Theoldmalecellscanberecognizedbytheirpositionontheouteredgesandbytheircapacity,measuringonanaveragethesameasacolumnofsand31millimetreshighinaglasstube5millimetreswide。（1。21x。195inches。——Translator'sNote。）Thesecellscontainmalesofthesecondorthirdgenerationandnonebutmales。Intheoldfemalecells,thoseinthemiddle,whosecapacityismeasuredbyasimilarcolumnofsand45millimetreshigh（1。75inches。——
　　Translator'sNote。）,arefemalesandnonebutfemales。
　　Thispresenceofbothsexesatatime,evenwhentherearebuttwocellsfree,onespaciousandtheothersmall,provesintheplainestfashionthattheregulardistributionobservedinthecompletenestsofrecentproductionisherereplacedbyanirregulardistribution,harmonizingwiththenumberandholding—capacityofthechamberstobestocked。TheMason—beehasbeforeher,letmesuppose,onlyfivevacantcells:twolargerandthreesmaller。Thetotalspaceatherdisposalwoulddoforaboutathirdofthelaying。Well,inthetwolargecells,sheputsfemales;inthethreesmallcells,sheputsmales。
　　Aswefindthesamesortofthinginalltheoldnests,wemustneedsadmitthatthemotherknowsthesexoftheeggwhichsheisgoingtolay,becausethateggisplacedinacellofthepropercapacity。Wecangofurtherandadmitthatthemotheralterstheorderofsuccessionofthesexesatherpleasure,becauseherlayings,betweenoneoldnestandanother,arebrokenupintosmallgroupsofmalesandfemalesaccordingtotheexigenciesofspaceintheactualnestwhichshehappenstobeoccupying。
　　Justnow,inthenewnest,wesawtheMason—beearranginghertotallayingintoseriesfirstoffemalesandnextofmales;andheresheis,mistressofanoldnestofwhichshehasnotthepowertoalterthearrangement,breakingupherlayingintosectionscomprisingbothsexesjustasrequiredbytheconditionsimposeduponher。Shethereforedecidesthesexoftheeggatwill,for,withoutthisprerogative,shecouldnot,inthechambersofthenestwhichsheowestochance,depositunerringlythesexforwhichthosechamberswereoriginallybuilt;andthishappenshoweversmallthenumberofchamberstobefilled。
　　Whenthenestisnew,IthinkIseeareasonwhytheMason—beeshouldseriateherlayingintofemalesandthenmales。Hernestisahalf—
　　sphere。ThatoftheMason—beeoftheShrubsisverynearlyasphere。
　　Ofallshapes,thesphericalshapeisthestrongest。Nowthesetwonestsrequireanexceptionalpowerofresistance。Withoutprotectionofanykind,theyhavetobravetheweather,oneonitspebble,theotheronitsbough。Theirsphericalconfigurationisthereforeverypractical。
　　ThenestoftheMason—beeoftheWallsconsistsofaclusterofuprightcellsbackingagainstoneanother。Forthewholetotakeasphericalform,theheightofthechambersmustdiminishfromthecentreofthedometothecircumference。Theirelevationisthesineofthemeridianarcstartingfromtheplaneofthepebble。Therefore,iftheyaretohaveanysolidity,theremustbelargecellsinthemiddleandsmallcellsattheedges。And,astheworkbeginswiththecentralchambersandendswiththoseonthecircumference,thelayingofthefemales,destinedforthelargecells,mustprecedethatofthemales,destinedforthesmallcells。Sothefemalescomefirstandthemalesatthefinish。
　　Thisisallverywellwhenthemotherherselffoundsthedwelling,whenshelaysthefirstrowsofbricks。But,whensheisinthepresenceofanoldnest,ofwhichsheisquiteunabletoalterthegeneralarrangement,howisshetomakeuseofthefewvacantrooms,thelargeandthesmallalike,ifthesexoftheeggbealreadyirrevocablyfixed？Shecanonlydosobyabandoningthearrangementintwoconsecutiverowsandaccommodatingherlayingtothevariedexigenciesofthehome。Eithershefindsitimpossibletomakeaneconomicaluseoftheoldnest,atheoryrefutedbytheevidence,orelseshedeterminesatwillthesexoftheeggwhichsheisabouttolay。
