——Theonlyquestionthereforeis，howfaritshouldbebroughtforward？——Theanswerisshort，andeasytobeunderstood；——bringitforwardasfaraspossible，withoutdiminishingtoomuchthepassagewhichmustbeleftforthesmoke。Nowasthispassage，which，initsnarrowestpart，IhavecalledtheTHROATOFTHE
CHIMNEY，ought，forreasonswhicharefullyexplainedintheforegoingChapter，tobeimmediately，orperpendicularlyovertheFire，itisevidentthatthebackoftheChimneymustalwaysbebuiltperfectlyupright——Todeterminethereforetheplaceforthenewback，orhowfarpreciselyitoughttobebroughtforward，nothingmoreisnecessarythantoascertainhowwidethethroatoftheChimneyoughttobeleft，orwhatspacemustbeleft，betweenthetopofthebreastoftheChimney，wheretheuprightcanaloftheChimneybegins，andthenewbackoftheFire-placecarriedupperpendicularlytothatheight。
InthecourseofmynumerousexperimentsuponChimnies，Ihavetakenmuchpainstodeterminethewidthpropertobegiventothispassage，andIhavefound，that，whenthebackoftheFire-placeisofaproperwidth，thebestwidthforthethroatofaChimney，whentheChimneyandtheFire-placeareattheusualformandsize，isFOURINCHES——Threeinchesmightsometimesanswer，especiallywheretheFire-placeisverysmall，andtheChimneygood，andwellsituated：butasitisalwaysofmuchimportancetopreventthoseaccidentalpuffsofsmokewhicharesometimesthrownintoroomsbythecarelessnessofservantsinputtingonsuddenlytoomanycoalsatonceuponthefire，andasIfoundtheseaccidentssometimeshappenedwhenthethroatsofChimneysweremadeverynarrow，Ifoundthat，uponthewhole，allcircumstancesbeingwellconsidered，andadvantagesanddisadvantagescomparedandbalanced，FOURINCHESisthebestwidththatcanbegiventothethroatofachimney；andthis，whethertheFire-placebedestinedtoburnwood，coals，turf，oranyotherfuelcommonlyusedforheatingroomsbyanopenfire。
InFire-placesdestinedforheatingverylargehalls，andwhereverygreatfiresarekeptup，thethroatoftheChimneymay，ifitshouldbethoughtnecessary，bemadefourinchesandanhalf，orfiveincheswide；——butIhavefrequentlymadeFire-placesforhallswhichhaveansweredperfectlywellwherethethroatsoftheChimnieshavenotbeenwiderthanfourinches。
Itmayperhapsappearextraordinary，uponthefirstviewofthematter，thatFire-placesofsuchdifferentsizesshouldallrequirethethroatoftheChimneytobeofthesamewidth；butwhenitisconsideredthattheCAPACITYofthethroatofaChimneydoesnotdependonitswidthalone，butonitswidthandLENGTHtakentogether；andthatinlargeFire-places，thewidthoftheback，andconsequentlythelengthofthethroatoftheChimney，isgreaterthaninthosewhicharesmaller，thisdifficultyvanishes。