesfemalesatthebackandmalestowardstheopening。Onecanseethatatubeatriflewiderandamotherslightlysmallerwouldaccountforthisdifferenceintheresults。Nevertheless,asthenecessaryspaceforafemaleisbarelyprovidedinthiscase,weseethatthemotheravoidsasfarasshecanatwo—sexarrangementbeginningwithmalesandthatsheadoptsitonlyinthelastextremity。Finally,whateverthecontentsofthesmalltubemaybe,thoseofthelargeone,followinguponit,nevervaryandconsistoffemalesatthebackandmalesinfront。
Thoughincomplete,becauseofcircumstancesverydifficulttocontrol,theresultoftheexperimentisnonethelessverystriking。
Twenty—fiveapparatuscontainonlymalesintheirnarrowgallery,innumbersvaryingfromaminimumofonetoamaximumoffive。Afterthesecomesthecolonyofthelargegallery,beginningwithfemalesandendingwithmales。Andthelayingsintheseapparatusdonotalwaysbelongtolatesummeroreventotheintermediateperiod:afewsmalltubescontaintheearliesteggsoftheOsmiae。AcoupleofOsmiae,moreforwardthantheothers,settoworkonthe23rdofApril。Bothofthemstartedtheirlayingbyplacingmalesinthenarrowtubes。Themeagresupplyofprovisionswasenoughinitselftoshowthesex,whichprovedlatertobeinaccordancewithmyanticipations。Weseethenthat,bymyartifices,thewholeswarmstartswiththeconverseofthenormalorder。Thisinversioniscontinued,atnomatterwhatperiod,fromthebeginningtotheendoftheoperations。Theserieswhich,accordingtorule,wouldbeginwithfemalesnowbeginswithmales。Oncethelargergalleryisreached,thelayingispursuedintheusualorder。
Wehaveadvancedonestepandthatnosmallone:wehaveseenthattheOsmia,whencircumstancesrequireit,iscapableofreversingthesequenceofthesexes。Woulditbepossible,providedthatthetubewerelongenough,toobtainacompleteinversion,inwhichtheentireseriesofthemalesshouldoccupythenarrowgalleryatthebackandtheentireseriesofthefemalestheroomygalleryinfront？Ithinknot;andIwilltellyouwhy。
LongandnarrowcylindersarebynomeanstotheOsmia'staste,notbecauseoftheirnarrownessbutbecauseoftheirlength。Rememberthatforeachloadofhoneybroughttheworkerisobligedtomovebackwardstwice。Sheenters,headfirst,tobeginbydisgorgingthehoney—syrupfromhercrop。Unabletoturninapassagewhichsheblocksentirely,shegoesoutbackwards,crawlingratherthanwalking,alaboriousperformanceonthepolishedsurfaceoftheglassandaperformancewhich,withanyothersurface,wouldstillbeveryawkward,asthewingsareboundtorubagainstthewallwiththeirfreeendandareliabletogetrumpledorbent。Shegoesoutbackwards,reachestheoutside,turnsroundandgoesinagain,butthistimetheoppositeway,soastobrushofftheloadofpollenfromherabdomenontotheheap。Ifthegalleryisatalllong,thiscrawlingbackwardsbecomestroublesomeafteratime;andtheOsmiasoonabandonsapassagethatistoosmalltoallowoffreemovement。
Ihavesaidthatthenarrowtubesofmyapparatusare,forthemostpart,onlyveryincompletelycolonized。TheBee,afterlodgingasmallnumberofmalesinthem,hastenstoleavethem。Inthewidefrontgallery,shecanstaywheresheisandstillbeabletoturnroundeasilyforherdifferentmanipulations;shewillavoidthosetwolongjourneysbackwards,whicharesoexhaustingandsobadforherwings。
Anotherreasonnodoubtpromptshernottomaketoogreatauseofthenarrowpassage,inwhichshewouldestablishmales,followedbyfemalesinthepartwherethegallerywidens。Themaleshavetoleavetheircellsacoupleofweeksormorebeforethefemales。Iftheyoccupythebackofthehouse,theywilldieprisonersorelsetheywilloverturneverythingontheirwayout。ThisriskisavoidedbytheorderwhichtheOsmiaadopts。