　　TheOsmiaethemselveswillfurnishthemostconclusiveevidenceonthelatterpoint。WehaveseenthattheseBeesarenotgenerallyminers,whothemselvesdigoutthefoundationoftheircells。Theymakeuseoftheoldstructuresofothers,orelseofnaturalretreats,suchashollowstems,thespiralsofemptyshellsandvarioushiding—placesinwalls,clayorwood。Theirworkisconfinedtorepairstothehouse,suchaspartitionsandcovers。Thereareplentyoftheseretreats;andtheinsectwouldalwaysfindfirst—
　　classonesifitthoughtofgoinganydistancetolookforthem。ButtheOsmiaisastay—at—home:shereturnstoherbirth—placeandclingstoitwithapatienceextremelydifficulttoexhaust。Itishere,inthislittlefamiliarcorner,thatshepreferstosettleherprogeny。Butthentheapartmentsarefewinnumberandofallshapesandsizes。Therearelongandshortones,spaciousonesandnarrow。
　　Shortofexpatriatingherself,aSpartancourse,shehastousethemall,fromfirsttolast,forshehasnochoice。Guidedbytheseconsiderations,IembarkedontheexperimentswhichIwillnowdescribe。
　　Ihavesaidhowmystudy,ontwoseparateoccasions,becameapopuloushive,inwhichtheThree—hornedOsmiabuilthernestsinthevariousapplianceswhichIhadpreparedforher。Amongtheseappliances,tubes,eitherofglassorreed,predominated。Thereweretubesofalllengthsandwidths。Inthelongtubes,entireoralmostentirelayings,withaseriesoffemalesfollowedbyaseriesofmales,weredeposited。AsIhavealreadyreferredtothisresult,I
　　willnotdiscussitagain。Theshorttubesweresufficientlyvariedinlengthtolodgeoneorotherportionofthetotallaying。Basingmycalculationsontherespectivelengthsofthecocoonsofthetwosexes,onthethicknessofthepartitionsandthefinallid,I
　　shortenedsomeofthesetotheexactdimensionsrequiredfortwococoonsonly,ofdifferentsexes。
　　Well,theseshorttubes,whetherofglassorreed,wereseizeduponaseagerlyasthelongtubes。Moreover,theyyieldedthissplendidresult:theircontents,onlyapartofthetotallaying,alwaysbeganwithfemaleandendedwithmalecocoons。Thisorderwasinvariable;
　　whatvariedwasthenumberofcellsinthelongtubesandtheproportionbetweenthetwosortsofcocoons,sometimesmalespredominatingandsometimesfemales。
　　Theexperimentisofparamountimportance;anditwillperhapsmaketheresultclearerifIquoteoneinstancefromamongamultitudeofsimilarcases。Igivethepreferencetothisparticularinstancebecauseoftheratherexceptionalfertilityofthelaying。AnOsmiamarkedonthethoraxiswatched,daybyday,fromthecommencementtotheendofherwork。Fromthe1sttothe10thofMay,sheoccupiesaglasstubeinwhichshelodgessevenfemalesfollowedbyamale,whichendstheseries。Fromthe10thtothe17thofMay,shecolonizesasecondtube,inwhichshelodgesfirstthreefemalesandthenthreemales。Fromthe17thtothe25thofMay,athirdtube,withthreefemalesandthentwomales。Onthe26thofMay,afourthtube,whichsheabandons,probablybecauseofitsexcessivewidth,afterlayingonefemaleinit。Lastly,fromthe26thtothe30thofMay,afifthtube,whichshecolonizeswithtwofemalesandthreemales。Total:twenty—fiveOsmiae,includingseventeenfemalesandeightmales。Anditwillnotbesuperfluoustoobservethattheseunfinishedseriesdonotinanywaycorrespondwithperiodsseparatedbyintervalsofrest。Thelayingiscontinuous,insofarasthevariableconditionoftheatmosphereallows。Assoonasonetubeisfullandclosed,anotherisoccupiedbytheOsmiawithoutdelay。
　　Thetubesreducedtotheexactlengthoftwocellsfulfilledmyexpectationinthegreatmajorityofcases:thelowercellwasoccupiedbyafemaleandtheupperbyamale。Therewereafewexceptions。MorediscerningthanIinherestimateofwhatwasstrictlynecessary,better—versedintheeconomyofspace,theOsmiahadfoundawayoflodgingtwofemaleswhereIhadonlyseenroomforonefemaleandamale。
