Moreover，aswesaidthatdifferentmenaresubjecttoillusions，eachaccordingtothedifferentemotionpresentinhim，soitisthatthesleeper，owingtosleep，andtothemovementsthengoingoninhissensoryorgans，aswellastotheotherfactsofthesensoryprocess，[isliabletoillusion]，sothatthedreampresentation，thoughbutlittlelikeit，appearsassomeactualgiventhing.Forwhenoneisasleep，inproportionasmostofthebloodsinksinwardstoitsfountain[theheart]，theinternal[sensory]movements，somepotential，othersactualaccompanyitinwards.Theyaresorelated[ingeneral]that，ifanythingmovetheblood，someonesensorymovementwillemergefromit，whileifthisperishesanotherwilltakeitsplace;whiletooneanotheralsotheyarerelatedinthesamewayastheartificialfrogsinwaterwhichseverallyrise[infixedsuccesion]tothesurfaceintheorderinwhichthesalt[whichkeepsthemdown]becomesdissolved.Theresiduarymovementsarelikethese：theyarewithinthesoulpotentially，butactualizethemselvesonlywhentheimpedimenttotheirdoingsohasbeenrelaxed;andaccordingastheyarethussetfree，theybegintomoveinthebloodwhichremainsinthesensoryorgans，andwhichisnowbutscanty，whiletheypossessverisimilitudeafterthemannerofcloud—shapes，whichintheirrapidmetamorphosesonecomparesnowtohumanbeingsandamomentafterwardstocentaurs.Eachofthemishowever，ashasbeensaid，theremnantofasensoryimpressiontakenwhensensewasactualizingitself;andwhenthis，thetrueimpression，hasdeparted，itsremnantisstillimmanent，anditiscorrecttosayofit，thatthoughnotactuallyKoriskos，itislikeKoriskos.Forwhenthepersonwasactuallyperceiving，hiscontrollingandjudgingsensoryfacultydidnotcallitKoriskos，but，promptedbythis[impression]，calledthegenuinepersonyonderKoriskos.Accordingly，thissensoryimpulse，which，whenactuallyperceiving，it[thecontrollingfaculty]describes（unlesscompletelyinhibitedbytheblood），itnow[indreams]whenquasi—perceiving，receivesfromthemovementspersistinginthesense—organs，andmistakesit—animpulsethatismerelylikethetrue[objective]impression—forthetrueimpressionitself，whiletheeffectofsleepissogreatthatitcausesthismistaketopassunnoticed.Accordingly，justasifafingerbeinsertedbeneaththeeyeballwithoutbeingobserved，oneobjectwillnotonlypresenttwovisualimages，butwillcreateanopinionofitsbeingtwoobjects;
whileifit[thefinger]beobserved，thepresentationwillbethesame，butthesameopinionwillnotbeformedofit;exactlysoitisinstatesofsleep：ifthesleeperperceivesthatheisasleep，andisconsciousofthesleepingstateduringwhichtheperceptioncomesbeforehismind，itpresentsitselfstill，butsomethingwithinhimspeakstothiseffect：\'theimageofKoriskospresentsitself，buttherealKoriskosisnotpresent\';foroften，whenoneisasleep，thereissomethinginconsciousnesswhichdeclaresthatwhatthenpresentsitselfisbutadream.If，however，heisnotawareofbeingasleep，thereisnothingwhichwillcontradictthetestimonyofthebarepresentation.
Thatwhatwehereurgeistrue，i.e.thattherearesuchpresentativemovementsinthesensoryorgans，anyonemayconvincehimself，ifheattendstoandtriestoremembertheaffectionsweexperiencewhensinkingintoslumberorwhenbeingawakened.Hewillsometimes，inthemomentofawakening，surprisetheimageswhichpresentthemselvestohiminsleep，andfindthattheyarereallybutmovementslurkingintheorgansofsense.Andindeedsomeveryyoungpersons，ifitisdark，thoughlookingwithwideopeneyes，seemultitudesofphantomfiguresmovingbeforethem，sothattheyoftencoveruptheirheadsinterror.
Fromallthis，then，theconclusiontobedrawnis，thatthedreamisasortofpresentation，and，moreparticularly，onewhichoccursinsleep;sincethephantomsjustmentionedarenotdreams，norisanyotheradreamwhichpresentsitselfwhenthesense—perceptionsareinastateoffreedom.Noriseverypresentationwhichoccursinsleepnecessarilyadream.Forinthefirstplace，somepersons[whenasleep]actually，inacertainway，perceivesounds，light，savour，andcontact;feebly，however，and，asitwere，remotely.Fortherehavebeencasesinwhichpersonswhileasleep，butwiththeeyespartlyopen，sawfaintlyintheirsleep（astheysupposed）thelightofalamp，andafterwards，onbeingawakened，straightwayrecognizeditastheactuallightofareallamp;while，inothercases，personswhofaintlyheardthecrowingofcocksorthebarkingofdogsidentifiedtheseclearlywiththerealsoundsassoonastheyawoke.Somepersons，too，returnanswerstoquestionsputtotheminsleep.Foritisquitepossiblethat，ofwakingorsleeping，whiletheoneispresentintheordinarysense，theotheralsoshouldbepresentinacertainway.Butnoneoftheseoccurrencesshouldbecalledadream.Norshouldthetruethoughts，asdistinctfromthemerepresentations，whichoccurinsleep[becalleddreams].Thedreamproperisapresentationbasedonthemovementofsenseimpressions，whensuchpresentationoccursduringsleep，takingsleepinthestrictsenseoftheterm.
Therearecasesofpersonswhointheirwholeliveshaveneverhadadream，whileothersdreamwhenconsiderablyadvancedinyears，havingneverdreamedbefore.Thecauseoftheirnothavingdreamsappearssomewhatlikethatwhichoperatesinthecaseofinfants，and[thatwhichoperates]immediatelyaftermeals.Itisintelligibleenoughthatnodream—presentationshouldoccurtopersonswhosenaturalconstitutionissuchthatinthemcopiousevaporationisborneupwards，which，whenbornebackdownwards，causesalargequantityofmotion.Butitisnotsurprisingthat，asageadvances，adreamshouldatlengthappeartothem.Indeed，itisinevitablethat，asachangeiswroughtintheminproportiontoageoremotionalexperience，thisreversal[fromnon—dreamingtodreaming]shouldoccuralso.