Inmytubeswiththeirunusualarrangement,themothermightwellfindthedilemmaperplexing:thereisthenarrownessofthespaceatherdisposalandthereistheemergencelateron。Inthenarrowtubes,thewidthisinsufficientforthefemales;ontheotherhand,ifshelodgesmalesthere,theyareliabletoperish,sincetheywillbepreventedfromissuingatthepropermoment。Thiswouldperhapsexplainthemother'shesitationandherobstinacyinsettlingfemalesinsomeofmyapparatuswhichlookedasiftheycouldsuitnonebutmales。
Asuspicionoccurstome,asuspicionarousedbymyattentiveexaminationofthenarrowtubes。All,whateverthenumberoftheirinmates,arecarefullypluggedattheopening,justasseparatetubeswouldbe。ItmightthereforebethecasethatthenarrowgalleryatthebackwaslookeduponbytheOsmianotastheprolongationofthelargefrontgallery,butasanindependenttube。Thefacilitywithwhichtheworkerturnsassoonasshereachesthewidetube,herlibertyofaction,whichisnowasgreatasinadoorwaycommunicatingwiththeouterair,mightwellbemisleadingandcausetheOsmiatotreatthenarrowpassageatthebackasthoughthewidepassageinfrontdidnotexist。Thiswouldaccountfortheplacingofthefemaleinthelargetubeabovethemalesinthesmalltube,anarrangementcontrarytohercustom。
Iwillnotundertaketodecidewhetherthemotherreallyappreciatesthedangerofmysnares,orwhethershemakesamistakeinconsideringonlythespaceatherdisposalandbeginningwithmales。
Atanyrate,Iperceiveinheratendencytodeviateaslittleaspossiblefromtheorderwhichsafeguardstheemergenceofthetwosexes。Thistendencyisdemonstratedbyherrepugnancetocolonizingmynarrowtubeswithlongseriesofmales。However,sofarasweareconcerned,itdoesnotmattermuchwhatpassesatsuchtimesintheOsmia'slittlebrain。Enoughforustoknowthatshedislikesnarrowandlongtubes,notbecausetheyarenarrow,butbecausetheyareatthesametimelong。
And,infact,shedoesverywellwithashorttubeofthesamediameter。SucharethecellsintheoldnestsoftheMason—beeoftheShrubsandtheemptyshellsoftheGardenSnail。Withtheshorttube,thetwodisadvantagesofthelongtubeareavoided。ShehasverylittleofthatcrawlingbackwardstodowhenshehasaSnail—shellforthehomeofhereggsandscarcelyanywhenthehomeisthecelloftheMason—bee。Moreover,asthestackofcocoonsnumberstwoorthreeatmost,thedeliverancewillbeexemptfromthedifficultiesattachedtoalongseries。TopersuadetheOsmiatonidifyinasingletubelongenoughtoreceivethewholeofherlayingandatthesametimenarrowenoughtoleaveheronlyjustthepossibilityofadmittanceappearstomeaprojectwithouttheslightestchanceofsuccess:theBeewouldstubbornlyrefusesuchadwellingorwouldcontentherselfwithentrustingonlyaverysmallportionofhereggstoit。Ontheotherhand,withnarrowbutshortcavities,success,withoutbeingeasy,seemstomeatleastquitepossible。Guidedbytheseconsiderations,Iembarkeduponthemostarduouspartofmyproblem:toobtainthecompleteoralmostcompletepermutationofonesexwiththeother;toproducealayingconsistingonlyofmalesbyofferingthemotheraseriesoflodgingssuitedonlytomales。