　　Thisexperimentspeaksvolumes。Whenconfrontedwithtubestoosmalltoreceiveallherfamily,sheisinthesameplightastheMason—beeinthepresenceofanoldnest。ShethereuponactsexactlyastheChalicodomadoes。Shebreaksupherlaying,dividesitintoseriesasshortastheroomatherdisposaldemands;andeachseriesbeginswithfemalesandendswithmales。Thisbreakingup,ontheonehand,intosectionsinallofwhichbothsexesarerepresentedandthedivision,ontheotherhand,oftheentirelayingintojusttwogroups,onefemale,theothermale,whenthelengthofthetubepermits,surelyprovideuswithampleevidenceoftheinsect'spowertoregulatethesexoftheeggaccordingtotheexigenciesofspace。
　　Andbesidestheexigenciesofspaceonemightperhapsventuretoaddthoseconnectedwiththeearlierdevelopmentofthemales。Thesebursttheircocoonsacoupleofweeksormorebeforethefemales;
　　theyarethefirstwhohastentothesweetsofthealmond—tree。Inordertoreleasethemselvesandemergeintothegladsunlightwithoutdisturbingthestringofcocoonswhereintheirsistersarestillsleeping,theymustoccupytheupperendoftherow;andthis,nodoubt,isthereasonthatmakestheOsmiaendeachofherbrokenlayingswithmales。Beingnexttothedoor,theseimpatientoneswillleavethehomewithoutupsettingtheshellsthatareslowerinhatching。
　　IexperimentedonLatreille'sOsmia,usingshortandevenveryshortstumpsofreed。AllthatIhadtodowastolaythemjustbesidethenestsoftheMason—beeoftheSheds,nestsbelovedbythisparticularOsmia。Old,disusedhurdlessuppliedmewithreedsinhabitedfromendtoendbytheHornedOsmia。InbothcasesIobtainedthesameresultsandthesameconclusionsaswiththeThree—hornedOsmia。
　　Ireturntothelatter,nidifyingundermyeyesinsomeoldnestsoftheMason—beeoftheWalls,whichIhadplacedwithinherreach,mixedupwiththetubes。Outsidemystudy,IhadneveryetseentheThree—hornedOsmiaadoptthatdomicile。Thismaybeduetothefactthatthesenestsareisolatedonebyoneinthefields;andtheOsmia,wholovestofeelherselfsurroundedbyherkinandtoworkinplentyofcompany,refusesthembecauseofthisisolation。Butonmytable,findingthemclosetothetubesinwhichtheothersareworking,sheadoptsthemwithouthesitation。
　　ThechamberspresentedbythoseoldnestsaremoreorlessspaciousaccordingtothethicknessofthecoatofmortarwhichtheChalicodomahaslaidovertheassembledchambers。Toleavehercell,theMason—beehastoperforatenotonlytheplug,thelidbuiltatthemouthofthecell,butalsothethickplasterwherewiththedomeisstrengthenedattheendofthework。Theperforationresultsinavestibulewhichgivesaccesstothechamberitself。Itisthisvestibulewhichissometimeslongerandsometimesshorter,whereasthecorrespondingchamberisofalmostconstantdimensions,inthecaseofthesamesex,ofcourse。
　　Wewillfirstconsidertheshortvestibule,atthemostlargeenoughtoreceivetheplugwithwhichtheOsmiawillcloseupthelodging。
　　Thereisthennothingatherdisposalexceptthecellproper,aspaciousapartmentinwhichoneoftheOsmia'sfemaleswillfindampleaccommodation,forsheismuchsmallerthantheoriginaloccupantofthechamber,nomatterthesex;butthereisnotroomfortwococoonsatatime,especiallyinviewofthespacetakenupbytheinterveningpartition。Well,inthoselarge,well—builtchambers,formerlythehomesofChalicodomae,theOsmiasettlesfemalesandnonebutfemales。
　　Letusnowconsiderthelongvestibule。Here,apartitionisconstructed,encroachingslightlyonthecellproper,andtheresidenceisdividedintotwounequalstoreys,alargeroombelow,housingafemale,andanarrowcabinabove,containingamale。
