Intheremainingcasesthesubjectsareliabletowritetheselectednameanywhere，generallyfirstorlast。NowmyobjectistosomanipulatemysubjectsastocausethemtowritetheselectednamewhenIwantthemtodoso。Thisisdonebycontinuoustalking，anddistractingtheirattentionuntilthepropermoment。
Ichoosethethirdspace，sincethis，beingtheonetheyaremostliabletochooseoftheirownaccord，istheeasiesttoforce。
Justastheybegintowritethefirstname，beforetheymakeamark，Isaysuddenly，“NowbesureandselectnamesoflivingpersonsthatIcouldnotpossiblyknow。“Thisisalmostcertaintoinsureapause，andthenameofalivingpersontobewrittenfirst。Icontinuemytalkinginanaturalmanner，takingtheattentiontoagreatextentfromthewriting，andnearlyalwaysobservinganotherpausejustbeforewritingthesecondname。WhenthesecondnameisalmostfinishedIexclaimsuddenly，“Nowwriteasrapidlyaspossible!“Ifthesubjectshavebeenproperlyimpressedwiththeseriousnessoftheexperiment，theywillalmostinvariably，onfinishingthesecondnameinobediencetomycommand“tobeasrapidaspossible，“andintheirdesiretopleaseme，hurryintothenamealreadyintheirminds，thuswritingtheselectednameinthethirdplace。Ifsuchisthecasetheywillnowmostsurelypausetothinkofafourthname。Ifso，IamcertainthatInowknowtheselectedname。However，iftheyshouldrapidlypassintothefourthname，itisthenuncertainwhethertheselectednameisinthethirdorfourthspace。This，however，seldomhappensifworkedinanexpertmanner。
Inrarecasesthesubjectcannotbemanipulatedbytheperformer，inwhichcaseitispurelyguesswork；eveninsuchcases，however，Istandonechanceinsixofsucceeding；andifImakeasecondtrialonfailingnotuncommonwithmediums，Istandonechanceinthreeofsucceeding。
ItishardlyworthwhiletosaythatasIfoldthebillets，Ifoldthethirdoneslightlydifferentfromtherest，sothatwhileitwillnotattractattention，Icanseeataglancewhatitiswhenthrownonthetable。Imemorizethename；also，ifindoubt，I