LetusinthefirstplaceconsulttheoldnestsoftheMason—beeoftheShrubs。Ihavesaidthatthesemortarspheroids,piercedalloverwithlittlecylindricalcavities,areadoptedprettyeagerlybytheThree—hornedOsmia,whocolonizesthembeforemyeyeswithfemalesinthedeepcellsandmalesintheshallowcells。Thatishowthingsgowhentheoldnestremainsinitsnaturalstate。Withagrater,however,Iscrapetheoutsideofanothernestsoastoreducethedepthofthecavitiestosometenmillimetres。（Abouttwo—fifthsofaninch。——Translator'sNote。）Thisleavesineachcelljustroomforonecocoon,surmountedbytheclosingstopper。Ofthefourteencavitiesinthenests,Ileavetwointact,measuringfifteenmillimetresindepth。（。585inch。——Translator'sNote。）Nothingcouldbemorestrikingthantheresultofthisexperiment,madeinthefirstyearofmyhomerearing。Thetwelvecavitieswhosedepthhadbeenreducedallreceivedmales;thetwocavitiesleftuntouchedreceivedfemales。
AyearpassesandIrepeattheexperimentwithanestoffifteencells;butthistimeallthecellsarereducedtotheminimumdepthwiththegrater。Well,thefifteencells,fromfirsttolast,areoccupiedbymales。Itmustbequiteunderstoodthat,ineachcase,alltheoffspringbelongedtoonemother,markedwithherdistinguishingspotandkeptinsightaslongasherlayinglasted。
Hewouldindeedbedifficulttopleasewhorefusedtobowbeforetheresultsofthesetwoexperiments。If,however,heisnotyetconvinced,hereissomethingtoremovehislastdoubts。
TheThree—hornedOsmiaoftensettlesherfamilyinoldshells,especiallythoseoftheCommonSnail（Helixaspersa）,whoissocommonunderthestone—heapsandinthecrevicesofthelittleunmortaredwallsthatsupportourterraces。Inthisspecies,thespiraliswideopen,sothattheOsmia,penetratingasfardownasthehelicalpassagepermits,finds,immediatelyabovethepointwhichistoonarrowtopass,thespacenecessaryforthecellofafemale。
Thiscellissucceededbyothers,widerstill,alwaysforfemales,arrangedinalineinthesamewayasinastraighttube。Inthelastwhorlofthespiral,thediameterwouldbetoogreatforasinglerow。Thenlongitudinalpartitionsareaddedtothetransversepartitions,thewholeresultingincellsofunequaldimensionsinwhichmalespredominate,mixedwithafewfemalesinthelowerstoreys。Thesequenceofthesexesisthereforewhatitwouldbeinastraighttubeandespeciallyinatubewithawidebore,wherethepartitioningiscomplicatedbysubdivisionsonthesamelevel。A
singleSnail—shellcontainsroomforsixoreightcells。Alarge,roughearthenstopperfinishesthenestattheentrancetotheshell。
Asadwellingofthissortcouldshowusnothingnew,IchoseformyswarmtheGardenSnail（Helixcaespitum）,whoseshell,shapedlikeasmall,swollenAmmonite,widensbyslowdegrees,thediameteroftheusableportion,rightuptothemouth,beinghardlygreaterthanthatrequiredbyamaleOsmia—cocoon。Moreover,thewidestpart,inwhichafemalemightfindroom,hastoreceiveathickstopping—plug,belowwhichtherewilloftenbeafreespace。Underalltheseconditions,thehousewillhardlysuitanybutmalesarrangedoneaftertheother。
Thecollectionofshellsplacedatthefootofeachhiveincludesspecimensofdifferentsizes。Thesmallestare18millimetres（。7
inch。——Translator'sNote。）indiameterandthelargest24millimetres（。936inch。——Translator'sNote。）Thereisroomfortwococoons,orthreeatmost,accordingtotheirdimensions。
Nowtheseshellswereusedbymyvisitorswithoutanyhesitation,perhapsevenwithmoreeagernessthantheglasstubes,whoseslipperysidesmighteasilybealittleannoyingtotheBee。Someofthemwereoccupiedonthefirstfewdaysofthelaying;andtheOsmiawhohadstartedwithahomeofthissortwouldpassnexttoasecondSnail—