　　Whenthelengthofthevestibulepermits,allowingforthespacerequiredbytheouterstopper,athirdstoreyisbuilt,smallerthanthesecond;andanothermaleislodgedinthiscrampedcorner。InthiswaytheoldnestoftheMason—beeofthePebblesiscolonized,cellaftercell,byasinglemother。
　　TheOsmia,aswesee,isveryfrugalofthelodgingthathasfallentohershare;shemakesthebestpossibleuseofit,givingtothefemalesthespaciouschambersoftheMason—beeandtothemalesthenarrowvestibules,subdividedintostoreyswhenthisisfeasible。
　　Economyofspaceisthechiefconsideration,sinceherstay—at—hometastesdonotallowhertoindulgeindistantquests。Shehastoemploythesitewhichchanceplacesatherdisposaljustasitis,nowforamaleandnowforafemale。Hereweseedisplayed,moreclearlythanever,herpowerofdecidingthesexoftheegg,inordertoadaptitjudiciouslytotheconditionsofthehouse—roomavailable。
　　IhadofferedatthesametimetotheOsmiaeinmystudysomeoldnestsoftheMason—beeoftheShrubs,whichareclayspheroidswithcylindricalcavitiesinthem。Thesecavitiesareformed,asintheoldnestsoftheMason—beeofthePebbles,ofthecellproperlyso—
　　calledandoftheexit—waywhichtheperfectinsectcutthroughtheoutercoatingatthetimeofitsdeliverance。Theirdiameterisaboutsevenmillimetres（。273inch。——Translator'sNote。）;theirdepthatthecentreoftheheapis23millimetres（。897inch。——Translator'sNote。）;andattheedgeaverages14millimetres（。546inch。——
　　Translator'sNote。）
　　ThedeepcentralcellsreceiveonlythefemalesoftheOsmia;
　　sometimeseventhetwosexestogether,withapartitioninthemiddle,thefemaleoccupyingthelowerandthemaletheupperstorey。
　　True,insuchcaseseconomyofspaceisstrainedtotheutmost,theapartmentsprovidedbytheMason—beeoftheShrubsbeingverysmallasitis,despitetheirentrance—halls。Lastly,thedeepercavitiesonthecircumferenceareallottedtofemalesandtheshallowertomales。
　　Iwilladdthatasinglemotherpeopleseachnestandalsothatsheproceedsfromcelltocellwithouttroublingtoascertainthedepth。
　　Shegoesfromthecentretotheedges,fromtheedgestothecentre,fromadeepcavitytoashallowcavityandviceversa,whichshewouldnotdoifthesexesweretofollowuponeachotherinasettledorder。Forgreatercertainty,Inumberedthecellsofonenestaseachofthemwasclosed。Onopeningthemlater,Iwasabletoseethatthesexeswerenotsubjectedtoachronologicalarrangement。
　　Femalesweresucceededbymalesandthesebyfemaleswithoutitsbeingpossibleformetomakeoutanyregularsequence。Only——andthisistheessentialpoint——thedeepcavitieswereallottedtothefemalesandtheshallowonestothemales。
　　WeknowthattheThree—hornedOsmiapreferstohauntthehabitationsoftheBeeswhonidifyinpopulouscolonies,suchastheMason—beeoftheShedsandtheHairy—footedAnthophora。Exercisingtheverygreatestcare,IbrokeupsomegreatlumpsofearthremovedfromthebanksinhabitedbytheAnthophoraandsenttomefromCarpentrasbymydearfriendandpupilM。Devillario。Iexaminedthemconscientiouslyinthequietofmystudy。IfoundtheOsmia'scocoonsarrangedinshortseries,inveryirregularpassages,theoriginalworkofwhichisduetotheAnthophora。Touchedupafterwards,madelargerorsmaller,lengthenedorshortened,intersectedwithanetworkofcrossingsbythenumerousgenerationsthathadsucceededoneanotherinthesamecity,theyformedaninextricablelabyrinth。
　　Sometimesthesecorridorsdidnotcommunicatewithanyadjoiningapartment;sometimestheygaveaccesstothespaciouschamberoftheAnthophora,whichcouldberecognized,inspiteofitsage,byitsovalshapeanditscoatingofglazedstucco。Inthelattercase,thebottomcell,whichonceconstituted,byitself,thechamberoftheAnthophora,wasalwaysoccupiedbyafemaleOsmia。Beyondit,inthenarrowcorridor,amalewaslodged,notseldomtwo,oreventhree。Ofcourse,claypartitions,theworkoftheOsmia,separatedthedifferentinhabitants,eachofwhomhadhisownstorey,hisownclosedcell。
　　Whentheaccommodationconsistedofnomorethanasimplecylinder,withnostate—bedroomattheendofit——abedroomalwaysreservedforafemale——thecontentsvariedwiththediameterofthecylinder。Theseries,ofwhichthelongestwereseriesoffour,included,withawiderdiameter,firstoneortwofemales,thenoneortwomales。Italsohappened,thoughrarely,thattheserieswasreversed,thatistosay,itbeganwithmalesandendedwithfemales。Lastly,therewereagoodmanyisolatedcocoons,ofonesexortheother。WhenthecocoonwasaloneandoccupiedtheAnthophora'scell,itinvariablybelongedtoafemale。
　　IhaveobservedthesamethinginthenestsoftheMason—beeoftheSheds,butnotsoeasily。Theseriesareshorterhere,becausetheMason—beedoesnotboregalleriesbutbuildscelluponcell。Theworkofthewholeswarmthusformsastratumofcellsthatgrowsthickerfromyeartoyear。ThecorridorsoccupiedbytheOsmiaaretheholeswhichtheMason—beeduginordertoreachdaylightfromthedeeplayers。Intheseshortseries,bothsexesareusuallypresent;and,iftheMason—bee'schamberisattheendofthepassage,itisinhabitedbyafemaleOsmia。
　　WecomebacktowhattheshorttubesandtheoldnestsoftheMason—
　　beeofthePebbleshavealreadytaughtus。TheOsmiawho,intubesofsufficientlength,dividesherwholelayingintoacontinuoussequenceoffemalesandacontinuoussequenceofmales,nowbreaksitupintoshortseriesinwhichbothsexesarepresent。Sheadaptshersectionallayingstotheexigenciesofachancelodging;shealwaysplacesafemaleinthesumptuouschamberwhichtheMason—beeortheAnthophoraoccupiedoriginally。
　　FactsevenmorestrikingaresuppliedbytheoldnestsoftheMaskedAnthophora（A。personata,ILLIG。）,oldnestswhichIhaveseenutilizedbytheHornedOsmiaandtheThree—hornedOsmiaatthesametime。Lessfrequently,thesamenestsserveforLatreille'sOsmia。
　　LetusfirstdescribetheMaskedAnthophora'snests。
　　Inasteepbankofsandyclay,wefindasetofround,wide—openholes。Therearegenerallyonlyafewofthem,eachabouthalfaninchindiameter。Theyaretheentrance—doorsleadingtotheAnthophora'sabode,doorsalwaysleftopen,evenafterthebuildingisfinished。Eachofthemgivesaccesstoashortpassage,sometimesstraight,sometimeswinding,nearlyhorizontal,polishedwithminutecareandvarnishedwithasortofwhiteglaze。Itlooksasifithadreceivedathincoatofwhitewash。Ontheinnersurfaceofthispassage,inthethicknessoftheearthybank,spaciousovalnicheshavebeenexcavated,communicatingwiththecorridorbymeansofanarrowbottle—neck,whichisclosed,whentheworkisdone,withasubstantialmortarstopper。TheAnthophorapolishestheoutsideofthisstoppersowell,smoothsitssurfacesoperfectly,bringingittothesamelevelasthatofthepassage,issocarefultogiveitthewhitetintoftherestofthewallthat,whenthejobisfinished,itbecomesabsolutelyimpossibletodistinguishtheentrance—doorcorrespondingwitheachcell。
　　Thecellisanovalcavitydugintheearthymass。Thewallhasthesamepolish,thesamechalkywhitenessasthegeneralpassage。ButtheAnthophoradoesnotlimitherselftodiggingovalniches:tomakeherworkmoresolid,shepoursoverthewallsofthechamberasalivaryliquidwhichnotonlywhitensandvarnishesbutalsopenetratestoadepthofsomemillimetresintothesandyearth,whichitturnsintoahardcement。Asimilarprecautionistakenwiththepassage;andthereforethewholeisasolidpieceofworkcapableofremaininginexcellentconditionforyears。
　　Moreover,thankstothewallhardenedbythesalivaryfluid,thestructurecanberemovedfromitsmatrixbychippingitcarefullyaway。Wethusobtain,atleastinfragments,aserpentinetubefromwhichhangsasingleordoublerowofovalnodulesthatlooklikelargegrapesdrawnoutlengthwise。Eachofthesenodulesisacell,theentrancetowhich,carefullyhidden,opensintothetubeorpassage。Whenshewishestoleavehercell,inthespring,theAnthophoradestroysthemortardiskthatclosesthejarandthusreachesthegeneralcorridor,whichisquiteopentotheouterair。