shell,intheimmediateneighbourhoodofthefirst,toathird,afourthandothersstill,alwaysclosetooneanother,untilherovarieswereemptied。ThewholefamilyofonemotherwouldthusbelodgedinSnail—shellswhichweredulymarkedwiththedateofthelayingandadescriptionoftheworker。ThefaithfuladherentsoftheSnail—shellwereintheminority。Thegreaternumberleftthetubestocometotheshellsandthenwentbackfromtheshellstothetubes。All,afterfillingthespiralstaircasewithtwoorthreecells,closedthehousewithathickearthenstopperonalevelwiththeopening。Itwasalongandtroublesometask,inwhichtheOsmiadisplayedallherpatienceasamotherandallhertalentsasaplasterer。Therewereevensomewho,scrupuloustoexcess,carefullycementedtheumbilicus,aholewhichseemedtoinspirethemwithdistrustasbeingabletogiveaccesstotheinteriorofthedwelling。Itwasadangerous—lookingcavity,whichforthegreatersafetyofthefamilyitwasprudenttoblockup。
Whenthepupaearesufficientlymatured,Iproceedtoexaminetheseelegantabodes。Thecontentsfillmewithjoy:theyfulfilmyanticipationstotheletter。Thegreat,theverygreatmajorityofthecocoonsturnouttobemales;hereandthere,inthebiggercells,afewrarefemalesappear。Thesmallnessofthespacehasalmostdoneawaywiththesixty—eightSnail—shellscolonized。But,ofthistotalnumber,ImustuseonlythoseserieswhichreceivedanentirelayingandwereoccupiedbythesameOsmiafromthebeginningtotheendoftheegg—season。Hereareafewexamples,takenfromamongthemostconclusive。
>Fromthe6thofMay,whenshestartedoperations,tothe25thofMay,thedateatwhichherlayingceased,theOsmiaoccupiedsevenSnail—
shellsinsuccession。Herfamilyconsistsoffourteencocoons,anumberveryneartheaverage;and,ofthesefourteencocoons,twelvebelongtomalesandonlytwotofemales。Theseoccupytheseventhandthirteenthplacesinchronologicalorder。
Another,betweenthe9thand27thofMay,stockedsixSnail—shellswithafamilyofthirteen,includingtenmalesandthreefemales。Theplacesoccupiedbythelatterintheserieswerenumbers3,4and5。
Athird,betweenthe2ndand29thofMay,colonizedelevenSnail—
shells,aprodigioustask。Thisindustriousonewasalsoexceedinglyprolific。Shesuppliedmewithafamilyoftwenty—six,thelargestwhichIhaveeverobtainedfromoneOsmia。Well,thisabnormalprogenyconsistedoftwenty—fivemalesandonefemale,onealone,occupyingplace17。
Thereisnoneedtogoon,afterthismagnificentexample,especiallyastheotherserieswouldall,withoutexception,giveusthesameresult。Twofactsareimmediatelyobvious。TheOsmiaisabletoreversetheorderofherlayingandtostartwithamoreorlesslongseriesofmalesbeforeproducinganyfemales。Inthefirstcase,thefirstfemaleappearsasnumber7;inthethird,asnumber17。Thereissomethingbetterstill;andthisisthepropositionwhichIwasparticularlyanxioustoprove:thefemalesexcanbepermutedwiththemalesexandcanbepermutedtothepointofdisappearingaltogether。Weseethisespeciallyinthethirdcase,wherethepresenceofasolitaryfemaleinafamilyoftwenty—sixisduetothesomewhatlargerdiameterofthecorrespondingSnail—shellandalso,nodoubt,tosomemistakeonthemother'spart,forthefemalecocoon,inaseriesoftwo,occupiestheupperstorey,theonenexttotheorifice,anarrangementwhichtheOsmiaappearstometodislike。
ThisresultthrowssomuchlightononeofthedarkestcornersofbiologythatImustattempttocorroborateitbymeansofevenmoreconclusiveexperiments。IproposenextyeartogivetheOsmiaenothingbutSnail—shellsforalodging,pickedoutonebyone,andrigorouslytodeprivetheswarmofanyotherretreatinwhichthelayingcouldbeeffected。Undertheseconditions,Ioughttoobtainnothingbutmales,ornearly,forthewholeswarm。