　　Theabandonednestprovidesaseriesofpear—shapedcavities,ofwhichthedistendedpartistheoldcellandthecontractedparttheexit—neck,ridofitsstopper。
　　Thesepear—shapedhollowsformsplendidlodgings,impregnablestrongholds,inwhichtheOsmiaefindasafeandcommodiousretreatfortheirfamilies。TheHornedOsmiaandtheThree—hornedOsmiaestablishthemselvesthereatthesametime。Althoughitisalittletoolargeforher,Latrielle'sOsmiaalsoappearsverywellsatisfiedwithit。
　　Ihaveexaminedsomefortyofthesuperbcellsutilizedbyeachofthefirsttwo。Thegreatmajorityaredividedintotwostoreysbymeansofatransversalpartition。ThelowerstoreyincludesthelargerportionoftheAnthophora'scell;theupperstoreyincludestherestofthecellandalittleofthebottle—neckthatsurmountsit。Thetwo—roomeddwellingisclosed,inthepassage,byashapeless,bulkymassofdriedmud。WhataclumsyartisttheOsmiais,comparedwiththeAnthophora！AgainsttheexquisiteworkoftheAnthophora,partitionandplugstrikeanoteashideouslyincongruousasalumpofdirtonpolishedmarble。
　　Thetwoapartmentsthusobtainedareofaveryunequalcapacity,whichatoncestrikestheobserver。Imeasuredthemwithmyfive—
　　millimetretube。Onanaverage,thebottomoneisrepresentedbyacolumnofsand50millimetresdeep（1。95inches。——Translator'sNote。）
　　andthetoponebyacolumnof15millimetres（。585inch。——
　　Translator'sNote。）。Theholding—capacityoftheoneisthereforeaboutthreetimesaslargeasthatoftheother。Thecocoonsenclosedpresentthesamedisparity。Thebottomoneisbig,thetoponesmall。
　　Lastly,theloweronebelongstoafemaleOsmiaandtheuppertoamaleOsmia。
　　Occasionallythelengthofthebottle—neckallowsofafresharrangementandthecavityisdividedintothreestoreys。Thebottomone,whichisalwaysthemostspacious,containsafemale;thetwoabove,bothsmallerthanthefirstandonesmallerthantheother,containmales。
　　Letuskeeptothefirstcase,whichisalwaysthemostfrequent。TheOsmiaisinthepresenceofoneofthesepear—shapedhollows。Itisafindthatmustbeemployedtothebestadvantage:aprizeofthissortisrareandfallsonlytofortune'sfavourites。Tolodgetwofemalesinitatonceisimpossible;thereisnotsufficientroom。Tolodgetwomalesinitwouldbeunduegenerositytoasexthatisentitledtobutthesmallestconsideration。Besides,thetwosexesmustberepresentedinalmostequalnumbers。TheOsmiadecidesupononefemale,whoseportionshallbethebetterroom,thelowerone,whichislarger,better—protectedandmorenicelypolished,andonemale,whoseportionshallbetheupperstorey,acrampedattic,unevenandruggedinthepartwhichencroachesonthebottle—neck。
　　Thisdecisionisprovedbynumerousundeniablefacts。BothOsmiaethereforecanchoosethesexoftheeggabouttobelaid,seeingthattheyarenowbreakingupthelayingintogroupsoftwo,afemaleandamale,asrequiredbytheconditionsofthelodging。
　　IhaveonlyoncefoundLatreille'sOsmiaestablishedinthenestoftheMaskedAnthophora。Shehadoccupiedbutasmallnumberofcells,becausetheotherswerenotfree,beinginhabitedbytheAnthophora。
　　Thecellsinquestionweredividedintothreestoreysbypartitionsofgreenmortar;thelowerstoreywasoccupiedbyafemale,thetwoothersbymales,withsmallercocoons。
　　Icametoanevenmoreremarkableexample。TwoAnthidiaofmydistrict,A。septemdentatum,LATR。,andA。bellicosum,LEP。,adoptasthehomeoftheiroffspringtheemptyshellsofdifferentsnails:
　　Helixaspersa,H。algira,H。nemoralis,H。caespitum。Thefirst—
　　named,theCommonSnail,isthemostoftenused,underthestone—
　　heapsandinthecrevicesofoldwalls。BothAnthidiacolonizeonlythesecondwhorlofthespiral。Thecentralpartistoosmallandremainsunoccupied。Evensowiththefrontwhorl,thelargest,whichisleftcompletelyempty,somuchsothat,onlookingthroughtheopening,itisimpossibletotellwhethertheshelldoesordoesnotcontaintheBee'snest。Wehavetobreakthislastwhorlifwewouldperceivethecuriousnesttuckedawayinthespiral。
　　Wethenfindfirstatransversalpartition,formedoftinybitsofgravelcementedbyaputtymadefromresin,whichiscollectedinfreshdropsfromtheoxycedrusandtheAleppopine。Beyondthisisastoutbarricademadeupofrubbishofallkinds:bitsofgravel,scrapsofearth,juniper—needles,thecatkinsoftheconifers,smallshells,driedexcretionsofSnails。Nextcomeapartitionofpureresin,alargecocooninaroomychamber,asecondpartitionofpureresinand,lastly,asmallercocooninanarrowchamber。Theinequalityofthetwocellsisthenecessaryconsequenceoftheshapeoftheshell,whoseinnerspacegainsrapidlyinwidthasthespiralgetsnearertotheorifice。Thus,bythemeregeneralarrangementofthehomeandwithoutanyworkontheBee'spartbeyondsomeslenderpartitions,alargeroomismarkedoutinfrontandamuchsmallerroomattheback。
　　Byaveryremarkableexception,whichIhavementionedcasuallyelsewhere,themalesofthegenusAnthidiumaregenerallylargerthanthefemales;andthisisthecasewiththetwospeciesinparticularthatdividetheSnail'sspiralwithresinpartitions。Icollectedsomedozensofnestsofbothspecies。Inatleasthalfthecases,thetwosexeswerepresenttogether;thefemale,thesmaller,occupiedthefrontcellandthemale,thebigger,thebackcell。Othercells,whichweresmallerortoomuchobstructedatthebackbythedried—upremainsoftheMollusc,containedonlyonecell,occupiedatonetimebyafemaleandatanotherbyamale。Afew,lastly,hadbothcellsinhabitednowbytwomalesandnowbytwofemales。Themostfrequentarrangementwasthesimultaneouspresenceofbothsexes,withthefemaleinfrontandthemalebehind。TheAnthidiawhomakeresin—
　　doughandliveinSnail—shellscanthereforealternatethesexesregularlytomeettheexigenciesofthespiraldwelling—house。
　　OnemorethingandIhavedone。Myapparatusofreeds,fixedagainstthewallsofthegarden,suppliedmewitharemarkablenestoftheHornedOsmia。Thenestisestablishedinabitofreed11millimetreswideinside。（。429inch——Translator'sNote。）Itcomprisesthirteencellsandoccupiesonlyhalfthecylinder,althoughtheorificeispluggedwiththeusualstopper。Thelayingthereforeseemsheretobecomplete。
　　Well,thislayingisarrangedinamostsingularfashion。Thereisfirst,atasuitabledistancefromthebottomorthenodeofthereed,atransversalpartition,perpendiculartotheaxisofthetube。
　　Thismarksoffacellofunusualsize,inwhichafemaleislodged。
　　Afterthat,inviewoftheexcessivewidthofthetunnel,whichistoogreatforaseriesinsinglefile,theOsmiaappearstoalterhermind。Shethereforebuildsapartitionperpendiculartothetransversalpartitionwhichshehasjustconstructedandthusdividesthesecondstoreyintotworooms,alargerroom,inwhichshelodgesafemale,andasmaller,inwhichshelodgesamale。Shenextbuildsasecondtransversalpartitionandasecondlongitudinalpartitionperpendiculartoit。Theseoncemoregivetwounequalchambers,stockedlikewise,thelargeonewithafemale,thesmalleronewithamale。
　　>Fromthisthirdstoreyonwards,theOsmiaabandonsgeometricalaccuracy;thearchitectseemstobealittleoutinherreckoning。
　　Thetransversalpartitionsbecomemoreandmoreslantingandtheworkgrowsirregular,butalwayswithasprinklingoflargechambersforthefemalesandsmallchambersforthemales。Threefemalesandtwomalesarehousedinthisway,thesexesalternating。